Processing defects detected on extreme ultraviolet masks

Abstract

The invention relates to processing defects detected on an extreme ultraviolet light mask. Methods and systems for photomask defect handling are provided. A method includes directing energy to a photomask and detecting energy from the photomask. The photomask is configured for use at one or more extreme ultraviolet wavelengths. The method also includes detecting a defect on the photomask based on the detected energy. In addition, the method includes generating a charged particle beam image of the photomask at the location of the detected defect. The method further includes processing the detected defect based on the charged particle beam image generated for the detected defect.

Description

[0001] Related information of divisional application

[0002] This application is a divisional application of the application patent application with the application date of September 19, 2019, the application number of 201980060748.1, and the invention name of “Processing defects detected on extreme ultraviolet photomasks”. TECHNICAL FIELD

[0003] The present disclosure relates generally to the field of photomask inspection and defect processing. More specifically, the present disclosure relates to methods and systems for inspecting and re-inspecting photomasks designed for use at extreme ultraviolet wavelengths. BACKGROUND

[0004] The following description and examples are not admitted to be prior art by virtue of their

[0005] Fabricating semiconductor devices such as logic and memory devices typically includes processing substrates such as semiconductor wafers using a large number of semiconductor fabrication processes to form various features and multiple levels of the semiconductor devices. For example, lithography is a semiconductor fabrication process that involves transferring patterns from a photomask to a resist arranged on a semiconductor wafer. Additional examples of semiconductor fabrication processes include, but are not limited to, chemical-mechanical polishing (CMP), etch, deposition, and ion implantation. Multiple semiconductor devices can be fabricated in an arrangement on a single semiconductor wafer and then separated into individual semiconductor devices.

[0006] Extreme ultraviolet (EUV) lithography has been widely developed as the performance of 193 nm immersion lithography reaches its limits and there are substantial high cost and yield problems associated with multi-patterning lithography processes. The EUV lithography is a leading candidate for next generation lithography (NGL) technology to extend Moore’s law to drive computer chips to be smaller, faster, and more efficient.

[0007] From a process yield management perspective, defectivity control of EUV photomasks that define patterns printed on wafers plays a critical role. However, defectivity control has been one of the high-risk areas of EUV lithography development due to the lack of actinic EUV photomasks or high-throughput charged-particle beam inspection tools that can inspect photomasks with the required resolution. Currently and for the foreseeable future, inspection of patterned EUV photomasks can have to rely on available inspection tools that operate in the deep ultraviolet (DUV) wavelength range (e.g., 193 nm to 257 nm) individually.

[0008] For EUV photomasks, such a significant wavelength difference between photomask use and photomask inspection has a significant impact on the performance of the DUV inspection tool. In comparison, older inspection methods for DUV masks had sufficient defect and pattern resolution at the 193 nm inspection wavelength, so there was effectively no need for separate re-inspection. Currently, EUV photomask inspection performed with DUV inspection tools aims to produce defects that are about a few hundred times inspected, so that subsequent re-inspection of these defects does not cause a nuisance for a scanning electron microscope (SEM) tool or operator to manually re-inspect each defect. DUV inspection tools also have limited resolution on mask dimensions targeted for EUV patterning. For example, the theoretical resolution limit of a DUV mask inspection tool is about 60 nm half pitch (HP) on 4X masks, which means that the main feature size on EUV masks will be far beyond the practical resolution of the DUV inspection tool.

[0009] However, DUV inspection tools with various illumination conditions are being explored to maximize defect sensitivity and / or main pattern modulation for initial EUV defect detection. Due to the generally low signal-to-noise ratio in DUV inspection imaging of EUV masks, these inspections typically result in hundreds or thousands of defects being detected. And if one wants to extend the detection capability even further, there can be tens of thousands of inspections per DUV inspection. Then accurate re-inspection and processing of each of these inspections is needed. Visual re-inspection of each defect is difficult due to poor DUV resolution, and the risk of manual defect misclassification rises with defect count. Furthermore, the lack of reliable aerial image processing systems makes processing EUV mask defects for printability even more challenging.

[0010] Therefore, it would be advantageous to develop a method and / or system for EUV photomask defect processing that does not have one or more of the drawbacks described above. SUMMARY

[0011] The following description of various embodiments should not be considered to be limiting the subject matter of the appended claims in any way.

[0012] One embodiment relates to a photomask inspection system including an inspection subsystem configured for directing energy to a photomask and detecting energy from the photomask. The photomask is configured for use at one or more extreme ultraviolet (EUV) light wavelengths. The system also includes one or more computer subsystems configured for detecting defects on the photomask based on the detected energy. In addition, the system includes a charged particle beam subsystem configured for generating charged particle beam images of the photomask at locations of the detected defects determined by the one or more computer subsystems. The one or more computer subsystems are configured for processing the detected defects based on the charged particle beam images generated for the detected defects. A photomask inspection system can be further configured as described herein.

[0013] Another embodiment relates to a method for inspecting a photomask. The method includes directing energy to a photomask and detecting energy from the photomask. The photomask is configured for use at one or more EUV light wavelengths. The method also includes detecting defects on the photomask based on the detected energy. In addition, the method includes generating charged particle beam images of the photomask at locations of the detected defects. The method further includes processing the detected defects based on the charged particle beam images generated for the detected defects.

[0014] The method described above can be performed as further described herein. In addition, the method described above can include any other steps of any other methods described herein. Moreover, the method described above can be performed by any of the systems described herein.

[0015] Another embodiment relates to a non-transitory computer-readable medium storing program instructions executable on a computer system for performing a computer-implemented method for inspecting a photomask. The computer-implemented method includes the steps of the method described above. The computer-readable medium can be further configured as described herein. The steps of the computer-implemented method can be performed as further described herein. In addition, the computer-implemented method that can be used to execute the program instructions can include any other steps of any other methods described herein. BRIEF DESCRIPTION OF DRAWINGS

[0016] Other objects and advantages of the present application will become apparent upon reading the following detailed description and upon inspecting the accompanying drawings in which:

[0017] Figure 1 is a schematic diagram illustrating a side view of one embodiment of a photomask inspection and defect processing system;

[0018] Figure 2 is a schematic illustration of a side view illustrating one embodiment of an optional subsystem that can be included in embodiments of photomask inspection systems;

[0019] Figure 3 is a schematic illustration of a side view illustrating one example of a portion of an extreme ultraviolet (EUV) photomask;

[0020] Figure 4 is a schematic illustration of a plan view illustrating one example of a portion of an EUV photomask pattern that does not have any defects;

[0021] Figure 5 is a schematic illustration of a plan view illustrating a portion of an EUV photomask pattern of Figure 4 is a schematic illustration of a plan view illustrating a portion of an EUV photomask pattern of

[0022] Figure 6 is a flowchart illustrating one embodiment of steps that can be performed by embodiments described herein for processing defects detected on EUV photomasks performed using a charged particle beam subsystem;

[0023] Figure 7 is a schematic illustration of a side view illustrating one example of a portion of an EUV photomask having different examples of buried defects formed thereon;

[0024] Figure 8 is a flowchart illustrating one embodiment of steps that can be performed by embodiments described herein for processing defects detected on EUV photomasks performed using an atomic force microscope subsystem; and

[0025] Figure 9 is a block diagram illustrating one embodiment of a non-transitory computer- readable medium storing program instructions executable on a computer system for performing one or more of the computer-implemented methods described herein.

[0026] While the application is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the drawings and detailed description thereto are not intended to limit the application to the particular form disclosed, but on the contrary, the intention is to cover all modifications, equivalents and alternatives falling within the spirit and scope of the application as defined by the appended claims. DETAILED DESCRIPTION

[0027] The terms "design" and "design data" as used herein refer to information and data generated by a semiconductor device designer in a design process and thus can be well used in the embodiments described herein prior to printing the design on any physical substrate. Thus, the terms "design" and "design data" as used herein generally refer to a physical design (layout) of an IC and data derived from the physical design by complex simulation or simple geometric construction and Boolean operations. The physical design can be stored in a data structure such as a Graphic Data Stream (GDS) file, any other standard machine-readable file, any other suitable file known in the art, and design databases. A GDSII file is one of a class of files used for representing design layout data. Other examples of such files include GL1 and OASIS files, and proprietary file formats such as RDF data specific to KLA of Milpitas, California. The design can include any other design data or design data agents described in commonly owned U.S. Patent No. 7,570,796 issued to Zafar et al. on August 4, 2009, and commonly owned U.S. Patent No. 7,676,077 issued to Kulkarni et al. on March 9, 2010, both of which are incorporated by reference as if fully set forth herein. Additionally, the design data can be standard cell library data, integrated layout data, design data for one or more layers, derivatives of design data, and all or part of a chip design data.

[0028] The "design" or "physical design" can also be a design that is ideally formed on a photomask. In this manner, a design as described herein can include all features of the design printed on a photomask, including any optical proximity correction (OPC) features, added to the design to enhance printing of features on a wafer without actually printing themselves.

[0029] Turning now to the drawings, it is noted that the various figures are not drawn to scale. In particular, the proportions of some of the elements of the figures have been greatly exaggerated to emphasize the features of the elements. It is also noted that the figures are not drawn to the same scale. The same reference numbers have been used to indicate elements that can be similarly configured shown in more than one figure. Unless otherwise mentioned herein, any of the described and shown elements can include any suitable commercially available element.

[0030] Embodiments described herein generally relate to extreme ultraviolet (EUV) photomask defect inspection using thermal inspection (e.g., deep ultraviolet (DUV) (e.g., 193 nm) optical inspection), followed by substantially fast defect processing using secondary charged particle beam images. EUV lithography is the leading technology for next generation lithography (NGL), whose feasibility depends on the availability of a proper functioning EUV photomask inspection tool that can capture all critical defects affecting yield. The use of existing photomask inspection tools can be extended to EUV photomask inspection by running the existing photomask inspection tools in "thermal" to achieve the required defect sensitivity (although a substantially high number of defects will be detected in the inspection), and then processing each detected defect (or at least a majority of the detected defects) using charged particle beam imaging (e.g., scanning electron microscope (SEM)) and possibly atomic force microscope (AFM) based on classification and printability modeling. As further described herein, the embodiments can be configured to process the detected defects using charged particle beam images and optionally AFM images. In addition, some embodiments described herein are configured for an automatic algorithm as well as a workflow of charged particle beam imaging / analysis and optionally AFM analysis, including defect isolation, classification, and printability analysis.

[0031] One embodiment relates to a photomask inspection system. Figure 1 One embodiment of this system is shown in FIG. 1. The photomask inspection system includes an inspection subsystem configured for directing energy to a photomask and detecting energy from the photomask. In one embodiment, the energy directed to the photomask by the inspection subsystem includes one or more DUV wavelengths of light. The photomask is configured for use at one or more EUV wavelengths of light. In other words, the photomask is configured for use in a lithography process that uses one or more EUV wavelengths of light to transfer a pattern from the photomask to a wafer (e.g., by reflecting EUV light from the photomask to the wafer). In this way, the actinic wavelength of the photomask in the embodiments described herein (i.e., the wavelength of light used to transfer a pattern from the photomask to a wafer to thereby cause a photochemical reaction in one or more materials (e.g., photoresist) on the wafer) can be different from the wavelength of light used for photomask inspection by the embodiments described herein.

[0032] In one embodiment, the energy directed to the photomask by the inspection subsystem includes light having a wavelength of 193 nm. In another embodiment, the energy directed to the photomask by the inspection subsystem includes light having one or more wavelengths in a range from 193 nm to 257 nm. In an additional embodiment, the energy directed to the photomask by the inspection subsystem includes light having a wavelength of 13.5 nm (or another EUV wavelength).

[0033] As shown in Figure 1 Figure 1 While the photomask inspection system is shown in Figure 1 as including one inspection subsystem, it should be understood that the photomask inspection system can include only one inspection subsystem or more than one inspection subsystem, such as the inspection subsystem 100 shown in Figure 2 and the inspection subsystem 200 shown in and further described herein.

[0034] As further shown in Figure 1 the inspection subsystem 100 includes a light source 102. The light source 102 can include any suitable light source known in the art, such as a laser. The light source 102 is configured to direct light to a beamsplitter 104, which is configured to reflect light from the light source 102 to a refractive optical element 106. The refractive optical element 106 is configured to focus light from the beamsplitter 104 to a photomask 108. The beamsplitter 104 can include any suitable beamsplitter, such as a 50 / 50 beamsplitter. The refractive optical element 106 can include any suitable refractive optical element, and while the refractive optical element 106 is shown in Figure 1 as a single refractive optical element, it can be replaced with one or more refractive optical elements and / or one or more reflective optical elements.

[0035] Thus, the light source 102, the beamsplitter 104, and the refractive optical element 106 can form an illumination channel of the inspection subsystem. The illumination channel can include any other suitable elements (not shown in Figure 1 ), such as one or more polarizing components, a diffractive optical element (DOE), and one or more filters, such as spectral filters. As shown in Figure 1 the light source, the beamsplitter, and the refractive optical element are configured so that light is directed to the photomask at a normal or substantially normal angle of incidence. The inspection subsystem can be configured so as to scan the light over the photomask in any suitable manner.

[0036] Light reflected from the photomask 108 as a result of the illumination can be collected by the refractive optical element 106 and directed through the beamsplitter 104 to a detector 110. Thus, the refractive optical element, the beamsplitter, and the detector can form a detection channel of the inspection subsystem. The detector can include any suitable imaging detector known in the art, such as a charge-coupled device (CCD) or a time delay integrator (TDI). This detection channel can also include one or more additional components (not shown in Figure 1 ​One or more polarizing components, one or more spatial filters, one or more spectral filters, etc. are not shown, for example. The detector 110 is configured to generate an output in response to the reflected light detected by the detector. The output can include a signal, signal data, an image, image data, and any other suitable output.

[0037] The inspection subsystem can be configured to have more than one mode in any suitable manner. In some examples, the inspection subsystem can have more than one mode sequentially (e.g., by changing one or more parameters of the imaging lens of the inspection subsystem, such as numerical aperture (NA), between scans of the photomask). Further, the inspection subsystem can scan the photomask in some modes simultaneously and in other modes sequentially. The photomask inspection system can be configured to control the optical mode used for any scan of any photomask in any suitable manner.

[0038] The photomask inspection system can include several other components not shown in Figure 1 For example, the system can include a load module, an alignment module, a handler such as a robotic transfer arm, and an environmental control module and can include any such components known in the art.

[0039] The photomask inspection system also includes one or more computer subsystems configured for detecting defects on the photomask based on the detected energy. The computer subsystems can be configured to detect defects in one or more different ways. For example, the computer subsystems can be configured to compare the output generated by the detector of the inspection subsystem in response to the detected energy (e.g., an image, image data, etc.) to corresponding design information for the photomask (e.g., GDS or other design data that can be stored in a design database). Thus, this defect detection is commonly referred to as die-to-database type of inspection. The results of the comparison can be compared to one or more thresholds. Output above the threshold(s) can be identified as corresponding to a defect or possible defect, and output not above the threshold(s) can not be identified as corresponding to a defect or possible defect. In another example, the output generated by the detector of the inspection subsystem in response to the detected energy can be compared to output generated at the same location in a different die on the photomask by the computer subsystems to each other. Thus, this defect detection is commonly referred to as die-to-die type of inspection. Such comparison results can be used to identify defects on the photomask as described above.

[0040] In one embodiment, the computer subsystem is configured for detecting defects using a hot threshold. For example, the threshold used in the defect detection described above can be a "hot" threshold. A "hot" threshold is generally defined as a threshold at, within, or substantially near the noise floor of the output generated by the inspection subsystem of the photomask. In this way, the defect detection can be more aggressive (hotter) than the defect detection that would normally be performed for a tuned inspection recipe, such that more events, including defects and nuisance events, are detected than would be expected in a tuned inspection. In this way, the inspection performed with this threshold is generally not useful for production monitoring due to the substantially high nuisance defect detection. An inspection performed with this threshold is generally referred to as a "hot" inspection, and a scan performed on a photomask during this inspection can be generally referred to as a "hot" scan.

[0041] In one embodiment, as shown in FIG. 1, Figure 1 The photomask inspection system includes a computer subsystem 116 coupled to the inspection subsystem 100. For example, the computer subsystem can be coupled to a detector (e.g., detector 110) of the inspection subsystem (e.g., by one or more transmission media, which can include any suitable transmission media known in the art, shown by dashed lines in FIG. 1). The computer subsystem can be coupled to the detector in any suitable manner such that the output (e.g., images) of the photomask generated by the inspection subsystem and any other information can be sent to the computer subsystem, and optionally such that the computer subsystem can send instructions to the inspection subsystem to perform one or more steps. Figure 1

[0042] This computer subsystem (as well as other computer subsystems described herein) can also be referred to herein as a computer system. Each of the computer subsystems or systems described herein can take various forms, including a personal computer system, image computer, mainframe computer system, workstation, network appliance, Internet appliance, or other device. In general, the term "computer system" can be broadly defined to encompass any device having one or more processors, which executes instructions from a memory medium. The computer subsystem or system can also include any suitable processor known in the art, such as a parallel processor. Additionally, the computer subsystem or system can include a computer platform having high speed processing and software as a standalone or networked appliance.

[0043] If the system includes more than one computer subsystem, the different computer subsystems can be coupled to each other such that images, data, information, instructions, etc. can be sent between the computer subsystems, as described further herein. For example, the computer subsystem 116 can be coupled to the computer subsystem 118 by any suitable transmission media, which can include any suitable wired and / or wireless transmission media known in the art, as by one or more transmission media, which can include any suitable transmission media known in the art, shown by dashed lines in FIG. 1. The computer subsystem 118 can be coupled to the computer subsystem 120 by any suitable transmission media, which can include any suitable wired and / or wireless transmission media known in the art, as by one or more transmission media, which can include any suitable transmission media known in the art, shown by dashed lines in FIG. 1. Figure 1 ​Two or more of such computer subsystems can also be effectively coupled by sharing a computer readable storage medium (not shown).

[0044] The photomask inspection system also includes a charged particle beam subsystem configured for generating a charged particle beam image of the photomask at the location of the detected defect determined by the one or more computer subsystems. In one embodiment, the charged particle beam subsystem is configured as an electron beam subsystem. For example, as shown in Figure 1 The electron beam subsystem can include an electron column 122 coupled to the computer subsystem 134, as shown in

[0045] The electron column includes an electron beam source 124 configured to generate electrons focused through one or more elements 126 to the photomask 128. The electron beam source can include, for example, a cathode source or emitter tip, and the one or more elements 126 can include, for example, a gun lens, an anode, a limiting aperture, a gate valve, a beam current selection aperture, an objective lens, and a scanning subsystem, all of which can include any such suitable elements known in the art.

[0046] Electrons (e.g., secondary electrons) returned from the photomask can be focused through one or more elements 130 to a detector 132. The one or more elements 130 can include, for example, a scanning subsystem, which can be the same scanning subsystem included in the elements 126.

[0047] The electron column can include any other suitable elements known in the art. Additionally, the electron column can be further configured as described in U.S. Patent No. 8,664,594 to Jiang et al., issued April 4, 2014, U.S. Patent No. 8,692,204 to Kojima et al., issued April 8, 2014, U.S. Patent No. 8,698,093 to Gubbens et al., issued April 15, 2014, and U.S. Patent No. 8,716,662 to MacDonald et al., issued May 6, 2014, which are incorporated by reference as if fully set forth herein.

[0048] While the electron column is shown in Figure 1 as being configured such that electrons are directed to the photomask at a tilted angle of incidence and scattered from the photomask at another tilted angle, it should be understood that the electron beam can be directed to the photomask and scattered from the photomask at any suitable angle. Additionally, the electron beam subsystem can be configured to generate images of the photomask using multiple modes (e.g., at different illumination angles, collection angles, etc.). The multiple modes of the electron beam subsystem can be different in any image generation parameter of the subsystem.

[0049] In another embodiment, the charged particle beam subsystem is configured as an ion beam subsystem. For example, in the electron column shown in Figure 2 In the electron column shown in

[0050] The computer subsystem 134 can be coupled to the detector 132 as described above. The detector can detect charged particles returning from the surface of the photomask, thereby forming a charged particle beam image of the photomask. The charged particle beam image can include any suitable charged particle beam image. The computer subsystem 134 can be configured to perform any of the functions described herein using the output of the detector and / or the charged particle beam image. The computer subsystem 134 can be configured to perform any additional step described herein. The computer subsystem 134 can be further configured as described herein.

[0051] While in the description of some of the embodiments provided herein, the charged particle beam subsystem can be referred to as a scanning electron microscope (SEM) and the charged particle beam image can be referred to as an "SEM image", all of the embodiments described herein are not limited to one or several SEM images. For example, the charged particle beam subsystem can have any suitable configuration for generating a charged particle beam image, including SEMs and other types of electron beam tools (e.g., a transmission electron microscope (TEM)). Additionally, the charged particle beam subsystem included in the system can include a commercially available electron beam tool, such as the commercially available mask DR-SEM E5600 series and mask MVM-SEM® E3600 series from Advantest America, Inc. of San Jose, California.

[0052] The computer subsystems included in the system can also include computer subsystems that are not coupled to the inspection or charged particle beam subsystem, such as the computer subsystem 118. In this way, one of the computer subsystems can be a computer subsystem that can be coupled to Figure 1The standalone computer subsystems can be configured to acquire images described herein and perform other steps described herein. In one such example, computer subsystems 116 and 134 can be configured to store images received from the inspection and charged particle beam subsystems to which they are respectively coupled and store the images in storage medium 120 which can be further configured as described above. The standalone computer subsystems can then acquire the images from the storage medium and use some combination of the images to perform one or more steps described herein.

[0053] As described above, the photomask inspection system can include more than one inspection subsystem. In embodiments described herein, Figure 1 The inspection subsystems shown in FIG. 1 can also be used in combination with Figure 2 The inspection subsystems shown in FIG. 1 can also be used in combination with Figure 2 The inspection subsystems shown in FIG. 1.

[0054] As Figure 1 As shown in FIG. 2, inspection subsystem 200 includes an illumination subsystem and a collection subsystem as described in greater detail herein. The illumination subsystem includes a light source 202. Light source 202 can be a coherent light source, such as a laser. The light source can be configured to emit monochromatic light having a wavelength of 248 nm, 193 nm, and / or another DUV or EUV wavelength described herein. Alternatively, the light source can be configured to emit light having a range of wavelengths and can be coupled to a spectral filter (not shown). Examples of broadband light sources include, but are not limited to, He-Xe arc lamps that produce light in the DUV wavelength range. In this manner, the light source and filter can emit monochromatic light having a wavelength as described above. The light source can be configured to emit light continuously or in pulses at various time intervals.

[0055] The illumination subsystem can also include several optical components coupled to the light source. For example, light from the light source 202 can first pass through a homogenizer 204. The homogenizer 204 can be configured to reduce the speckle of the light from the light source. The illumination subsystem can also include a diaphragm 206. The diaphragm 206 can have an adjustable NA. For example, the diaphragm can be coupled to a control mechanism that can be configured to mechanically alter the diaphragm depending on a control signal received from a user or in accordance with program instructions received from a prescription at a program running on the system. In this way, the light can have various partial coherence factors, σ. For example, the diaphragm 206 can be altered to adjust the pupil of a condenser lens 208. The pupil of the condenser lens controls the NA of the system. As the pupil of the condenser is decreased, the coherence of the illumination is increased, thereby decreasing the value of σ. The value of σ can be expressed as the ratio of the NA of the condenser lens to the NA of the objective lens. The exposure system can have a value of σ in a range between about 0.3 to about 0.9. Thus, the diaphragm 206 can be altered so that the inspection subsystem has a value of σ between about 0.3 and about 0.9. The value of σ can be altered depending on the features on the photomask. For example, a higher value of σ can be used if the photomask includes lines and spaces than if the photomask includes contact holes. The control mechanism can also be configured to alter the diaphragm to provide annular or off-axis illumination. The diaphragm can also be configured to provide other types of illumination such as quadrupole or dipole illumination. The diaphragm can further be configured to alter the shape of the light beam. For example, the diaphragm can be a diffractive optical element or an apodized diaphragm.

[0056] The illumination subsystem can also include several additional optical components (not shown). For example, the illumination subsystem can also include a telescope configured to alter the beam diameter of the light. Additionally, the illumination subsystem can include one or more relay lenses, additional lenses such as field lenses, folding mirrors, additional diaphragms, and beam splitters.

[0057] The illumination subsystem can also include a condenser lens 208. The condenser lens 208 can be configured to change the diameter of the light in the object (reticle) plane to be approximately or larger than the field of view of the subsystem. The light exiting the condenser lens can be directed to a beamsplitter 209, which directs the light from the condenser lens to a reticle 210 to thereby illuminate the reticle 210 supported on a stage 212. The beamsplitter 209 can include any suitable beamsplitter known in the art. The stage is configured to support the reticle by contacting the reticle proximate to its outer lateral edges. The stage 212 can be configured to move the reticle so that the alignment of the reticle can be changed and so that light can be scanned across the reticle. Alternatively, the illumination system can include a scanning element (not shown), such as an acousto-optic deflector or a mechanical scanning assembly, so that the reticle can remain substantially stationary while light is scanned across the reticle. The stage 212 can also be configured to move the reticle through the focal point, thereby changing the focal setting of the inspection subsystem. The stage can also be coupled to an autofocus device (not shown) configured to change the position of the stage, thereby changing the position of the reticle to maintain the focal setting of the inspection subsystem during inspection. Alternatively, the autofocus device can be coupled to the objective lens to change the position of the objective lens to maintain the focal setting during inspection.

[0058] The inspection subsystem can also include several optical components arranged to form a collection subsystem. For example, the collection subsystem includes an objective lens 214. Light reflected by the reticle is collected by the objective lens 214. The collection subsystem also includes an aperture 216 having an adjustable NA. The NA of the aperture 216 can also be selected so that the light exiting the aperture has a selected magnification. The aperture 216 is positioned between the objective lens 214 and a lens 218, which can be configured as a tube lens. Light from the lens 218 can be directed to a beamsplitter 220. The beamsplitter 220 can be configured to direct the light to three detectors 222, 224, and 226. The collection subsystem can also include several additional optical components (not shown), such as magnification lenses. The magnification lenses can be positioned between the lens 218 and the beamsplitter 220.

[0059] The detectors 222, 224, and 226 can be configured to form an image of the light reflected by the illuminated portion of the reticle. This image can be referred to as an "airborne image." The detectors can include, for example, CCD or TDI cameras. The detectors can also have one- or two-dimensional pixel arrays. Each of the three detectors can have a different focal setting. In this way, the three detectors can form images of the reticle substantially simultaneously at three different focal settings. For example, one detector can be substantially in focus, and the other two detectors can be out of focus in opposite directions relative to the in-focus condition. In addition, the inspection subsystem can include any number of such detectors depending on the mechanical or physical constraints of the inspection subsystem.

[0060] Alternatively, the inspection subsystem can include only one detector configured to form an image of the photomask. The detector can have a focal setting approximately equal to the focal setting of the exposure system. After each image is formed, the image of the photomask can be formed at a different focal setting by changing the focal setting of the detector. In this embodiment, the beam splitter 220 will not have to split the light to multiple detectors.

[0061] The computer subsystem 228 can be coupled to the inspection subsystem 200. For example, the computer subsystem can be coupled to the detectors of the inspection subsystem (e.g., detectors 222, 224, and 226) (e.g., by a wired or wireless connection, by a network, etc.). The computer subsystem can be coupled to the inspection subsystem in any other suitable manner such that the images of the photomask and any other information generated by the inspection subsystem can be sent to the computer subsystem, and optionally such that the computer subsystem can send instructions to the inspection subsystem to perform one or more steps described herein. Figure 1 The computer subsystem can be coupled to the detectors in any suitable manner. The computer subsystem can be coupled to the inspection subsystem in any other suitable manner such that the images of the photomask and any other information generated by the inspection subsystem can be sent to the computer subsystem, and optionally such that the computer subsystem can send instructions to the inspection subsystem to perform one or more steps described herein.

[0062] Note that the inspection and charged particle beam subsystems described herein Figure 1 and 2 to generally illustrate some configurations of inspection and charged particle beam subsystems that can be included in embodiments described herein. Obviously, the configurations of the inspection and charged particle beam subsystems described herein can be altered to optimize the performance of the systems, as is typically performed when designing a commercial inspection system. In addition, the photomask inspection systems described herein can be implemented using existing inspection and charged particle beam subsystems (e.g., photomask inspection tools commercially available from KLA) (e.g., by adding the functionality described herein to an existing inspection and / or charged particle beam inspection system). For some such systems, the embodiments described herein can be provided as optional functionality of the systems (e.g., in addition to other functionality of the systems). Alternatively, the photomask inspection systems described herein can be designed "from scratch" to provide a completely new system.

[0063] As described above, the inspection subsystem can be configured as an optical inspection subsystem configured to scan the photomask with light having one or more wavelengths. However, the inspection subsystem can be a different type of inspection subsystem. For example, in one embodiment, the energy directed by the inspection subsystem to the photomask includes electrons. In another embodiment, the energy directed by the inspection subsystem to the photomask includes ions. In such embodiments, the inspection subsystem can be configured in a similar manner as shown by the electron column 122 in Figure 1 The inspection subsystem can be configured in a similar manner as shown by the electron column 122 in Figure 1The optical inspection subsystems shown in the figures can be replaced with electron beam or ion beam inspection subsystems, and the system can include two charged particle beam subsystems (one for inspection and the other for generating charged particle beam images of detected defects).

[0064] The two charged particle beam subsystems can differ in one or more parameters, such that one of the subsystems is particularly suitable for inspection and the other is particularly suitable for generating charged particle beam images. For example, the two subsystems can differ in resolution capability (such that the subsystem used for inspection has a lower resolution capability than one of the subsystems used for charged particle beam imaging). In another alternative, the system can include one charged particle beam subsystem for both inspection and charged particle beam imaging, and one or more parameters of the subsystem can be varied between inspection and imaging, such that the subsystem can be used for both tasks. For example, if Figure 3 If the charged particle beam subsystems shown in the figures are used for imaging rather than for inspection, they can be configured to have a higher resolution. In other words, Figure 3 Embodiments of the charged particle beam subsystems shown in the figures describe some general and various configurations of charged particle beam subsystems that can be adapted in several ways that will be apparent to those skilled in the art to produce subsystems with different imaging capabilities that are more or less suitable for different applications. Electron beam inspection of photomasks can also be performed as described in Naka et al., "Capability of Model EBEYE M for EUV Mask Production," SPIE Photomask Technology 2012 Proceedings Vol. 8522, p. 14, published November 8, 2012, which is incorporated by reference as if fully set forth herein. Embodiments described herein can be further configured as described in the figures with reference to Naka et al.

[0065] In Figure 6 A typical EUV photomask stack is illustrated in the figure. The mask substrate (not shown) is covered with 40 pairs of molybdenum (Mo) / silicon (Si) multilayer (ML) thin films, capped with a relatively thin layer of ruthenium (Ru), in Figure 6 The dual layer film of tantalum boron nitride (TaBN) 302 and tantalum boron oxide (TaBO) 304, shown collectively as thin film 300 in the figure, acts as an absorber that is selectively etched to form the mask pattern. EUV mask inspection can be done by a DUV inspection tool to retrieve defect sites. When the DUV inspection tool is used in a relatively high sensitivity mode, the number of defect sites can be in the tens of thousands due to the relatively low signal-to-noise ratio (S / N or SNR) generated in the DUV inspection tool for EUV mask defects.

[0066] The computer subsystems are configured for processing the detected defects based on charged particle beam images generated for the detected defects. As used herein, the term "processing" is defined as determining additional information for the detected defects that can be used to make a final decision on how to properly handle the detected defects (e.g., the detected defects should be repaired, the detected defects are nuisance and can be ignored, the detected defects do not need to be repaired but should be monitored for the impact on the wafer pattern formed with the photomask for the detected defects, etc.). For example, the secondary charged particle beam (electron beam or ion) imaging of the mask described herein provides higher resolution than inspection with relatively high acceleration voltage, allowing for better re-detection of the detected defects on the mask. Thus, the charged particle beam images can be used to determine information for the detected defects more accurately (at greater resolution) than the inspection images of the detected defects, thereby enabling additional information to be determined from the charged particle beam images that can be used to more effectively make decisions regarding (handling) the detected defects. Recent advances in multi-beam and multi-line electron beam imaging also allow for large amounts of SEM data collection. In this way, after inspection is complete, a test SEM image (or ion beam image can be taken using one of the ion beam tools described herein) can be taken at each detected defect location (or a selected subset of the detected defect locations) determined from the previous step of output from inspection on a mask re-detection, CD-SEM or electron beam inspection tool. Each test charged particle beam image can then be processed in accordance with one or more of the following aspects described further herein— defect isolation, classification and printability.

[0067] In one embodiment, one or more computer subsystems are configured for detecting defects using a hot threshold, which can be performed as described further herein. In this way, the embodiments described herein can provide substantially high sensitivity EUV photomask defect inspection using a hot inspection run, followed by subsequent classification and wafer printability simulation of substantially larger number of detected defects using secondary charged particle beam images. Since the charged particle beam images provide higher resolution of the detected defects than inspection, the charged particle beam images provide more accurate information (e.g., more accurate location, more accurate size and other characteristics such as texture, shape, etc.) for the detected defects than the inspection images or signals.

[0068] In another embodiment, processing the detected defects includes determining whether the detected defects are real defects or false defects. For example, each charged particle beam image can be re-inspected to evaluate whether the detected defects are real. Determining whether the detected defects are real can include determining one or more characteristics of the detected defects and comparing the one or more characteristics to predetermined criteria that separate real defects from false defects. For example, a charged particle beam image of a detected defect can be processed by the computer subsystem to determine a size of the detected defect. The determined size can then be compared by the computer subsystem to a threshold value that separates real defects from false defects based on size. Other characteristics of the detected defects determined from the charged particle beam image can be used in a similar manner to separate real defects from false defects.

[0069] Because the computer subsystem can be configured for detecting defects on the photomask by applying a hot threshold to the output of the inspection subsystem, most of the detected defects can include "false" or "nuisance" defects. As used herein, the term "false defect" is generally defined as a defect that is detected on the photomask itself but is not actually a real defect on the photomask. Rather, a "false defect" can be detected due to noise sources on the photomask (e.g., line edge roughness (LER), relatively small critical dimension (CD) variations in the patterned features, thickness variations, etc.) and / or due to edge effects in the inspection subsystem itself or its configuration for inspection.

[0070] Accordingly, in general, the goal of mask inspection is not to detect false defects on the photomask. However, in the embodiments described herein, the computer subsystem can run a hot scan (i.e., a scan in which output is generated for the photomask as energy is scanned across the photomask and a hot threshold is applied to the output to detect defects, thereby making the scan a "hot scan") to ensure that all real defects (even real defects with relatively low SNR) are captured by the inspection. In other words, because it is a hot scan, a significant number of false defects will be detected and at least some real defects or defects of interest (DOIs) will also be detected (because the defects will also be detected by the hot scan). Regardless of the number of real defects and false defects detected by the inspection, charged particle beam images can be used to separate real defects from false defects. Unlike false defects, the term "real" defect as used herein can be generally defined as a defect that is detected by the inspection and confirmed by charged particle beam image processing to be an actual defect and / or a DOI. Accordingly, the conditions for ultimately being identified as a real defect or a false defect can be controlled by quality specifications set by the photomask user.

[0071] In some embodiments, processing the detected defects includes determining whether the detected defects are real defects or false defects (which can be performed as described further herein), and determining the printability of real defects on the wafer if the wafer is printed with the photomask in a lithography process. In other words, determining the printability simulates a wafer printing (lithography) process that would be performed using the photomask, thereby predicting how real defects would affect the pattern printed on the wafer. For example, each charged particle beam image generated for a detected defect can be re-detected to assess whether the detected defect is real, and if real, whether the detected defect is critical and whether it causes a relatively large error when printed on the wafer or nuisance and whether it causes a nearly unobservable error when printed on the wafer. Instead of performing this manually and using human judgment, the embodiments described herein provide an automated and substantially accurate way of modeling the EUV lithography process, thereby enabling a substantially accurate prediction of the printability of mask absorber defects starting with the detected defect charged particle beam image.

[0072] In one such embodiment, determining printability includes inputting the charged particle beam image generated for the real defect into a model of the lithography process, thereby generating a simulated wafer image that illustrates how the real defect affects one or more patterns printed on a wafer in the lithography process. Determining printability predicts the defect print impact through a wafer exposure condition simulation. This simulation can include a mask near-field simulation, which can use a Kirchoff approximation, a rigorous finite-difference time-domain (FDTD) solver, a rigorous coupled wave analysis (RCWA), or a compact approximation model such as the Defect Printability Simulator (DPS) Mask Model (examples of which are described in Clifford et al., "Compensation methods using a new model for buried defects in extreme ultraviolet lithography masks," SPIE Proceedings Vol. 7823, 78230V, p. 10, published September 24, 2010, incorporated by reference as if fully set forth herein), optics, and resist exposure and development models, which can include any suitable such models known in the art. If the wafer process is not yet determined or the mask near-field model is not well determined, printability can be checked directly on the mask binary layout without wafer exposure condition simulation for a rough estimate. The embodiments described herein can be configured as described in the referenced paper above. Determining printability can also be performed as described in U.S. Patent No. 9,002,497, issued April 7, 2015 to Marella et al., and U.S. Patent No. 9,478,019, issued October 25, 2016 to Sezginer et al., incorporated by reference as if fully set forth herein.

[0073] In another embodiment, processing the detected defect includes, if a wafer is printed with the photomask in a lithography process, determining printability of the detected defect on the wafer, and determining printability includes inputting design information of the photomask into a model of the charged particle beam subsystem, thereby generating a simulated charged particle beam reference image of a defect-free version of the photomask, simulating aerial images for the charged particle beam image generated for the detected defect and for the simulated charged particle beam reference image, and determining how the detected defect affects one or more patterns printed on a wafer in the lithography process based on the simulated aerial images.

[0074] In the above embodiment, a defect-free mask can be reproduced by clipping from a post-OPC design database corresponding to the location of the detected defect using a die-to-database method. Next, the charged particle beam-based processing workflow involves... Figure 6 The three main stages shown in the document.

[0075] In the first stage, the raw image from the charged particle beam tool (an image of the photomask generated during charged particle beam imaging) can first be denoised and refined to improve image quality. Figure 6 In the example shown, the original charged particle beam image 600 can be denoised and refined to produce a denoised test image 602. The denoising and refinement of the original charged particle beam image can be performed in any suitable manner known in this art. Next, the post-OPC design database is clipped (i.e., extracted) at the same location as the detected defect, and a charged particle beam model is applied to reproduce the defect-free reference charged particle beam image. The charged particle beam model can include any suitable charged particle beam model known in this art. Figure 6 In the example shown, a lookup function (which can be performed using a test charged particle beam image) is used in the design data to generate a design data clip 604 (e.g., a portion of GDS data), and a charged particle beam model is applied to the design data clip to generate a defect-free reference charged particle beam image 606. The reproduction can be calibrated using an actual defect-free charged particle beam image and its corresponding design data clip. The reproduced defect-free charged particle beam image can be a grayscale image. The denoised test image is aligned with and then subtracted from the reproduced reference charged particle beam image to generate a grayscale difference image, which is then used to isolate defect locations through local grayscale level variations. Alignment and subtraction can be performed in any suitable manner known in this art. Figure 6 As shown, a differential grayscale image 608 can be generated by subtracting the reference image 606 from the denoised test image 602. The differential image 608 can be used to perform defect isolation to generate a defect location image 610, as further described herein.

[0076] In the second stage, binary contours can be extracted from the denoised charged particle beam test image to generate a test binary image containing the detected defects, and binary contours can also be extracted from the reconstructed charged particle beam image to generate a reference binary image. Contour extraction can be applied to both the test image and the reconstructed charged particle beam image in any suitable manner known in this art. Figure 6As shown in the middle, binary contours can be extracted from the denoised test image 602, whereby a test binary image 612 is produced, and binary contours can be extracted from the defect-free reference charged-particle beam image 606, whereby a reference binary image 614 is produced. A binary difference image can then be produced by subtracting the reference binary image from the test binary image. For example, as shown in the bottom left, Figure 6 As shown in the middle, the test binary image 612 can be subtracted from the reference binary image 614, whereby a binary difference image 616 is produced.

[0077] The resulting gray-scale and binary difference images can then be used to calculate defect metrics from the defects isolated in the first stage, and to effectively determine the type of defects, e.g., line edge roughness (LER), pinhole (PinHole), pin dot (PinDot), intrusion, extrusion, etc. For example, as shown in the bottom right, Figure 6 As shown in the middle, the difference image 608 and the binary difference image 616 can be used to calculate defect metrics 618. The defect metrics can be organized in a data structure such as a defect information table and can include information such as: defect area in the binary difference image, size in x and y directions in the binary difference image, defect area in the gray-scale difference image, and size in x and y directions in the gray-scale difference image. The defect metrics can be used with or input to guidelines 620 that define different defect types to produce a defect classification 622. Otherwise, the defect metrics can be determined from the gray-scale difference image and the binary difference image in any suitable manner known in the art. Defect classification can also be performed as further described herein.

[0078] In a third stage, a suitable mask model can be applied to both the test and reference binary images with associated scanner exposure conditions to produce test and EUV aerial images. The mask model can be applied to the images as further described herein. In this way, the computer subsystem can run printability simulations under scanner optical conditions. For example, as shown in the bottom right, Figure 1 As shown in the middle, the test binary image 612 can be input to an EUV lithography simulation to produce a test aerial image 624, and the reference binary image 614 can be input to an EUV lithography simulation to produce a reference aerial image 626. An aerial image analyzer (AIA) can then be run to calculate the printability of all features within the field of view (FOV) of the charged-particle beam subsystem in which the defects are located. For example, as shown in the bottom right, Figure 4As shown in the middle, test aerial image 624 and reference aerial image 626 can be input to an AIA that can produce simulated image 628, showing how a defect would affect a feature of a mask in an aerial image projected on a wafer during a lithography process. An example of an AIA that can be used in the embodiments described herein is described in Verduijn et al., "Printability and actinic AIMS review of programmed mask blank defects," Proc. of SPIE Vol. 10143 Extreme Ultraviolet (EUV) Lithography VIII, 101430K, 13, published March 24, 2017, which is incorporated by reference as if fully set forth herein. The embodiments described herein can be further configured as described in the present disclosure.

[0079] In some embodiments, the charged particle beam subsystem is configured for automatically generating charged particle beam images at all locations of detected defects, processing the detected defects includes determining whether the detected defects are real defects or false defects, and the one or more computer subsystems are configured for automatically processing all detected defects that are determined to be real defects. For example, the charged particle beam subsystem can be configured for automatically generating a charged particle beam image for each of the detected defects that are reported by the inspection. In particular, a recipe (i.e., a set of instructions) for generating a charged particle beam image can instruct the charged particle beam subsystem to generate a charged particle beam image at each of the locations of the detected defects determined by the computer subsystem, thereby enabling automatic generation of charged particle beam images for all detected defects. In addition, the computer subsystem can be configured for automatically processing each of the charged particle beam images generated for the detected defects, thereby automatically processing the detected defects, which can include determining whether the detected defects are real defects or false defects, possibly in combination with any other processing described herein. Once the computer subsystem determines which of the detected defects are real or false, any further processing can be performed only for the real defects, thereby making the processing more efficient.

[0080] In another embodiment, processing the detected defect includes determining an isolation of the detected defect with respect to patterned features in a FOV of the charged-particle beam subsystem, the FOV centered on (or in which is located) the location of the detected defect. The defect isolation identifies a detected defect location within the FOV of the charged-particle beam image. For example, a local gray scale variation and a differential image generated from the charged-particle beam image taken at the defect location can be used to determine the detected defect location. In one such example, the local gray scale variation can be determined as a function of position within the differential image, and a maximum of the local gray scale variation can be determined as the defect location. However, the determination of the defect location within the FOV of the charged-particle beam image can be performed in any other suitable manner known in the art.

[0081] Determining defect isolation can also include determining which of the patterned features of the photomask that the detected defect is closest to in the FOV of the charged particle beam image and how close the detected defect is to those patterned features. For example, once the location of the detected defect within the FOV of the charged particle beam image has been determined as described above, the defect location information can be used to identify spatial information of the detected defect using the test image of the detected defect relative to the patterned features on the photomask. Determining defect isolation can include identifying which of the patterned features that the detected defect overlaps, whether the defect is located within one or more patterned features or at least partially overlaps one or more patterned features, or which of the patterned features is defective. Defect isolation can also include determining the location of the detected defect relative to the patterned features in the FOV, regardless of whether the detected defect overlaps the patterned features themselves or is within the patterned features. For example, if the detected defect spatially coincides with a patterned feature, determining defect isolation can include determining the location of the detected defect relative to the perimeter or outer boundary of the patterned feature. If the detected defect does not spatially coincide with a patterned feature in the FOV of the charged particle beam subsystem, determining defect isolation can include identifying the closest patterned feature to the defect and then determining how close the defect is to the patterned feature. Defect isolation can also include determining how close the detected defect is to a particular portion of the patterned feature (e.g., a corner, side, end, etc.) in some length dimension. The spatial distance between the detected defect and the closest patterned feature can be expressed as a single value, a range of values, an average value, a function, or in other suitable ways. Furthermore, a user can be particularly interested in how close the detected defect is to a subset of the patterned features of the photomask as compared to other patterned features of the photomask. For example, one or more parameters of the defect isolation step can be set such that if a type of patterned feature is within the FOV of the charged particle beam subsystem, the detected defect location relative to the closest patterned feature to the detected defect is determined and / or the detected defect location relative to any type of patterned feature is determined. The relative location of the detected defect to a patterned feature that can not be the closest to the defect can be determined in other ways as described herein.

[0082] In one such embodiment, the computer subsystem is configured for sending the determined isolation of the detected defect to a photomask repair tool, and the photomask repair tool uses the determined isolation of the detected defect in a repair process performed on the photomask. In this way, when a fix is needed for the mask, the defect isolation information can be used in the repair tool. In one such example, as Figure 5As shown in the middle, computer subsystem 118 (or any other computer subsystem of the system) can be configured to send the determined isolation of the detected defect to a photomask repair tool 140, which can or can not be part of the system. The photomask repair tool can have any suitable configuration known in the art. Further, the photomask repair tool can be a commercially available photomask repair tool, such as the MeRiT neXT system commercially available from Carl Zeiss SBE, LLC of Thornwood, New York. The photomask repair tool can use the determined isolation to determine one or more parameters of a repair process (e.g., repair location and area) targeted to repair the detected defect without altering any properly formed features proximate to the detected defect.

[0083] In another embodiment, processing the detected defect includes classifying the detected defect based on charged particle beam images generated for the detected defect by identifying a type of the detected defect. The classification defines a type of the defect. Figure 7 One example of a defect-free mask pattern 400 is shown, where colored regions indicate absorbers and non-colored regions indicate ML. Figure 7 Some examples of different types of defects are shown. If there is no significant defect and only local line edge roughness (LER) contributes to the detected inspection signal or image, the detected defect is referred to as LER, as shown by defect example 500. If a defect, such as a protrusion or an intrusion, is connected to an absorber, the detected defect is referred to as a hard defect, as shown by defect example 502. If an absorber has a pinhole and the bottom ML is exposed through the pinhole, the detected defect is a pinhole defect, as shown by defect example 504. If a part of a particle or the entire particle is on the ML, the particle or particle part has a printable impact on wafer exposure and is classified as a particle on ML, as shown by defect example 506. If a particle is completely on an absorber, the particle has no printable impact on wafer exposure and goes into a different bin, the particle is a particle on absorber, as shown by defect example 508.

[0084] Classification of detected defects based on charged particle beam images generated for the detected defects can be performed in any suitable manner. For example, a computer subsystem can determine one or more characteristics of a detected defect based on the charged particle beam images. Those characteristics can include, for example, size, shape, orientation, position, position relative to any nearby patterned features, texture, etc. Any determined characteristics and possibly the charged particle beam images can be input to a defect classifier by the computer subsystem. The defect classifier can be configured for determining a classification (or type, ranking, etc.) of a detected defect based on the determined characteristics of the detected defect and / or the charged particle beam images. The defect classifier can be any suitable defect classification method or algorithm known in the art. One example of such a defect classifier is a relatively simple decision tree in which different types of defects are separated by applying different cut lines to defect characteristics that separate different types of defects from one another. Other examples of suitable defect classifiers are machine learning type defect classifiers, some examples of which are described in U.S. Patent Application Publication No. 2018 / 0107928 by Zhang et al., published April 19, 2018, and U.S. Patent Application Publication No. 2019 / 0073568 by He et al., published March 7, 2019, which publications are incorporated by reference as if fully set forth herein. The embodiments described herein can be further configured as described in those publications.

[0085] In another embodiment, the system includes an atomic force microscope (AFM) subsystem configured for scanning the photomask at the location of the detected defect, thereby generating height information of the location of the detected defect. For example, in addition to absorber-related defects, the embodiments described herein can also be used for buried defects in the ML, where the defects are primarily located on the bottom of the ML stack but the morphology can be transferred from the bottom to the surface, e.g., as shown in Figure 1 Figure 3 ​An example of a typical ML embedded defect on an EUV photomask is shown. Figure 700 shows a protrusion type defect, and figure 702 shows a recess type defect. As shown in figures 700 and 702, the EUV photomask includes an ML stack 704 and a patterned absorber stack 706. In some examples, an ML protrusion defect 708 on the bottom of the ML stack can be transferred to the topmost surface of the ML stack as an ML protrusion defect on the surface 710. Further, an ML recess defect 712 on the bottom of the ML stack can be transferred to the topmost surface of the ML stack as an ML recess defect on the surface 714. The nature of the defect comes from the bottom of the ML, and the defect can transfer the morphology up to the topmost surface of the ML. Since such defects can cause phase differences in the light projected onto a wafer, the defects can induce printability defects on the wafer. An AFM tool can thus be used to scan the surface of the EUV photomask to obtain height information as an image that can be used to address the defects (i.e., in terms of height information across x and y positions across the EUV photomask).

[0086] As shown in Figure 1 The system can include an AFM subsystem 136 coupled to the computer subsystem 138, as shown in FIG. 1. The computer subsystem 138 can be coupled to other computer subsystems of the system, as further described herein, such that information, data, etc. can be transmitted between the computer subsystems. The AFM subsystem can have any suitable configuration known in the art. For example, suitable AFM tools that can be used as the AFM subsystem described herein are commercially available from Bruker Corp. of Billerica, MA (e.g., InSight series products), and in photomask repair tools such as the Rapid Probe Microscope (RPM) in the MeRiTneXT system commercially available from Zeiss. The AFM subsystem can also not be part of the system, but can be coupled to the system in some manner (e.g., through a computer subsystem of the system).

[0087] In some such embodiments, the computer subsystem can be configured to send results of defect detection to the AFM subsystem (e.g., as an inspection results file), and the AFM subsystem can automatically generate height information for all detected defects. In another embodiment, the computer subsystem can identify one or more of the detected defects as described further herein and select the identified one or more detected defects for scanning by the AFM subsystem. If less than all of the detected defects are selected for AFM scanning, only the defect detection results for the selected defects can be sent to the AFM subsystem or a computer subsystem coupled to the AFM subsystem. Regardless of how many detected defects are selected or sent for AFM scanning, the AFM subsystem can be configured for automatic scanning of the detected defects, and the scanning can be performed automatically as described further herein. The parameters for AFM scanning of the detected defects can all be the same (some pre-determined best known AFM parameters) or can be dynamically selected prior to or during the AFM scanning based on any information generated for the detected defects.

[0088] In one such embodiment, the computer subsystem is configured for identifying one or more of the detected defects that do not appear in a charged particle beam image generated at the location of the one or more of the detected defects, and the AFM subsystem is configured for automatically scanning the photomask only at the location of the identified one or more of the detected defects. For example, if there is a significant defect signal detected by photomask inspection, but no defect is observed in the charged particle beam image (this can be the case especially when the defect is detected on a non-absorber covered portion of the photomask and thus is not an absorber related defect), the AFM subsystem can be used to scan the surface at the location of the non-re-detected defect to acquire height information as can be used for processing the image of the detected defect. In particular, if a defect location reported by inspection is scanned by a charged particle beam subsystem as described herein and no defect is found in the charged particle beam image generated for the defect location (this can be determined by performing any suitable defect location method on the charged particle beam image), the defect location can be selected for AFM subsystem scanning. Defect re-detection performed using charged particle beam images can be performed as described herein with respect to inspection or in any other suitable manner known in the art. Since the AFM subsystem can generate height information for the detected defect, the information generated by the AFM subsystem can be more suitable for defect processing than the available inspection image (or information determined therefrom) and the available charged particle beam image in which the defect was not re-detected.

[0089] In another embodiment, the computer subsystem is configured to acquire information about additional defects detected on the blank substrate before fabricating the photomask using the blank substrate, and the AFM subsystem is configured to scan the photomask at the location of the additional defects, thereby generating height information about the location of the additional defects. For example, before forming the photomask pattern, the blank substrate (e.g., made of...) can be scanned. Figure 7 The blank substrate (which is a stack of MLs, not a substrate composed of patterned absorbers) is inspected. If the blank substrate passes inspection (e.g., is determined to be sufficiently defect-free), a patterned absorber layer can be formed on the blank substrate to form a photomask. The photomask can then be inspected as described herein. If the photomask blank inspection notices the location of ML defects before absorber patterning, an AFM tool can be used to scan the surface to obtain height information, which is then used as an image for processing the detected defects (performed as described herein).

[0090] The computer subsystem can be implemented in any suitable manner (e.g., from an inspection system that performs blank substrate testing (not shown) or from a storage medium (e.g., Figure 7 The storage medium 120 shown herein stores information on additional defects detected by the blank substrate inspection. This additional defect information may include any and / or all information from the additional defect report of the blank substrate inspection, including information about blank substrate defects. This information may be used by the computer subsystem described herein to determine the location on the photomask where an AFM scan will be performed to generate those blank substrate defects. The AFM subsystem may scan the photomask at the location of the additional defects as further described herein.

[0091] In some such embodiments, the computer subsystem is configured to process the detected defect based on height information generated for the detected defect. Processing of the detected defect based on height information can be performed in any suitable manner. For example, the height information can be input into a defect classifier as further described herein, and the defect classifier can determine the defect type based on the height information. In one such example, if the height information indicates that the detected defect protrudes above the upper surface of the ML stack of the photomask, then the defect classifier can determine that the defect is an ML protrusion defect, for example... Figure 8 The defects shown in the image. Conversely, if the height information indicates that the detected defect intrudes into the surface of the ML stack of the photomask, then defect processing can determine that the defect is an ML depression defect, for example. Figure 8 The defects shown in the figure. Processing the detected defects based on height information may also, or alternatively, include any of the other processes described herein, such as those performed using charged particle beam imaging.

[0092] In one such embodiment, processing the detected defects based on height information includes determining printability of the detected defects on a wafer if the wafer is printed in a lithography process with a photomask, and determining the printability includes inputting design information for the photomask into a model of the AFM subsystem, thereby generating simulated reference height information for a defect-free version of the photomask; simulating an aerial image for the height information generated for the detected defects and for the simulated reference height information; and determining how the detected defects affect one or more patterns printed on the wafer in the lithography process based on the simulated aerial image.

[0093] In this embodiment, a defect-free mask can be reproduced from a corresponding post-OPC design database clip in a die-to-database type approach. Then, the AFM-based workflow involves three main stages, as shown in Figure 8 In the first stage, raw images from the AFM tool can first be denoised and refined to improve the quality of the images. For example, as shown in Figure 8 In the first stage, raw images from the AFM tool can first be denoised and refined to improve the quality of the images. For example, as shown in Figure 8 In the first stage, raw images from the AFM tool can first be denoised and refined to improve the quality of the images. For example, as shown in Figure 8 In the first stage, raw images from the AFM tool can first be denoised and refined to improve the quality of the images. For example, as shown in

[0094] In the second stage, binary contours and ML morphology can be extracted from the denoised AFM test image to produce both a test binary image and a test ML morphology containing detected defects, and also binary contours and ML morphology can be extracted from the reproduced AFM image to produce a reference binary mask and a flat ML morphology. For example, as shown in Figure 8 In the second stage, binary contours and ML morphology can be extracted from the denoised AFM test image to produce both a test binary image and a test ML morphology containing detected defects, and also binary contours and ML morphology can be extracted from the reproduced AFM image to produce a reference binary mask and a flat ML morphology. For example, as shown in Figure 8 In the second stage, binary contours and ML morphology can be extracted from the denoised AFM test image to produce both a test binary image and a test ML morphology containing detected defects, and also binary contours and ML morphology can be extracted from the reproduced AFM image to produce a reference binary mask and a flat ML morphology. For example, as shown in Figure 9 In the second stage, binary contours and ML morphology can be extracted from the denoised AFM test image to produce both a test binary image and a test ML morphology containing detected defects, and also binary contours and ML morphology can be extracted from the reproduced AFM image to produce a reference binary mask and a flat ML morphology. For example, as shown in

[0095] In the third stage, the DPS mask model is applied to both the test and reference AFM images with the associated scanner exposure conditions to produce test and reference EUV aerial images. This step can be performed as further described herein. Additionally, this simulation can include a rigorous FDTD solver and RCWA. For example, as shown in Figure 9 In the third stage, the DPS mask model is applied to both the test and reference AFM images with the associated scanner exposure conditions to produce test and reference EUV aerial images. This step can be performed as further described herein. Additionally, this simulation can include a rigorous FDTD solver and RCWA. For example, as shown in

[0096] In another such embodiment, the computer subsystem is configured for automatically processing detected defects based on height information generated for the detected defects. For example, any or all of the processing described herein can be automatically performed by the computer subsystem. In other words, once the AFM subsystem has generated height information for a detected defect, the computer subsystem can use the height information and any other information for the detected defect that is generated by the system or otherwise available to the system to automatically perform defect processing. As further described herein, implementing automatic processing, which is provided by the embodiments described herein, is advantageous.

[0097] In another embodiment, one or more computer subsystems are configured for processing detected defects based on charged particle beam images generated for the detected defects in conjunction with images generated from energy directed to a photomask by an inspection subsystem. The images can be generated as further described herein. In one such embodiment, the energy directed to the photomask by the inspection subsystem includes light having a wavelength of 193 nm. For example, inspection of a photomask can be performed with 193 nm light, thereby making 193 nm images of the photomask readily available for defect processing. In particular, a computer subsystem can store output of the inspection subsystem for any defects detected on the photomask and then can use the stored output to process the defects in conjunction with charged particle images. Using such images in conjunction with charged particle images can be particularly advantageous for relatively shallow multi-layer defects that do not have any signal in the charged particle beam images. For example, as further described herein, if a defect is below the uppermost surface of the photomask, the defect can not generate a signal in the charged particle beam images. If high resolution AFM is not fast or stable enough for some applications, re-inspecting the optical 193 nm images along with the charged particle beam images to thereby process the defects can also be particularly advantageous. The computer subsystem can also use the charged particle beam images in conjunction with any other inspection images that can be used for defects detected on the photomask (i.e., not just 193 nm images) and / or other optical images. Defect processing performed using charged particle beams in conjunction with other images can be performed in other ways as further described herein.

[0098] The computer subsystems can also be configured for defect processing using other combinations of images and information generated or acquired by the systems described herein. For example, height information for a defect determined using an AFM subsystem can be used in conjunction with an optical image of the defect. In one such embodiment, one or more computer subsystems are configured for processing a detected defect based on height information generated for the detected defect in conjunction with an image generated from energy detected by the inspection subsystem from the photomask. The image can include any of the images described herein. In one such example, the image can be a 193 nm optical image. For example, in one embodiment, the energy directed to the photomask by the inspection subsystem includes light having a wavelength of 193 nm. However, the optical image used with the height information for defect processing can also include any of the other optical images described above. Defect processing performed using optical images and height information can be performed in other ways as further described herein.

[0099] Each of the embodiments of each of the systems described above can be combined together into a single embodiment.

[0100] The embodiments described herein have several advantages over other methods and systems for EUV photomask inspection. For example, one advantage of the embodiments described herein is their ability to isolate accurate defect locations by comparing secondary images (e.g., charged particle beam images and / or AFM images) to a design database. Another advantage is that the embodiments enable defect type classification by substantially high resolution charged particle beam / AFM images. An additional advantage is that printability can be predicted without relying on any operator's experience or actinic imaging tools. Furthermore, since the entire workflow can be fully automated by software and processed in parallel by computing servers, the throughput of the entire process can be much faster than manual processing.

[0101] To the inventors' knowledge, the embodiments described herein were generated by first attempting to implement a substantially high sensitivity EUV mask inspection followed by the fusion of the processing power of fully automated charged particle beam / AFM image analysis, which advantageously extends EUV photomask inspection capabilities and improves defect processing accuracy and throughput of patterned EUV mask inspection. Unlike the embodiments described herein, actinic imaging tools (such as the AIMS tool commercially available from Zeiss) can be used for EUV photomask defect processing, but are too slow to handle hundreds of defects in a production line. Another alternative to the embodiments described herein can be to utilize actinic inspection tools with low numerical aperture (NA) inspection (LNI) for defect processing, but such inspection tools are not currently available.

[0102] The embodiments described herein also differ from wafer inspection methods and systems in several aspects. For example, for photomasks, generally, it is desirable to perform inspection at the actinic wavelength of the photomask for several reasons including that the printability of detected defects can be assessed with a properly configured inspection tool, and the use of the actinic wavelength ensures that defects can be detected on the photomask with sufficient sensitivity. In contrast, wafer inspection methods and systems are generally designed primarily based on the minimum size of defects that need to be detected on the wafer. However, there is no need to perform this wafer inspection to determine the printability of any defects on the wafer because the defects have already been printed on the wafer and the wafer is not used to print any other substrates. In contrast, even if a photomask defect is imaged by a high resolution tool, there can still be a question of whether the defect would print on a wafer. To address this, an actinic re-detection tool takes images through scanner optics, which are low resolution, to mimic wafer effects. The embodiments described herein replace the actinic re-detection process with simulation, which is not necessary for wafer inspection and wafer defect analysis. In this way, the actinic wavelength of the photomask and the wafer printed with the photomask can be the same (i.e., the wavelength of light used to transfer a pattern from the photomask to the wafer thereby causing a photochemical reaction in one or more materials on the wafer (e.g., photoresist)), but the wafer inspection process is not designed based on that actinic wavelength or the lack of a wafer inspection tool capable of inspecting at that actinic wavelength.

[0103] Thus, unlike wafer inspection, the embodiments described herein have been designed to overcome the lack of a suitable inspection method or system for EUV photomasks. Features of the embodiments described herein that enable inspection of EUV photomasks include the ability to efficiently process the large number of defects detected by inspecting EUV photomasks using charged particle beam images (and optionally AFM images) for defect processing. This feature enables the use of a hot threshold in the inspection, resulting in the detection of a larger number of both real and false defects without significantly impacting the inspection process. Thus, the inspection can detect defects of the minimum size that need to be detected for EUV photomask inspection, and the resulting detected false defects can be separated from real defects by the processing described herein. Furthermore, because defect processing (and any image generation performed to facilitate defect processing) can be automated, the embodiments described herein will be efficient, easy to use, and accurate, especially compared to other available options for EUV photomask inspection.

[0104] Another embodiment relates to a method for inspecting a photomask. The method includes directing energy to and detecting energy from the photomask. The photomask is configured as described herein. The method further includes detecting defects on the photomask based on the detected energy and generating a charged particle beam image of the photomask at the location of the detected defect. The method further includes processing the detected defect based on the charged particle beam image generated for the detected defect.

[0105] Each of the steps of the method may be performed as further described herein. The method may also include any other steps that can be performed by the verification subsystem, the charged particle beam subsystem, the AFM subsystem, and / or the computer subsystem or system described herein (which may be configured according to any of the embodiments described herein). Furthermore, the methods described above may be performed by any of the system embodiments described herein.

[0106] Additional embodiments relate to a non-transitory computer-readable medium storing program instructions executable on a computer system to perform a computer-implemented method for detecting a photomask. ​ An example of this embodiment is shown below. Specifically, as... ​ As shown herein, the non-transitory computer-readable medium 900 contains program instructions 902 that can be executed on a computer system 904. A computer-implemented method may include any step of any method described herein.

[0107] Program instructions 902 for implementing methods such as those described herein may be stored on a computer-readable medium 900. The computer-readable medium may be, for example, a storage medium such as a disk or optical disc, magnetic tape, or any other suitable non-transitory computer-readable medium known in the art.

[0108] The program instructions can be implemented using any of the following methods: programmatic, component-based, and / or object-oriented, as well as other technologies. For example, ActiveX controls, C++ objects, JavaBeans, Microsoft Foundation Classes (“MFC”), SSE (Streaming SIMD Extensions), or other technologies or methods may be used as needed to implement the program instructions.

[0109] Computer system 904 may be configured according to any of the embodiments described herein.

[0110] Further modifications and alternative embodiments of various aspects of the application will be apparent to those skilled in the art in view of this description. For example, methods and systems are provided for processing defects detected on photomasks. Accordingly, this description is to be construed as illustrative only and is for the purpose of teaching those skilled in the art the general manner of carrying out the application. It is to be understood that the forms of the application shown and described herein are to be taken as the presently preferred embodiments. Elements and materials can be substituted for those illustrated and described herein, parts and processes can be reversed, and certain features of the application can be utilized independently, combinations of which are to be considered within the scope of the application, as defined by the appended claims, where the spirit and scope of the application are to be considered as limited only by the pertinences of the claims.

Claims

1. A photomask inspection system, comprising: an inspection subsystem configured for directing energy to a photomask, detecting energy from the photomask, and generating an output in response to the energy from the photomask, wherein the photomask is configured for use at one or more extreme ultraviolet wavelengths; one or more computer subsystems configured to detect a defect on the photomask by applying a hot threshold to the output; and a charged particle beam subsystem configured for generating a charged particle beam image of the photomask at a location of a detected defect determined by the one or more computer subsystems, wherein the one or more computer subsystems are further configured for processing the detected defect based on the charged particle beam image generated for the detected defect, wherein the one or more computer subsystems are further configured to: denoise and refine the charged particle beam image to generate a denoised test image; extract binary contours from the denoised test image to generate a test binary image containing the detected defect, and extract binary contours from the charged particle beam image to generate a reference binary image; and apply a mask model to the test binary image and the reference binary image to generate test and reference extreme ultraviolet (EUV) aerial images.

2. The system of claim 1, wherein the energy directed to the photomask by the inspection subsystem comprises light having a wavelength of 193 nm.

3. The system of claim 1, wherein the energy directed to the photomask by the inspection subsystem comprises light having one or more wavelengths ranging from 193 nm to 257 nm.

4. The system of claim 1, wherein the energy directed to the photomask by the inspection subsystem comprises light having a wavelength of 13.5 nm.

5. The system of claim 1, wherein the energy directed to the photomask by the inspection subsystem comprises electrons.

6. The system of claim 1, wherein the energy directed to the photomask by the inspection subsystem comprises ions.

7. The system of claim 1, wherein processing the detected defect comprises determining whether the detected defect is a real defect or a false defect.

8. The system of claim 1, wherein processing the detected defect comprises determining whether the detected defect is a real defect or a false defect, and determining a printability of the real defect on a wafer if the photomask is utilized in a lithography process to print the wafer.

9. The system of claim 8, wherein determining the printability comprises inputting the charged particle beam image generated for the real defect into a model of the lithography process, thereby generating a simulated wafer image that illustrates how the real defect affects one or more patterns printed on the wafer in the lithography process.

10. The system of claim 1, wherein processing the detected defects comprises: If the wafer is printed with the photomask in a lithography process, a printability of the detected defect on the wafer is determined, and wherein determining the printability comprises inputting design information for the photomask into a model of the charged particle beam subsystem, thereby generating simulated charged particle beam reference images of a defect-free version of the photomask; simulating aerial images for the charged particle beam images generated for the detected defect and for the simulated charged particle beam reference images; and determining how the detected defect affects one or more patterns printed on the wafer in the lithography process based on the simulated aerial images.

11. The system of claim 1, wherein the charged particle beam subsystem is further configured for automatically generating the charged particle beam images at all of the locations of the detected defects, wherein processing the detected defects comprises determining whether the detected defects are real defects or false defects, and wherein the one or more computer subsystems are further configured for automatically processing all of the detected defects determined to be the real defects.

12. The system of claim 1, wherein processing the detected defects comprises determining an isolation of the detected defect with respect to patterned features in a field of view of the charged particle beam subsystem centered on the location of the detected defect.

13. The system of claim 12, wherein the one or more computer subsystems are further configured for sending the determined isolation of the detected defect to a photomask repair tool, and wherein the photomask repair tool uses the determined isolation of the detected defect in a repair process performed on the photomask.

14. The system of claim 1, further comprising an atomic force microscope subsystem configured for scanning the photomask at the locations of the detected defects, thereby generating height information for the locations of the detected defects.

15. The system of claim 14, wherein the one or more computer subsystems are further configured for identifying one or more of the detected defects that do not appear in the charged particle beam images generated at the locations of the one or more of the detected defects, and wherein the atomic force microscope subsystem is further configured for automatically scanning the photomask only at the locations of the identified one or more of the detected defects.

16. The system of claim 14, wherein the one or more computer subsystems are further configured for acquiring information for additional defects detected on a blank substrate before the photomask is made with the blank substrate, and wherein the atomic force microscope subsystem is further configured for scanning the photomask at locations of the additional defects, thereby generating height information for the locations of the additional defects.

17. The system of claim 14, wherein the one or more computer subsystems are further configured for processing the detected defects based on the height information generated for the detected defects.

18. The system of claim 17, wherein processing the detected defects based on the height information comprises: If a wafer is printed in a lithography process using the photomask, a printability of the detected defects on the wafer is determined, and wherein determining the printability includes inputting design information for the photomask into a model of the atomic force microscope subsystem, thereby generating simulated reference height information for a defect-free version of the photomask; simulating an aerial image for the height information generated for the detected defects and for the simulated reference height information; and determining how the detected defects affect one or more patterns printed on the wafer in the lithography process based on the simulated aerial image.

19. The system of claim 17, wherein the one or more computer subsystems are further configured for processing the detected defects based on the height information generated for the detected defects in conjunction with images generated from the energy detected by the inspection subsystem from the photomask.

20. The system of claim 19, wherein the energy directed by the inspection subsystem to the photomask includes light having a wavelength of 193 nm.

21. The system of claim 14, wherein the one or more computer subsystems are further configured for automatically processing the detected defects based on the height information generated for the detected defects.

22. The system of claim 1, wherein the one or more computer subsystems are further configured for processing the detected defects based on the charged particle beam images generated for the detected defects in conjunction with images generated from the energy detected by the inspection subsystem from the photomask.

23. The system of claim 22, wherein the energy directed by the inspection subsystem to the photomask includes light having a wavelength of 193 nm.

24. The system of claim 1, wherein the charged particle beam subsystem is further configured as an electron beam subsystem.

25. The system of claim 1, wherein the charged particle beam subsystem is further configured as an ion beam subsystem.

26. The system of claim 1, wherein processing the detected defects includes classifying the detected defects based on the charged particle beam images generated for the detected defects by identifying a type of the detected defects.

27. A non-transitory computer readable medium storing program instructions executable on a computer system for performing a computer-implemented method for inspecting a photomask, wherein the computer-implemented method comprises: directing energy to a photomask, detecting energy from the photomask, and generating an output in response to the detected energy, wherein the photomask is configured for use at one or more extreme ultraviolet wavelengths; detecting defects on the photomask by applying a thermal threshold to the output; producing a charged-particle beam image of the photomask at a location of the detected defect; and processing the detected defect based on the charged-particle beam image produced for the detected defect, wherein the computer-implemented method further comprises: denoising and refining the charged-particle beam image to produce a denoised test image; extracting binary contours from the denoised test image to produce a test binary image containing the detected defect and extracting binary contours from the charged-particle beam image to produce a reference binary image; and applying a mask model to the test binary image and the reference binary image to produce test and reference extreme ultraviolet, EUV, aerial images.

28. A method for inspecting a photomask, comprising: directing energy to a photomask, detecting energy from the photomask, and producing an output in response to the detected energy, wherein the photomask is configured for use at one or more extreme ultraviolet wavelengths; detecting a defect on the photomask by applying a thermal threshold to the output; producing a charged-particle beam image of the photomask at a location of the detected defect; and processing the detected defect based on the charged-particle beam image produced for the detected defect, wherein the method further comprises: denoising and refining the charged-particle beam image to produce a denoised test image; extracting binary contours from the denoised test image to produce a test binary image containing the detected defect and extracting binary contours from the charged-particle beam image to produce a reference binary image; and applying a mask model to the test binary image and the reference binary image to produce test and reference extreme ultraviolet, EUV, aerial images.

Images