Method and apparatus for repairing defects in a lithographic mask

Abstract

The present invention relates to a method (1900) for repairing at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800), comprising the step of identifying at least one repair shape parameter for the at least one defect—the step of identifying the parameter includes the step of assigning at least one numerical value to the parameter, wherein the numerical value deviates from a numerical value predetermined by the at least one defect for the parameter.

Description

Technology Field

[0001] This patent application claims priority to German patent application DE 10 2020 208 980.9, titled "Method and apparatus for repairing defects in a lithographic mask," filed with the German Patent Office on July 17, 2020. The entirety of German patent application DE 10 2020 208 980.9 is incorporated by reference into this patent application.

[0002] The present invention relates to a method and apparatus for repairing at least one defect in a lithographic mask. Additionally, the present invention relates to a method and apparatus for repairing at least one pattern element in a lithographic mask. Background Technology

[0003] As integration density increases in the semiconductor industry, lithographic masks must image increasingly smaller structures on wafers. One possibility to account for this trend is to use lithographic or photolithographic masks in which the chemical line wavelength is shifted to a shorter wavelength. Currently, ArF (argon fluoride) excimer lasers emitting at a wavelength of approximately 193 nm are frequently used as light sources in photolithography. The use of double or multi-exposure masks makes it possible to create photoresist structures with dimensional sizes that cannot be achieved with a single exposure step.

[0004] Currently, photolithography systems using electromagnetic radiation in the EUV (extreme ultraviolet) wavelength range (preferably in the range of 10 nm to 15 nm) are being developed. Since optically transparent materials are not currently available in the specified EUV range, these EUV photolithography systems are based on a completely new beam guiding concept using reflective optical elements. The technical challenges in developing EUV systems are immense, and enormous development efforts are required to bring the systems to a level ready for industrial application.

[0005] Lithographic masks, photolithographic masks, exposure masks, photomasks, or simple masks make a critical contribution to the imaging of much smaller structures in photoresists arranged on a wafer. As integration density increases, reducing the minimum structure size of exposure masks is becoming increasingly important. To ensure that structures of increasingly smaller masks can be reliably imaged in the photoresists applied to the wafer, resolution enhancement techniques (RETs), such as Optical Proximity Correction (OPC), are being used more frequently. The following exemplary literature describes various aspects of RET or OPC techniques: W.-M. Gan et al.: "Placement of sub-resolution assist features based on a generic algorithm", DOI 10.1109 / ACCESS.2019.2926102, IEEE ACCESS; P. Gupta et al.: "Manufacturing-aware design methodology for assist feature correctness", Design and Process Integration for Microelectronic Manufacturing III, Proc. of SPIE Vol. 4756, Bellingham, WA, 2005, doi: 10.1117 / 12.604872, US 2006 / 0 046 160 A1, US 8 739 080 B1, US 8 498 469 B2, US 2008 / 0 077 907 A1, US 2009 0 258 302 A1 and US 10 318 697 B2.

[0006] By combining the use of RET technology with increasingly smaller structural or pattern elements, the process for manufacturing photolithographic masks becomes increasingly complex, requiring more time and ultimately increasing costs.

[0007] Due to the small structural size of pattern elements, much smaller deviations from the design specifications during mask manufacturing appear as visual or "printable" errors on the wafer. These must be repaired whenever possible. Due to the small dimensional size of defects visualized on the wafer, reaching values ​​in the low double-digit nanometer range, the cost required to detect, for example, said defects is becoming very high. Furthermore, repairing increasingly smaller defects is becoming more difficult. First, determining the location of repair tools for identified defects is possible only through very complex measurements, and second, setting up repair tools for specific defects requires a significant amount of time.

[0008] Accordingly, the present invention solves the problem of specifying a method and apparatus for repairing lithographic masks, particularly for improving small defects.

[0009] According to an exemplary embodiment of the present invention, this problem is solved by the method according to claim 1 and the apparatus according to claim 22. In a further exemplary embodiment, this problem is solved by the method according to claim 18 and the apparatus according to claim 23.

[0010] In one embodiment, a method for repairing at least one defect in a lithographic mask comprises the step of identifying at least one repair shape parameter for at least one defect, wherein the step of identifying the parameter comprises the step of assigning at least one numeric value to the parameter, wherein the numeric value deviates from a numeric value predetermined by at least one defect for said parameter. In particular, this may be advantageous for small defects, i.e., defects having at least one dimensional size that is smaller than 10 times the resolution limit of the mask, smaller than 5 times the resolution limit, smaller than 3 times the resolution limit, or smaller than the resolution limit. It may be particularly advantageous when the small defect has at least one dimensional size within the range of 2% to 50% of the resolution limit.

[0011] The defect to be repaired typically defines the parameters of the repair shape used to repair the defect. Through comprehensive analysis, the inventors have discovered that if at least one of the parameters of the repair shape is assigned a numerical value different from what is actually required by the repair of the defect, the compensation or repair of very small defects, which until now could only be removed with extreme difficulty, can be significantly simplified. In particular, this may be advantageous for small defects, namely defects having at least one dimensional size within the range of 2% to 50% of the resolution limit of the mask.

[0012] A repair shape in which parameters have only numeric values ​​identified based on measurements is hereinafter also referred to as a nominal repair shape. Typically, a nominal repair shape is formed from the difference between a measured mask segment having defects and a defect-free mask segment of the measured equivalent. Alternatively and / or additionally, it is also possible to generate a nominal repair shape by subtracting the design data of the measured defective mask segment from the latter. The repair shape used in the method according to the present invention differs from the nominal repair shape in that, in the first repair shape mentioned, at least one parameter has a numeric value that deviates from the numeric value identified from the measurement result.

[0013] Simply put, it may be advantageous to perform repairs (e.g., edge errors) based not only on numerical values ​​for parameters predetermined by each defect (e.g., in the case of edge errors caused by the difference between the target edge position and the error edge position predetermined by the mask design). This is because very precise repairs are required to correct these (small) defects, especially when the numerical value of the parameter is in or below the resolution limit region (e.g., when there is only a very slight error in the edge position). Small errors during repair may result in repairs that do not entail significant improvement. However, if a targeted departure is made from the numerical value predetermined by each defect for the corresponding parameter, what can be achieved is that the error during repair has a significantly less pronounced effect, so that the requirements of the repair process can be relaxed where appropriate.

[0014] The method according to the present invention is described below based on an example of an edge placement error of a pattern element. The deviation (dx) of the edge of the pattern element from the specifications of the mask design is converted by the mask into an edge placement error (EPE), given by the product of the deviation (dx) and the possible mask enhancement factor (MEEF) (mask error enhancement factor) for defects or deviations and magnification or reduction of the projection lens placed downstream of the mask during the exposure process. The projection lens of the photolithographic exposure system may have a magnification of M=1 / 4 or M=1 / 5.

[0015] However, when the deviation (dx) of the edge of a pattern element becomes smaller than the resolution limit of the photolithographic mask, the latter converts the significantly reduced deviation (dx) during the exposure process into a reduced edge parameter error. This is caused by the chemical radiation of the photomask being averaged over structures whose dimensional size is smaller than the resolution capability of the mask. The details of this averaging process depend on the structures considered and the details of the exposure process used to image these structural elements. The present application utilizes this practical problem to facilitate the repair of defects whose dimensional size is smaller than the resolution limit of the lithographic mask.

[0016] In this regard, during edge repair of pattern elements, for example, the parameters of this repair (or the corresponding parameters of the nominal repair shape) actually require the deposition of material from a measured edge (numerically predetermined by the defect) to a target edge, and the material can be intentionally deposited away from the measured edge, where the material is selectively deposited over a shorter length. Thus, errors in the precise positioning of the deposited material have a significantly smaller impact on the quality of the repaired mask (for example, due to diffraction effects). This applies similarly to errors requiring the etching of the material, and in the case of edge errors, for example, the material can be etched away from the measured edge.

[0017] A repair geometry combines all instructions performed by a repair tool to remove defects in a photomask. As an example of a defect involving excess material, the repair geometry describes a local etching process that can be used to remove excess material from the mask. The repair geometry typically has a base area scanned by a particle beam in a manner defined in the repair geometry. This means that the repair geometry defines the particle energy of the particle beam, the spot width at the focus point of the particle beam, its residence time at any location, the distance between adjacent incident points of the particle beam to the mask or defect, and the time period until the particle beam reaches the starting point again. The numerical values ​​of the area scanned by the particle beam and the parameters described above may also change during the implementation of the repair geometry. In the case of a local etching process, the repair geometry also specifies the temporal progression of the supply of etching gas(s); that is, the repair geometry controls the gas volumetric flow rate of the etching gas(s) during the processing of at least one defect.

[0018] In this application, the term lithographic mask includes photolithographic mask.

[0019] At least one parameter may include at least one lateral dimension of at least one repair feature, at least one height dimension of at least one repair feature, the distance between at least one repair feature and at least one defect, the material composition of at least one defect, the geometric shape of at least one repair feature, and at least one element from a group of surroundings of at least one defect on a lithographic mask. The material composition of at least one defect primarily determines its complex refractive index.

[0020] The ease or improvement of repairing or compensating for minor defects can be achieved by various means. In one aspect, the defect does not need to be repaired 1:1 as predetermined by the parameters of the nominal repair shape. In this respect, the basic area of ​​the repair shape may be smaller than the basic area of ​​the defect. In another aspect, the height of the repair shape may be smaller than the height of the defect. Thus, the method described above enables defect correction in which material removal from the mask or material accumulation on the mask may be smaller than the volume of the defect. This situation has a favorable effect on the defect processing time.

[0021] In another embodiment, the averaging of the aforementioned chemical radiation over structures smaller than the resolution capability allows the repair shape to be positioned somewhat away from the location of the original defect. As a result, the very strict requirements for positional accuracy during the placement of the repair shape for the defect are certainly relaxed without having a significantly adverse effect on the quality of defect repair or defect compensation.

[0022] Furthermore, the geometric shape of the repair shape can deviate from the actual defect shape. This ensures that defect repair is significantly simplified because the shape of the repair shape can be selected at least partially independently of the defect shape and thus has a geometric shape that can be generated significantly more simply.

[0023] At least one parameter may deviate from a predetermined numerical value by a predetermined absolute value due to at least one defect in the said parameter.

[0024] The deviation of the numerical value of at least one lateral parameter can be selected from a range in which the lower limit is greater than 0 and the upper limit is smaller than the resolution limit (of the defect-free area) of the lithographic mask.

[0025] The resolution limit of a lithographic mask can be determined by the chemical line wavelength of the lithographic mask, the numerical aperture (NA), and the exposure setting suitable for projecting the pattern of the lithographic mask onto the photoresist arranged on the wafer.

[0026] The numerical aperture may include the NA of the projection lens of an exposure system suitable for projecting the pattern of a lithographic mask onto a photoresist. The NA of the projection lens is the mask-side numerical aperture (NA M It may include ).

[0027] Typically, lithographic masks are designed for operation in a specific exposure system. This is because the lithographic mask is designed based on the chemical beam wavelength and the numerical aperture (NA) of the projection lens of the exposure system. M This means that it is specifically designed for an exposure setting specifically used by an exposure system for projecting the pattern of the lithographic mask onto the photoresist. In this sense, the design of the lithographic mask determines or fixes its resolution limit.

[0028] The resolution limit of the lithographic mask may include the mask-side resolution limit of an exposure system suitable for exposing a wafer using the lithographic mask.

[0029] Mask-side resolution limit (R M )Is It can be determined by, where λ is the wavelength of the chemical line of the lithographic mask, and NA M is the numerical aperture on the mask side of the projection lens of the exposure system, and σ is the exposure setting of the exposure system that exposes the lithographic mask.

[0030] Mask-side resolution limit (R M ) is proportional to the chemical line wavelength (λ), and the mask numerical aperture (NA M It is inversely proportional to ). NA Mσ can have a numerical value within the range of approximately 0.1 to 0.5. Additionally, the resolution limit of the lithographic mask depends on the exposure setting (σ) of the exposure system that exposes the wafer using the lithographic mask. The numerical value of σ is in the range of 0 (for central illumination) to 1 (for maximum oblique illumination). Oblique exposure is also referred to as off-axis exposure in the technical field.

[0031] Narrowing (NA) M Depending on ) and exposure setting (σ), the resolution limit of a photomask exposed to a deep ultraviolet (DUV) wavelength of λ=193 nm is 150 nm ≤ R M It is between ≤300nm. In the case of an EUV mask with a chemical line wavelength of λ=13.5nm, the resolution limit is currently 50nm≤R M It is in the range of ≤100nm. This means that the range in which a single numerical value of one parameter of the repair shape or a numerical value of multiple parameters can be selected is reduced as the chemical line wavelength of the photolithographic mask decreases. In other words, the method defined above opens up a wider range of new degrees of freedom in the DUV wavelength range than in the EUV range.

[0032] The deviation of the numerical value of at least one parameter may include a range of 2% to 80%, preferably 2% to 50%, and most preferably 2% to 30% of the resolution limit of the lithographic mask.

[0033] The dimensional size of at least one dimension of at least one repair shape may include a range of 10% to 90%, preferably 20% to 80%, more preferably 30% to 70%, and most preferably 40% to 60% of the dimensional size at the corresponding dimension of at least one defect.

[0034] The dimension size of at least one dimension of at least one repair shape may include the lateral dimension and / or height of at least one repair shape.

[0035] The distance between at least one repair feature and at least one defect may include a range of 2% to 80%, preferably 2% to 50%, more preferably 2% to 30%, and most preferably 2% to 10% of the resolution limit of the defect-free area of ​​the lithographic mask.

[0036] The repair shape may have lateral displacement relative to the nominal repair shape, may have lateral deviation of its dimensional size relative to the nominal repair shape, and may have height deviation relative to the nominal repair shape, so that the repair shape and the nominal repair shape produce substantially the same optical intensity distribution in the photoresist. This means that the lateral dimension(s) and height of the repair shape can be combined or correlated. For example, a small height of the repair shape can be compensated by increasing one or both lateral dimensions, and vice versa. Additionally, a correlation is possible between the height or lateral dimension(s) of the repair shape and the lateral displacement of the repair shape associated with the defect. The details depend on the repair shape, the surrounding environment of the mask on which the repair shape is created, and the exposure process.

[0037] At least one defect may include at least one element from the group of edge placement errors of pattern elements, interrupted and / or bridging connections of pattern elements, outliers in edge roughness of pattern elements, particles attached to the lithographic mask, lateral defects having only low printability, defect residues of performed defect repairs, sidewall angle errors of pattern elements, and distance ranges between two pattern elements and / or center of gravity errors of pattern elements.

[0038] The step of verifying parameters of at least one repair shape may include the step of recording at least one aerial image of at least one defect. The aerial image may be measured using a mask inspection device. The mask inspection device may include an optical inspection device and / or an inspection device that scans the mask surface. The optical mask inspection device may include, for example, a laser interferometer, and the mask inspection device that scans the mask surface may include, for example, an atomic force microscope. The optical mask inspection device may be designed to record an aerial image of a photomask and / or an aerial image focus stack.

[0039] The step of recording at least one aerial image may include recording at least one aerial image of at least one defect at the chemical line wavelength of the lithographic mask, and / or recording a focus stack of at least one aerial image of the defect. Recording an aerial image at the chemical line wavelength allows details that are imaged during the later exposure of the photomask in the photoresist to appear in the aerial image. Therefore, it is advantageous to record an aerial image at the chemical line wavelength of the mask. It is much more advantageous to determine the imaging behavior of the defect region of the photolithographic mask while tuning through the focus of the mask.

[0040] The step of verifying parameters of at least one repair shape may include scanning at least one defect by a scanning particle microscope and / or a scanning probe microscope. Additionally, the step of verifying parameters of at least one repair shape may include recording at least one aerial image of at least one defect and scanning at least one defect by a scanning particle microscope and / or a scanning probe microscope.

[0041] A scanning particle microscope may include at least one element from the group consisting of a scanning electron microscope (SEM), a scanning ion microscope (FIB, focused ion beam), and a scanning electron microscope using polarization analysis (SEMPA).

[0042] A scanning probe microscope may include at least one element from the group consisting of a scanning tunneling microscope (STM), an atomic force microscope (AFM), a magnetic force microscope (MFM), a scanning near-field optical microscope (SNOM), and a scanning near-field acoustic microscope (SNAM).

[0043] The step of verifying at least one parameter of a repair shape may additionally be based on at least one element from the group of exposure settings in which the lithographic mask is exposed during operation, design data of the lithographic mask, refractive index data of the deposited material for repairing defects of the missing material, and RET technology for resolution enhancement for the lithographic mask.

[0044] The step of verifying at least one parameter of a repair shape may include the step of applying at least one algorithm to the measurement data and design data of at least one defect of a lithographic mask.

[0045] At least one algorithm may be implemented using hardware, software, firmware, or a combination thereof. Additionally, at least one algorithm may be stored in non-volatile memory. In particular, at least one algorithm may be stored in solid-state memory (SSD, solid-state drive).

[0046] The step of assigning a numerical value may include the step of applying a trained machine learning model that determines the deviation of the numerical value of at least one parameter from a numerical value predetermined by at least one defect for the parameter.

[0047] A machine learning model may include a transformation model having at least two transformation blocks, wherein each of the at least two transformation blocks includes at least one generally learnable function that transforms an input into an output used as an input to a subsequent transformation block. The machine learning model may include at least one element from the group of parametric mapping, artificial neural networks, deep neural networks, time-delayed neural networks, convolutional neural networks, recurrent neural networks, long short-term memory networks, generative models, kernel density estimators, statistical models, decision trees, linear models, and time-invariant models.

[0048] The machine learning model may include (a) at least one encoder block for determining an image of at least one defect and an information-carrying feature of design data assigned to the image of at least one defect; and (b) at least one decoder block for generating at least one effect of at least one defect from the determined information-carrying feature, wherein at least one decoder block shows what the superposition of at least one defect with the corresponding design data of the image of at least one defect looks like.

[0049] The step of verifying at least one parameter of a mathematical shape may include the step of applying a trained machine learning model to verify at least one parameter of a mathematical shape.

[0050] A trained machine learning model may be used in at least two embodiments of the method described above. First, a correspondingly trained machine learning model may be used to assign numerical values ​​or numerical values ​​deviating from numerical values ​​as a result of at least one defect to one or more parameters of a repair shape. However, it is also possible—which is a currently preferred embodiment—for a correspondingly trained machine learning model to identify all parameters of at least one repair shape for repairing at least one defect based on measurement data, e.g., one or more aerial images of the defect, design data of the mask, and optionally the structure of one or more RET techniques.

[0051] The method according to the present invention may further include the step of generating at least one repair element on a lithographic mask by an identified repair shape. Implementing the repair shape generates a repair element on a photomask designed to eliminate, i.e., repair or compensate for, at least one defect as much as possible.

[0052] At least one repair element may not be imaged during the exposure of the photolithographic mask. At least one repair element may change the imaging behavior of at least one defect during the exposure of the lithographic mask.

[0053] The generated repair elements typically have deviations from the nominal repair geometry, which is below the resolution limit of the photolithographic mask, and for this reason, they may not be imaged in the photoresist and thus on the underlying wafer during exposure of the mask with a chemical line wavelength. However, the repair elements generated on the mask are designed to alter the imaging behavior of the defective region of the photomask, so that the defective region combined with the repair elements results in imaging behavior very similar to that of a defect-free region with the same pattern arrangement. As a result, during the exposure process, the repaired mask produces substantially the same edge positions as the defect-free mask. The effect of at least one repair element is based at least partially on the diffraction effect of the chemical line exposure radiation in the repair element.

[0054] The step of generating at least one repair element may include the step of performing at least one local etching process and / or the step of performing at least one local deposition process by at least one focused particle beam and at least one precursor gas.

[0055] At least one focused particle beam may include at least one element from the group of photon beams, electron beams, ion beams, atomic beams, and molecular beams.

[0056] At least one precursor gas may include at least one element from the group of etching gas, deposition gas and additive gas.

[0057] The etching gas may include at least one element from the group consisting of halogen (F2, Cl2, Br2, I2), oxygen (O2), ozone (O3), hydrochloric acid (HCl), hydrogen fluoride (HF), xenon difluoride (XeF2), xenon tetrafluoride (XeF4), xenon hexafluoride (XeF6), xenon chloride (XeCl), argon fluoride (ArF), krypton fluoride (KrF), sulfur difluoride (SF2), sulfur tetrafluoride (SF4), sulfur hexafluoride (SF6), nitrosyl chloride (NOCl), phosphorus trichloride (PCl3), phosphorus nitrate (PCl5), phosphorus trifluoride (PF3), nitrogen trifluoride (NF3), water vapor (H2O), hydrogen peroxide (H2O2), dinitrogen oxide (N2O), nitrogen oxide (NO), nitrogen dioxide (NO2), and nitric acid (HNO3).

[0058] At least one deposited gas may include at least one element from the group of metal alkyl, transition element alkyl, main group alkyl, metal carbonyl, transition element carbonyl, main group carbonyl, metal alkoxide, transition element alkoxide, main group alkoxide, metal complex, transition element complex, main group complex, and organic compound.

[0059] Metal alkyl, transition element alkyl, and main group alkyl may include at least one element from the group of cyclopentadienyl (Cp)trimethylplatinum (CpPtMe3), methylcyclopentadienyl (MeCp)trimethylplatinum (MeCpPtMe3), tetramethyltin (SnMe4), trimethylgallium (GaMe3), ferrocene (Co2Fe), and bisarylchrome (Ar2Cr).

[0060] Metal carbonyls, transition element carbonyls, and main group carbonyls include chrome hexacarbonyl (Cr(CO)6), molybdenum hexacarbonyl (Mo(CO)6), tungsten hexacarbonyl (W(CO)6), dicobalt octacarbonyl (CO2(CO)8), and triruthenium dodecacarbonyl (Ru3(CO)6). 12 It may include at least one element from the group of iron pentacarbonyl (Fe(CO)5).

[0061] Metal alkoxides, transition element alkoxides, and main group alkoxides may include at least one element from the group of tetraethylorthosilicate (TEOS, Si(OC2H5)4) and tetraisopropoxytitanium (Ti(OC3H7)4). Metal halides, transition element halides, and main group halides may include at least one element from the group of tungsten hexafluoride (WF6), tungsten hexachloride (WCl6), titanium hexachloride (TiCl6), boron trichloride (BCl3), and silicon tetrachloride (SiCl4).

[0062] The metal complex, transition element complex, and main group complex may include at least one element from the group of copper bis(hexafluoroacetylacetonate) (Cu(C5F6HO2)2) and dimethylgold trifluoroacetylacetonate (Me2Au(C5F3H4O2)).

[0063] The organic compound may include at least one element from the group of carbon monoxide (CO), carbon dioxide (CO2), aliphatic hydrocarbons, aromatic hydrocarbons, vacuum pump oil, and components of volatile organic compounds. Aromatic hydrocarbons may include styrene.

[0064] At least one additive gas may include at least one element from the group of oxidizing agents, halides, and reducing agents.

[0065] The oxidizing agent may include at least one element from the group consisting of oxygen (O2), ozone (O3), water vapor (H2O), hydrogen peroxide (H2O2), dinitrogen oxide (N2O), nitric oxide (NO), nitrogen dioxide (NO2), and nitric acid (HNO3). The halogen may include at least one element from the group consisting of chlorine (Cl2), hydrochloric acid (HCl), xenon difluoride (XeF2), hydrogen fluoride (HF), iodine (I2), hydrogen iodide (HI), bromine (Br2), hydrogen bromide (HBr), nitrosyl chloride (NOCl), phosphorus trichloride (PCl3), phosphorus pentachloride (PCl5), and phosphorus trifluoride (PF3). The reducing agent may include at least one element from the group consisting of hydrogen (H2), ammonia (NH3), and methane (CH4).

[0066] At least one generated repair element may overlap at least partially with at least one defect. The deposited repair element may comprise the material of the lithographic mask. The deposited repair element may be a metal, e.g., chromium (Cr), a metal compound, e.g., tantalum nitride (TaN), silicon (Si), silicon dioxide (SiO2), and silicon molybdenum oxynitride (Mo x SiO y N z It may include ), where 0 <x≤0.5, 0≤y≤2 및 0≤z≤4 / 3이다. 에칭된 수리 요소는 포토리소그래픽 마스크의 재료를 에칭할 수 있다. 에칭된 수리 요소는 상기 언급된 마스크 재료를 포함할 수 있다.

[0067] The above-described method may include: (a) generating at least one repair element by at least one repair shape in which a parameter is defined by at least one defect; and (b) identifying a parameter of a repair shape for a residual defect residue—the step of identifying a parameter of a repair shape for a residual defect residue includes assigning at least one numeric value to the parameter that deviates from a numeric value predetermined by the residual defect residue for the parameter.

[0068] The method defined above can be used as a second step of a general defect repair process. In this case, in the first step, large defects—that is, defects larger than the resolution limit of the photomask—can be repaired by implementing a repair shape in the form of a local etching process or a local deposition process, or by generating a corresponding repair element. Then, the repaired mask is inspected. During the inspection of the mask, if it is established that the repaired location still does not always satisfy the specifications, a repair shape is determined for the residual defect, and the parameters of this repair shape are determined according to the method of the present invention. Here, the residual defect is assumed to constitute small defects, that is, defects having at least one dimensional size at at least one dimension smaller than the resolution capability or resolution limit of the photolithographic mask. The residual defect can subsequently be repaired or compensated by generating a corresponding repair element based on the repair shape identified in the second step.

[0069] In a second embodiment, a method for repairing at least one defective pattern element of a lithographic mask comprises: (a) determining at least one repair element of a lithographic mask that does not image the lithographic mask during exposure, wherein the at least one repair element is configured to change the imaging behavior of the at least one defective pattern element; and (b) generating at least one repair element on the lithographic mask using at least one focused particle beam and at least one precursor gas.

[0070] At least one generated repair element may have a dimensional size at least one dimension smaller than the resolution limit (R) of the photomask. As previously described, averaging the chemical radiation for a structure having a dimensional size lower than the resolution capability of the mask results in a reduced effect on the placement error of the repair element. This situation significantly facilitates the localization of the repair element(s) in relation to the location of the defect to be repaired. Furthermore, due to the non-imaging of the repair element(s), its geometric shape(s) can deviate significantly from the shape of the defect without significantly adversely affecting the defect compensation. This fact significantly simplifies the repair or compensation of small defects.

[0071] However, the repair element(s) generated on the mask locally alter the diffraction behavior of the mask at the chemical line wavelength. Then, the repair element(s) are designed such that the repair element(s) combined with the defective pattern element substantially realize the imaging behavior of the corresponding defect-free region of the photolithographic mask.

[0072] At least one repair element may have at least one dimensional size including a range of 10% to 90%, preferably 20% to 80%, more preferably 30% to 70%, and most preferably 40% to 60% of the resolution limit of the lithographic mask.

[0073] The distance between at least one repair element and at least one defect pattern element may include a range of 2% to 80%, preferably 2% to 50%, more preferably 2% to 30%, and most preferably 2% to 10% of the resolution limit of the lithographic mask.

[0074] At least one dimensional size of at least one repair element may include a range of 10% to 90%, preferably 20% to 80%, more preferably 30% to 70%, and most preferably 40% to 60% of the resolution limit of the lithographic mask.

[0075] Additional aspects of the repair element have been described in detail in relation to the first embodiment.

[0076] A computer program may include instructions that, when executed by a computer system, cause the computer system to perform steps of one of the modes described above.

[0077] In one embodiment, an apparatus for repairing at least one defect of a lithographic mask includes means for verifying at least one repair shape parameter for at least one defect, wherein the means for verifying the parameter includes means for assigning to at least one parameter a numeric value that deviates from a predetermined numeric value for at least one defect for said parameter.

[0078] The device may also include means for generating at least one repair element on a lithographic mask by a verified repair shape.

[0079] Means for verifying parameters of at least one repair shape may include at least one coprocessor configured to determine parameters of at least one repair shape from measurement data and design data of at least one defect of a lithographic mask. Additionally, at least one coprocessor may be configured to assign at least one numerical value deviating from a numerical value predetermined by at least one defect to at least one corresponding parameter. Assigning the deviation numerical value of the corresponding at least one parameter may be performed based on the resolution limit of the lithographic mask and the resolution limit when generating at least one repair element.

[0080] The resolution limit when generating at least one repair element is fundamentally influenced by two parameters. The first parameter is the minimum spot diameter at which the particle beam for generating the repair element can be focused. In the case of a photon beam, the achievable spot diameter is determined by the wavelength of the photon. To generate a repair element on a mask in the DUV wavelength range using a photon beam, photons from the EUV wavelength range are required. EUV photon sources are still very expensive at present. Therefore, for the purpose of generating repair elements, it is advantageous to use a massive particle beam, such as an electron beam, whose resolution limit is given by the de Broglie wavelength. Electron beams can currently be focused to spot diameters in the range of several nanometers. The positioning accuracy of electron beams is significantly higher and extends to the sub-nanometer range.

[0081] A second parameter determining the resolution limit when generating repair elements is the interaction region or scattering cone of secondary electrons generated by a beam of massive particles. The diameter of said interaction region on the mask surface determines the extent of the local chemical reaction initiated by the particle beam and at least one precursor gas. The size of the interaction region depends on the energy of the particles incident on the photomask. Additionally, the local material composition of the photomask at the interaction site significantly affects the size of the interaction region. Currently, the local chemical reaction can be limited to a lateral dimension of about 5 nm.

[0082] Means for verifying parameters of at least one repair shape may include at least one algorithm implemented as an application-oriented integrated circuit (ASIC), a complex programmable logic circuit (CPLD, complex programmable logic element) and / or a field programmable gate array (FPGA).

[0083] Means for verifying at least one parameter of a repair shape may include at least one trained machine learning model. Additionally, means for assigning a numerical value deviating from a predetermined numerical value by at least one defect to a corresponding at least one parameter may include a trained machine learning model.

[0084] Means for verifying parameters of at least one repair shape may include at least one element from the group of a mask inspection device, an interferometer, a confocal microscope, a scanning particle microscope, and a scanning probe microscope. The element from said group may record measurement data from at least one defect.

[0085] Means for generating at least one repair element may include at least one focused particle beam configured to perform a local chemical reaction and at least one precursor gas.

[0086] In a second embodiment, an apparatus for repairing at least one defective pattern element of a lithographic mask comprises: (a) means for determining at least one repair element of a lithographic mask that does not image the lithographic mask during exposure, wherein the repair element is configured to change the imaging behavior of at least one defective pattern element; and (b) means for providing a focused particle beam and at least one precursor gas configured to generate at least one repair element on the lithographic mask.

[0087] Means for determining at least one repair element may include at least one coprocessor configured to determine at least one repair element from measurement data and design data of at least one defect of a lithographic mask. Determining at least one repair element may be performed based on the resolution limit of the lithographic mask and the minimum spot size of the focused particle beam at its focusing portion.

[0088] Means for determining at least one repair element of a lithographic mask may include at least one algorithm implemented as a specific application-oriented integrated circuit (ASIC), a complex programmable logic circuit (CPLD, complex programmable logic element) and / or a field programmable gate array (FPGA).

[0089] Means for determining at least one mathematical element may include at least one trained machine learning model.

[0090] Means for determining at least one repair element may include at least one element from the group of a mask inspection device, an interferometer, a confocal microscope, a scanning particle microscope, and a scanning probe microscope.

[0091] Further embodiments include a lithographic mask repaired according to any of the methods described above. The repaired lithographic mask may be used in an exposure system. Additionally, the repaired lithographic mask may be repaired based on at least one repair shape determined according to any of the embodiments described above. Additionally, the repaired lithographic mask may include at least one repair element created by using the determined at least one repair shape. The lithographic mask may be repaired by performing a particle beam-induced local deposition process and / or a local etching process.

[0092] In another embodiment, the exposure system uses a lithographic mask repaired according to any of the steps of the method described above.

[0093] The exposure system may be a photolithographic exposure system. In particular, the exposure system may be a microlithographic projection exposure system. The exposure system may be any type of exposure system, for example, an exposure system suitable for using a transmissive lithographic mask and an exposure system suitable for using a reflective lithographic mask.

[0094] The resolution limit of the exposure system can be determined by the wavelength of the light source of the exposure system, the numerical aperture of the projection lens of the exposure system, and the exposure settings of the exposure system.

[0095] At least two resolution limits may be specified for an exposure system applied to two ends of the projection lens or projection objective lens of the exposure system.

[0096] The resolution limit of the exposure system may be the resolution limit of the wafer side of the projection lens of the exposure system.

[0097] The wafer-side resolution limit is It can be determined by, where λ is the chemical line wavelength of the lithographic mask, and NA Wis the wafer-side numerical aperture of the projection lens of the exposure system, and σ is the exposure setting of the exposure system suitable for exposing a lithographic mask to project a pattern of the lithographic mask onto the photoresist arranged on the wafer.

[0098] NA on the wafer side of the projection lens, i.e., NA W It is typically selected to be as large as possible to obtain the lowest possible wafer-side resolution limit. When using an immersion solution, NA W can be greater than 1, for example, 1.3.

[0099] The mask-side resolution limit of the projection lens of the exposure system is It can be determined by, where λ is the chemical line wavelength of the lithographic mask, and NA M is the mask-side numerical aperture of the projection lens of the exposure system, and σ is the exposure setting of the exposure system suitable for exposing a lithographic mask to project a pattern of the lithographic mask onto the photoresist arranged on the wafer.

[0100] Mask opening (NA) M ) and wafer-side numerical aperture (NA W The ratio of ) defines the magnification M of the projection lens: For transmission lithographic masks, the magnification typically has a numerical value of M=1 / 4 or 1 / 5 for 4x and 5x reduction projection lenses. This is NA W For an immersion projection lens having =1.2 and M=4, the mask-side numerical aperture (NA M This means that ) is 0.30. In the case of a reflective lithographic mask, the magnification may depend on the direction in which the exposure beam scans across the wafer. For example, M can be 1 / 8 in the scanning direction, and M can be 1 / 4 in the direction perpendicular to the scanning direction. Brief explanation of the drawing

[0101] The following detailed description describes a currently preferred exemplary embodiment of the present invention with reference to the drawings, in which: The upper portion image of FIG. 1 illustrates a schematic section through a pattern element of a photolithographic mask, said pattern element having edges at positions predetermined by design, and the lower portion image schematically shows the effective dose distribution of optical intensity around the mask segment shown in the upper portion image of the photoresist during exposure of the photomask. The upper part of Figure 2 shows a schematic section through a mask with defect pattern elements, and the lower part shows the effect of defect pattern elements on the effective dose distribution applied to the photoresist. The upper part image of FIG. 3 shows the repair of a defect pattern element of a photolithographic mask from FIG. 2 according to the prior art, and the lower part image shows the effect of the defect repair on the optical intensity distribution of the repaired mask segment from the upper part image. The upper part image of FIG. 4 shows the ideal repair of a defect pattern element of a photolithographic mask from FIG. 2 based on the method according to the present invention, and the lower part image shows the optical intensity distribution generated by the optimally repaired photomask in the photoresist of a wafer. The upper part image of FIG. 5 shows the actual repair of the defect pattern element of the mask from FIG. 2 based on the method according to the present invention, and the lower part image shows the effect of the repaired photomask on the optical intensity distribution generated in the photoresist. FIG. 6 schematically illustrates an alternative exemplary embodiment of a repair element for compensating for a defect from FIG. 2. The upper portion image (715) of FIG. 7 schematically shows the perfectly placed edge of a pattern element, the portion image (735) shows a pattern element having a defect of excess absorber material, the portion image (755) shows the repair of the defect from the portion image (735) according to the prior art, the portion image (775) shows the compensation of the defect from the portion image (735) by the creation of a repair element described in the present application, and the portion image (795) shows an exemplary second embodiment of a repair element according to the present invention. The upper portion image (805) of FIG. 8 shows a pattern element of a photolithographic mask, the edge of which is accurately positioned at a location provided by the design, the central portion image (835) shows the repair of a defect in the missing absorber material by the creation of a repair element according to the present invention, and the lower portion image (855) shows the repair of a defect in the excess absorber material by the creation of a repair element according to the present invention. The upper part of the image in Fig. 9 shows a schematic plan view of a strip structure with defects, and the lower part shows a strip structure after repair of defects according to the prior art. The upper portion image of FIG. 10 reproduces the upper portion image of FIG. 9, the central portion image illustrates an exemplary repair of a defect from the upper portion image according to one of the methods described in the present application, and the lower portion image illustrates the averaging of the complex amplitude of chemical radiation over a length scale of optical resolution. The upper portion image (1105) of FIG. 11 schematically illustrates a plan view of an absorbent angled pattern element, the portion image (1125) shows a pattern element having a defect of excess absorber material, the portion image (1145) shows the repair of the defect from the portion image (1125) according to the prior art, the portion image (1165) shows the compensation of the defect from the portion image (1125) by the creation of a repair element described in the present application, and the portion image (1185) shows the creation of a second example of a repair element according to the present invention. Figure 12 shows the repair of a defect in the missing absorber material according to the scheme shown in Figure 11. The top part image (1305) of FIG. 13 shows a strip structure of a mask arranged on a substrate, where the pattern element has a defect of missing absorber material, the middle part image (1335) shows a precise simulation of a defective mask segment, and the bottom part image (1365) shows the optical intensity distribution of the defective mask segment. FIG. 14 shows the mask segment of FIG. 13 after the defect has been optimally repaired by the method according to the present invention. Figure 15 shows the mask segment of Figure 13 after repairing the defect of Figure 13 with a repair element that is not optimally positioned. The upper left portion image (1605) of FIG. 16 shows a mask segment having a defect, said mask segment is measured by a scanning electron microscope, the upper right portion image (1635) shows a repair element in the mask segment from the upper left portion image, said repair element compensates for the defect, and the lower left portion image (1665) shows the repaired mask segment from the upper left portion image (1605). The upper portion image (1705) of FIG. 17 shows the normalized CD variation within the defect mask segment from the upper left portion image of FIG. 16, and the lower portion image (1755) shows the normalized CD variation of the repaired mask segment from the lower left portion image (1665) of FIG. 16. The upper portion image (1805) of FIG. 18 shows a schematic section through a strip structure having a defect, the portion image (1825) shows a mask segment from the portion image (1805) around the repaired defect, the portion image (1835) shows CD variation caused by the defect from the upper portion image, and the lower portion image (1865) shows CD variation still remaining after the defect repair. FIG. 19 shows a flowchart of a first embodiment of a method for repairing at least one defect in a lithographic mask. FIG. 20 shows a flowchart of a second embodiment of a method for repairing at least one defect in a lithographic mask. FIG. 21 shows a schematic section through a device for repairing one or more defects in a lithographic mask. FIG. 22 illustrates a schematic section through an optical mask inspection device and a comparison with the scanner of a photolithographic exposure system. FIG. 23 shows a schematic section through a device for repairing at least one defect pattern element of a lithographic mask. FIG. 24 illustrates a schematic section through a device realizing a particle beam source and gas supply system from FIG. 23. Specific details for implementing the invention

[0102] Current preferred embodiments of the method and apparatus according to the present invention for repairing one or more defects in a lithographic mask are described in more detail below. Additionally, exemplary embodiments of the method and apparatus according to the present invention for repairing defective pattern elements in a lithographic mask are described in detail below. The method according to the present invention is described based on an example of a binary photomask in the deep ultraviolet (DUV) wavelength range. However, it is not limited to improving the repair of defective DUV masks. Furthermore, the method according to the present invention is described primarily based on edge placement errors of pattern elements. However, these methods are not limited to repairing this type of error. Rather, they can be used particularly to facilitate the repair of any type of small defect and various types of photolithographic masks. The masks may include transmissive and reflective photomasks. Additionally, small defects in binary and / or phase-shifting masks can be repaired in the same way as small defects in multi-exposure masks. Hereinafter, the terms mask or photomask are also intended to encompass templates for nanoimprint lithography.

[0103] Additionally, the apparatus according to the present invention for repairing one or more defects in a lithographic mask is described based on an example of a modified scanning electron microscope. However, the apparatus according to the present invention is not limited to being realized only based on a scanning electron microscope. Rather, the apparatus according to the present invention may be based on any scanning particle microscope, that is, the apparatus defined in this application may use any type of particle, preferably a type of particle having mass, to inspect and / or create one or more repair elements of a photomask.

[0104] The top portion image (105) of FIG. 1 shows a schematic section through a one-dimensional (1D) segment of a photolithographic mask (100). The mask (100) may be a transmissive or reflective mask (100). In the example of FIG. 1, the photomask (100) comprises a binary transmissive mask (100). The photolithographic mask (100) comprises a substrate (110) having a surface (115). A pattern element (120) or a structural element (120) having a surface (125) is arranged on the surface (115) of the substrate (110). The substrate (110) may comprise a quartz substrate and / or a material with a low coefficient of thermal expansion (LTE (low thermal expansion) substrate). In the case of a transmissive photomask (100), its substrate (110) is substantially optically transparent to electromagnetic radiation at chemical line wavelengths. The pattern element (120) may be a structural element (120) of the binary photomask (100). In this case, the pattern element (120) may include elements of the absorber structure (120), and may include, for example, chromium. The absorbent pattern element (120) substantially absorbs all electromagnetic radiation at the chemical line wavelength incident on the pattern element (120). In the case of a DUV mask, the thickness of the pattern element (120) is in the range of 60 nm to 200 nm. The absorbent pattern element of an EUV mask currently has a layer thickness in the range of 50 nm to 70 nm (not shown in FIG. 1).

[0105] In this specification and in other parts thereof, the expression “substantially” indicates a measurement variable within a conventional measurement error when a measuring instrument according to the prior art is used to determine the measurement variable.

[0106] Additionally, the pattern element (120) may include a structural element (120) that shifts the phase of the chemical radiation with respect to the radiation incident on the substrate (110) and absorbs a portion of the light of the chemical radiation wavelength incident on the pattern element (120). An example of such a mask is an AttPSM (attenuated phase shifting mask) based on molybdenum silicide or silicon nitride. Such a mask typically transmits 6% to 20% of the incident optical intensity in the dark region with a phase shift of 180° compared to the transparent region of the photomask.

[0107] However, the pattern element (120) may be a structural element (120) of a simple phase-shifting photomask (100). A simple phase-shifting mask (100) can be created, for example, by etching a corresponding pattern into the substrate (110) of the mask (100), in which case the substrate substantially comprises quartz (SiO2). This type of mask is called a CPL (chromium-free phase-shifting) mask. An additional example of a simple phase-shifting mask type is an AltPSM (alternating phase-shifting mask) mask.

[0108] The upper portion image (105) of FIG. 1 shows an ideal edge (130) predetermined by design, said edge having a sidewall angle (135) of substantially 90°. Additionally, the edge (130) is precisely positioned at a predetermined location by design.

[0109] The bottom portion image (155) of FIG. 1 schematically illustrates the effective dose distribution (160) of optical intensity during exposure of the pattern element (120) of the photomask (100). The pattern element (120) absorbs radiation incident on the mask (100) from above, so that no optical intensity is substantially detected in the photoresist of the wafer at a certain distance from the edge (130) below the pattern element (120). In the transparent substrate area (110) of the mask (100), the substrate area is separated from the edge (130) of the pattern element (120), and the photoresist applied to the wafer is exposed to maximum optical intensity. The edge (130) of the pattern element (120) is typically defined as the location where the optical intensity of the photoresist reaches 50% of the maximum optical intensity or maximum effective dose. In the lower part image (155) of FIG. 1, this is illustrated by the intersection of the vertical line (180) and the horizontal line (170) indicated by a dotted line.

[0110] In the upper portion image (205) of FIG. 2, the edge (230) of the pattern element (220) is not positioned at a location predetermined by the design. Rather, the edge (230) is at a distance (dx) and is indicated by the horizontal double arrow (240). The distance (dx) from the target position of the edge (230) as predetermined by the design may occur as a result of the incorrect placement of the pattern element (220), and its dimensional size has a value predetermined by the design. In this case, the pattern element (220) can be corrected by identifying two repair shapes and generating two repair elements using said repair shapes. Meanwhile, the first repair element defines the deposition of missing absorber material in the area identified by the double arrow (240) in FIG. 2. Meanwhile, the second repair shape specifies the removal of excess absorber material of the pattern element (220) or absorber material located in an incorrect position (not shown in FIG. 2).

[0111] In the following, it is assumed that the edge of the pattern element (220), which is not reproduced in 1D in FIG. 2, is correctly positioned, and that the pattern element (220) is corrected only by the deposition of missing absorber material in the region of the edge (230). Additionally, in the following, it is assumed that the defect (240) is a small defect. This means that the dimensional size (dx) of the one-dimensional defect (240) is smaller than the resolution limit of the photolithographic mask (200) at its chemical line wavelength. The positioning or placement accuracy of the edge (130, 230) of the pattern element (120, 220) is very important, particularly due to the overlay problem for a lithography system that performs multiple exposure steps to fix the pattern element to the photoresist.

[0112] The bottom portion image (255) of FIG. 2 shows an optical intensity distribution (260) or effective dose distribution (260) generated by the misplacement or positioning of the edges (230) of the pattern elements (220) arranged on the wafer during the exposure of the photolithographic mask (200) with chemical electromagnetic radiation. During the exposure of the wafer, the positional error of the edges (230) of the pattern elements (220) (said as dx in the one-dimensional example of FIG. 2) is converted by the mask (200) into an edge placement error (EPE), schematically illustrated by the two arrows (280) of the bottom portion image (255) of FIG. 2. The relationship between the EPE (280) on the wafer and the mask defect (240) is given by the following equation: This is explained by. In this case, in the example reproduced in FIG. 2, the mask defect (240) or mask error (240) indicates that the edge (230) of the pattern element (220) is mispositioned by dx. The mask enhancement factor (MEEF) (mask error enhancement factor) signifies the magnification or enhancement of the mask defect (240) that may occur due to the mask (200). For simplification, MEEF=1 is assumed to be true below. Factor M represents the magnification or reduction by which the projection lens of the exposure system images the pattern element (220) of the mask (200) on the wafer. For the projection lens currently in use, M=1 / 4 or M=1 / 5 is true. Like FIG. 2, subsequent drawings do not depict the relationship between the mask defect (240) and the EPE (280) in a scale-following manner.

[0113] Ideal repair of the mask defect (240) by perfect implementation of the perfect repair shape is achieved by placing a repair element on the defect edge (230) of the pattern element (220) so that the defective pattern element (220) looks like the pattern element (120) shown in FIG. 1. However, this would require the ideal positioning of the repair tool in relation to the edge (230) of the defective pattern element (220). Additionally, perfect repair of the defect (240) would be based on the premise that the resolution limit of the repair tool is very small, and ideally 0.

[0114] The upper portion image (305) of FIG. 3 illustrates the actual repair of the defect (240) from FIG. 2 according to the prior art. Due to the limited positioning accuracy of the repair tool, the latter cannot be perfectly aligned with the edge (230) of the defect pattern element (220). Therefore, the repair element (310) created by the repair shape does not ideally correspond to the 1D dimensional size (dx) of the defect (240). Rather, due to the limited positioning accuracy of the repair tool, the repair shape of the defect (240) creates a small portion (320) of the repair element (310) on the surface (125) of the defect pattern element (220). For this reason, the edge (330) of the repair element (310) deviates from the position predetermined by the design. The remaining defect (340) has a dimensional size (dx').

[0115] The lower portion image (355) of FIG. 3 illustrates the change in optical intensity distribution (360) or EPE in the photoresist during exposure of the repaired mask segment (300) from the upper portion image of FIG. 3. Compared to the exposure of the unrepaired mask (200), the repaired mask (300) is much closer to the pattern geometry predetermined by the design from FIG. 1. However, the difference or EPE (380) indicated by the two arrows (380) still remains, resulting in a Critical Dimension (CD) deviation of the repaired pattern element (220) that is outside the acceptable error budget of the mask (300). The defect (340) remaining after repair according to the prior art may be repaired or compensated for as described below. However, it is more advantageous to repair the defect (240) directly, i.e., in a single step as described below with reference to FIG. 4.

[0116] The upper portion image (405) of FIG. 4 illustrates the best possible repair of a defective pattern element (220) from FIG. 2 according to one of the methods described in the present application. The repair shape for correcting the defective pattern element (200) from FIG. 2 is designed such that the implementation of the repair shape creates a repair element (410) at a distance (420) from the edge (230) of the defective pattern element (220). The ideal distance (420) between the edge (230) of the defective pattern element (220) and the repair element (410) is approximately dx / 2, that is, half the distance between the edge (230) and the vertical line (170) or the edge error (240) of the defective pattern element (220). When the height (430) of the repair element (410) corresponds substantially to the height of the pattern element (220), the repair element (410) and thus the repair shape underlying the repair element (410) have a dimensional size corresponding to approximately 80% of the distance (240) between the edge (230) of the pattern element (220) and the vertical line (170). As already described above, the mask-side resolution limit of the DUV mask is within the range of 150 nm to about 300 nm, and the resolution limit of the EUV mask is within the range of about 50 nm to 100 nm.

[0117] The bottom portion image (455) of FIG. 4 shows the optical intensity distribution (460) of a repaired photomask (400) having repair elements (410) arranged at a distance (420) from the edge (230) of the defective pattern element (220) of the photomask (200). A comparison of FIG. 1 and FIG. 4 makes it clear that the optical intensity distribution (460) resulting from the combined effect of the defective pattern element (220) and the repair elements (410) creates a pattern element in the photoresist that precisely realizes the edge (130) predetermined by design. The repair elements (420) result in the ideal creation of a pattern element predetermined by design in the photoresist and thus on the wafer, even if the repair elements (410) are not adjacent to the edge (230) and the extent of the defect (240) does not correspond to the one-dimensional dimension size (440) of the repair elements (410).

[0118] The repair element (410) may be created by a repair tool described below in the context of FIGS. 21 through 24, based on a corresponding repair shape on the photomask (400). To this end, the repair tool may perform a local chemical deposition reaction using a particle beam and at least one precursor gas. The material composition of the repair element (410) may correspond to the material composition of the pattern element (220). However, the material composition of the repair element (410) may also deviate from the material of the pattern element (220) as long as the material of the repair element (410) substantially completely absorbs the chemical line wavelength of the photolithographic mask (420).

[0119] FIG. 4 illustrates the complete compensation of a defective pattern element (220) by creating a repair element (410) at a distance (420) from the edge (230) of the defective pattern element (220). The top part image (505) of FIG. 5 illustrates the repair of the defective pattern element (220) that can actually be performed. Due to the limited positioning accuracy of the repair tool, the identified repair shape cannot accurately position the repair element (510) at the location provided for the optimal compensation of the defective pattern element. In the example illustrated in FIG. 5, the distance (520) from the edge (230) of the defective pattern element (220) is smaller than in the optimal repair of the defective pattern element (220) as illustrated in FIG. 4. As is evident from the bottom portion image (555) of FIG. 5, the placement error (540) of the repair element (510) results in an almost imperceptible deviation of the optical intensity distribution (560) compared to the ideal optical intensity distribution (460) of FIG. 4 in the photoresist applied to the wafer. This means that the EPE of the mask (500), as illustrated by the arrow (580) of FIG. 5, is negligibly small after the repair of the mask by the creation of the repair element (510).

[0120] This is a placement error (540) of the repair element (510) during transfer to the photoresist of approximately factor 1 / R M It is caused by a stepwise downward decrease, where R M represents the resolution limit of the photolithographic mask. This is one of the main advantages of the method described herein for repairing small defects in a photolithographic mask. The sensitivity of the placement of repair elements (410, 510) to the EPE (580) is significantly reduced when the method according to the present invention is performed. This reduction in the sensitivity of the placement of one or more repair elements (410, 510) is the resolution limit (R) of the photolithographic mask (400, 500). MIt is explained by ). The repair element (410, 510) is the resolution limit (R) of the photolithographic mask (400, 500) without repair or compensation of the pattern element (220), i.e., EPE (580) that deviates more than 10% from the best possible optical intensity distribution (460) of FIG. 4. M It can be placed at a distance of about 1% to 30% of ).

[0121] FIG. 6 illustrates a second example of a repair element (610) that can be used to repair or compensate for a defective pattern element (220) from FIG. 2. Unlike the height (530) of the repair element (510) in FIG. 5, the height (630) of the repair element (610) in the top part image (605) of FIG. 6 is smaller than the height or thickness of the pattern element (220) of the mask (600). To compensate for the smaller height (630) of the repair element (610), the repair element (610) has a larger 1D dimension size (635) than the repair element (510) from FIG. 5. As can be seen from the bottom part image (655) of FIG. 6, the altered repair element (610) does not substantially lead to a change in the optical intensity distribution (660) in the photoresist. The EPE (680) is negligibly small—in a manner similar to that of FIG. 5.

[0122] In addition to reduced placement sensitivity of the repair elements (410, 510, 610), by implementing the identified repair shape, it is possible to generate repair elements (410, 510, 610) in which the lateral dimension size (435, 535, 635) deviates significantly from the 1D dimension size (dx) of the defect (240). Thus, the defect repair method disclosed in this application has reduced sensitivity in that the parameterized repair shape transfers one or more lateral dimension sizes of the defect (240) to the repair elements (410, 510, 610). In addition to the mitigated placement sensitivity, this is a second significant advantage of the method for repairing small defects in a photolithographic mask as described in this application. In particular, the repair elements (410, 510, 610) may be significantly smaller than the defect (240) to be repaired. This situation has a favorable effect on the time required to repair defects, that is, the generation of repair elements (410, 510, 610).

[0123] The application of the method described in this application for repairing defects in missing absorber material, i.e., clear defects, is illustrated in FIGS. 1 through 6. FIG. 7 illustrates how these methods can be used to repair defects in excess material, i.e., so-called dark defects. The upper portion image (705) of FIG. 7 shows pattern elements (720) arranged on the surface (715) of the substrate (710) of the mask (700). The edges (730) of the pattern elements (720) are precisely positioned at locations provided by the design, said locations identified by dotted vertical lines (170).

[0124] In the second partial image (735) from the top, the pattern element (740) has a defect (750) of excess material. The defect (750) has a 1D dimensional size (725) that extends beyond the edge (730) of the defect-free pattern element (720). The defect (750) may be caused by an incorrect placement of the pattern element (720) having a dimensional size that is exactly the size provided by the design. However, as already discussed in the context of FIG. 2, it is also possible that too little absorber material is removed in the area of ​​the edge (730) during the patterning of the mask (700).

[0125] Partial image (755) shows the repair of a defect (750) according to the prior art. When the ideal creation of the repair element (760) is achieved, the defect (750) can be completely repaired as shown in partial image (755). The challenges related to the repair of a defect (750), particularly a small defect, according to the prior art have already been described in the discussion of FIG. 3.

[0126] The repair of a defect (750) of excess material by the creation of a repair element (770) according to the present invention is illustrated in a partial image (775). The repair element (770) is not designed to remove excess material of the defect (750) by performing a local etching process. Rather, the creation of the repair element (770) removes a portion of the defect (750) and a small portion of the defective pattern element (740). After the creation of the repair element (770), the remaining defect residue (780) combined with the repaired pattern element (740) affects the imaging behavior of the repaired mask (700) in the area of ​​the repaired pattern element so that it has the same imaging behavior as the pattern element (720).

[0127] Finally, the bottom portion image (795) of FIG. 7 shows a second repair element (790) that can be created as an alternative to the repair element (770). The creation of the repair element (790) removes a portion of the material of the defective pattern element (740) in the region of the edge (730) of the error-free pattern element (720). The 1D dimension size of the repair element (790) is larger than that of the repair element (770). The larger 1D dimension size of the repair element (790) is compensated by the fact that the implementation of the repair shape underlying the repair element (790) does not etch the pattern element (740) under the surface (715) of the substrate (710) of the mask (700).

[0128] The implementation of repair elements (770, 790) for repairing defects (740) opens up additional degrees of freedom described above in the context of FIGS. 5 and 6.

[0129] The upper portion image (805) of FIG. 8 shows an absorbent pattern element (720) having a surface (725), and this pattern element is arranged on the surface (715) of a substrate (710) that is optically transparent to the chemical line wavelength of the photolithographic mask (800). The edge (730) of the pattern element (720) is precisely positioned at a location provided by the design. The position is indicated by a vertical line (170) shown as a dotted line in FIG. 8.

[0130] The central portion image (835) of FIG. 8 shows a defective pattern element (840). The edge (830) of the pattern element (840) is not located at the position provided by the design of the mask. As a result, the pattern element (840) forms a bright defect or a defect of missing absorber material. In the portion image (835), the defect is compensated by implementing a corresponding repair shape to create a repair element (850). In the example of FIG. 8, the repair element (850) is adjacent to the edge (830) of the defective pattern element (840). The repair element (850) includes deposited absorber material that extends beyond the range of the missing absorber material by a 1D dimension size. The height of the exemplary repair element (850) is approximately half the height of the pattern elements (820, 840).

[0131] The bottom portion image (855) of FIG. 8 shows a pattern element (860) having a defect in excess absorber material. The defect in excess material is repaired by removing the top portion of the excess absorber material by creating a repair shape (870). The edge (880) of the repaired pattern element (860) does not correspond to a position predetermined by the design, said position is indicated by a dashed line (170) in FIG. 8. Nevertheless, the repair element (870) removes the effect of the defect in excess material except for a negligibly small portion.

[0132] Due to the reduced placement requirements of the repair tool, the creation of the repair element (850, 870) can be performed at a lower cost compared to the creation of the repair element (310, 760) according to the prior art.

[0133] The upper portion image (905) of FIG. 9 shows a schematic plan view of a mask segment (900) having a strip structure ("line and space pattern") with two absorbent strips (920 and 930) arranged on an optically transparent substrate (910). In the example shown in FIG. 9, the left pattern element (920) has two-dimensional (2D) defects (940) of excess absorbent material.

[0134] The bottom portion image (955) of FIG. 9 illustrates the repair of a defect (940) by performing a local etching process using a particle beam and an etching gas. In the example of the repair process shown in the bottom portion image (955) of FIG. 9, the defect (940) is almost completely removed, except for a negligible defect residue (950). Generally, laterally extended defects (940) cannot be removed in the same precise manner as shown in the bottom portion image (955) of FIG. 9. As already described above, the precision with which the repair shape of the defect (940) or the repair tool can be positioned is limited in relation to the defect (940) or the pattern element (920). Additionally, both the photomask (900) and the repair shape undergo drift, for example, thermal drift, during the creation of the repair element. Accordingly, the repaired pattern element (920) has edge roughness that deviates from a predetermined error interval (not shown in FIG. 9) along the processed right edge. Therefore, creating a corresponding repair shape to remove a defect (940) extending along the pattern element (930) is a very complex and time-consuming process.

[0135] FIG. 10 illustrates the repair of a 2D defect (940) from FIG. 9 using one of the methods described in the present application. The top portion image (1005) of FIG. 10 again reproduces the top portion image (905) from FIG. 9. The central portion image (1035) of FIG. 10 illustrates the compensation of the defect (940) of excess material by the creation of a repair element (1010). The repair element (1010) created a locally etched structure in the defect (940), partially in the substrate (910) of the mask (900), and also partially in the pattern element (920). Various repair elements (1010) can be created based on a single repair shape created at various locations along the defect extending from the right edge of the pattern element. However, as illustrated in the central portion image (1035) of FIG. 10, it is also possible to identify a single repair shape that creates various repair elements (1010).

[0136] A major advantage of repairing a defect (940) as shown in partial image (1035) is the significantly reduced sensitivity of the placement or positioning of the repair element (1010) in relation to the defect (940). The lower partial image (1065) of FIG. 10 schematically illustrates the reduced sensitivity during the placement of the repair element (1010). The dashed ellipse (1050) illustrates an area of ​​the photolithographic mask (900) or a length scale in which the mask (900) averages the complex amplitude of the chemical line electromagnetic radiation. If a structure exists on the photomask with a dimensional size in one or two dimensions smaller than the resolution limit of the photomask, the chemical line radiation is averaged over the details of said structure. This means that while the microroughness of pattern elements has little to no effect on the imaging behavior of the photolithographic mask, the complex amplitude of electromagnetic chemical radiation averaged over dimensional sizes in the resolution limit region affects the imaging behavior of the photomask.

[0137] The dotted ellipse (1050) indicates an area of ​​the mask that provides an optical intensity portion at the imaging point of the mask in the photoresist. The scale of this averaging area is—as already described above—the resolution limit (R) of the photolithographic mask. M It is determined by the resolution limit (R). The dimensional size of the repair element (1010) is determined by the resolution limit (R). M Insofar as it is small compared to ), its size and placement are a factor 1 / R compared to both the size and location determination of repair elements from the prior art. M It is less sensitive up to. This means that a very wide range of different repair elements (1010) lead to the same or very similar results in relation to the imaging behavior of the photolithographic mask. The methods described in this application utilize this situation to improve the repair of small defects that have been particularly difficult to repair until now.

[0138] The resolution limits of today's DUV and EUV masks are described in detail.

[0139] FIGS. 11 and 12 illustrate the repair of defects in excess absorber material and missing absorber material, wherein the repair element may have a geometric shape different from the defect to be repaired. Partial image (1105) shows a mask segment (1100) having an angular pattern element (1120) arranged on a mask substrate (1110) and whose shape satisfies design specifications. In the partial image (1125) of FIG. 11, the pattern element (1130) has a defect (1140) of excess absorber material, and this defect is located inside the angular region of the pattern element (1130).

[0140] A partial image (1145) of FIG. 11 illustrates the repair of a defect by implementing a repair shape assigned to the defect according to the prior art, and this repair shape creates a repair element (1150) in the area of ​​the defect (1140). The problem related to this has already been thoroughly described in the context of FIG. 3 and FIG. 9.

[0141] Partial image (1165) of FIG. 11 illustrates a first example of the creation of a repair element (1170) that is much less sensitive than the repair element (1150) from partial image (1145) in relation to the lateral dimension size of the repair element (1170) with respect to the defect (1140) first, and the dimensional size of the repair element (1170) with respect to the dimensional size of the defect (1140). Finally, partial image (1185) shows a repair element (1190), the geometric shape of which is distinctly different from the shape of the defect to be repaired. A circular or circular repair element (1190) can be created much more simply than a rectangular or square repair element (1170). Nevertheless, the repair element (1190) in partial image (1185) results in a compensation of the defect (1140) that is by no means inferior to the compensation of the repair element (1170). The reason for this was explained in the discussion of Fig. 10.

[0142] Now, referring to FIG. 12, the repair of a defect in the missing absorber material is described. A partial image (1205) of FIG. 12 shows a square pattern element (1220) placed on an optically transparent substrate (1210) of a photolithographic mask (1200). In the pattern element (1230) in the partial image (1225), a part (1240) of the pattern element (1230) is missing. This means that the latter has a bright defect (1240) or a defect (1240) of the missing absorbent material.

[0143] A partial image (1245) of FIG. 12 illustrates the repair of a defect (1240) of missing material by applying a corresponding repair element (1250) across the defect (1240) by implementing a repair shape identified for the defect. The repair element (1250) is implemented according to the prior art.

[0144] A partial image (1245) of FIG. 12 illustrates a first example of a repair element (1270) that is generated according to one of the methods described in the present application and avoids sensitivity of the repair method in the partial image (1245). Additionally, a second exemplary example of a repair element (1290) illustrates that the repair element (1290) only partially reproduces the complex contour of the defect (1240) and thus can be generated much more simply than the repair element (1270).

[0145] The top portion image (1305) of FIG. 13 shows a top view of a mask segment (1300) having a strip structure with an absorbent strip-shaped pattern element (1320) applied on a substrate (1310) of a photomask (1300). The second pattern element (1320) from the left has a defect (1330) of missing material. The width of the defect (1330) was selected as 15% of the half-pitch of the strip structure. The middle portion image (1335) of FIG. 13 shows a precise simulation of the imaging behavior of the mask segment (1300) from the top portion image (1305). The defect (1330) is clearly visualized as a defect (1340) in the middle portion image (1335).

[0146] The bottom portion image (1365) of FIG. 13 shows the optical intensity distribution of the defective mask segment (1300). The dashed horizontal line (1360) indicates the maximum value of the optical intensity distribution or effective dose distribution of the error-free mask segment. The intersection point between the intensity distribution and the dashed horizontal line (1370) describes the point where the photoresist is considered to be exposed. Thus, the intersection point indicates the width of the strip-shaped pattern element created on the wafer by the mask segment (1300). The defect (1330) is mirrored in the wide central pattern element created on the wafer. Variations in the width of the pattern element imaged in the photoresist by the mask segment (1300) result in variations in CD (critical dimension) along the imaged pattern structure.

[0147] The upper portion image (1405) of FIG. 14 shows a defect mask segment (1300) for repairing a defect (1330) by the creation of a repair element (1440). In the example illustrated in FIG. 14, the repair element (1440) is not placed on the defect (1330) as is customary in the prior art, but rather is located right next to the defect (1330) of the missing absorber material. As can be seen from the central portion image (1435), the repaired defect (1330) is no longer visualized in the imaging simulation. Additionally, it can be clearly seen from the lower portion image (1465) that the repair element (1450) completely repairs the defect (1330) and thus there is no variation in optical intensity along the length of the pattern element (1320).

[0148] The top portion image (1505) of FIG. 15 illustrates the repair of a defect (1330) from FIG. 13, and during this repair, a repair element (1540) is placed at a distance of 15% of the half-pitch from the pattern element (1320). The absence of deviation from the defect-free mask structure is identifiable in the imaging simulation shown in the middle portion image (1535) of FIG. 15. A slight decrease in the maximum value of the optical intensity distribution is visualized in the bottom portion image (1565) of FIG. 15, resulting in a smaller width of the central pattern element on the wafer. However, the deviation in width is still within a predetermined error budget for the mask (1300).

[0149] The upper left portion of the image in FIG. 16 shows a mask segment (1600) recorded by a scanning electron microscope, said mask segment having a strip pattern (line and space pattern) with a half-pitch of 152 nm on the mask (1600). As indicated by the dotted vertical line (1625) with curvature, the pattern element (1620) of the mask segment (1600) has a defect (1630) of missing absorbent material. The exemplary defect (1630) in FIG. 16 has a range of about 10 nm perpendicular to the pattern element (1620). This range is the resolution limit (R) of the photolithographic mask (1600). M It is significantly smaller than ).

[0150] The repair element (1640) is schematically illustrated in the upper right portion image (1635), and the repair element is designed to be combined with an adjacent pattern element (1620) to compensate for the defect (1630). The lower left portion image (1665) of FIG. 16 is a difference image for the upper right portion image (1635) and represents the repaired mask segment from the upper left portion image (1605). As indicated by the straight vertical line (1645), the defect (1630) is no longer visualized in the record of the mask segment (1600).

[0151] The top portion image (1705) of FIG. 17 shows variations in CD for various pattern elements from the top left portion image (1605) of FIG. 16 following the pattern elements. The dashed line (1710) specifies the CD of the pattern element created on the wafer by the mask (1600)—predefined by the design—and in the example shown in FIG. 17, the CD is CD = 37.68 nm. The dashed horizontal lines (1720 and 1730) indicate the lower and upper limits of the allowable CD variation (ΔCD), respectively. The CD variation range (ΔCD) is ±2.5% in the example of FIG. 17. The curve in the top portion image (1705) indicates that the curve (1750) reaches the upper limit of the CD tolerance interval, and the CD curve (1760) proceeds far beyond the ±2.5% CD tolerance interval over most of its profile. The maximum value of the normalized CD variation for the curve (1760) is ΔCD / CD = 8.6%.

[0152] The bottom portion image (1755) of FIG. 17 shows CD variation for various pattern elements (1620) of the repaired mask segment from the bottom left portion image (1665) of FIG. 16. All CD curve profiles are within an acceptable tolerance range. That is, all CD curves satisfy the requirement ΔCD / CD < ±2.5%. This means that the repair element (1640) completely compensates for the defect (1630).

[0153] The top partial image (1805) of FIG. 18 shows a 1D section through a mask segment (1800) having eight pattern elements (1820) in the form of a strip structure arranged on an optically transparent substrate (1810) of the mask (1800). The eight pattern elements (1820) create seven optically transparent strips numbered consecutively from 0 to 6 in the partial image (1805). The half-pitch of the strip structure on the mask (1800) is 152 nm or 38 nm on the wafer. In the top partial image (1805), the second pattern element (1820) from the left has a defect (1840) of an excess absorbent and phase-shifting molybdenum silicide (MoSi) material at its right edge. To analyze the effect of the defect (1840), the strip structure of the defective mask segment (1800) shown in the partial image (1805) is simulated. The simulation parameters are NA=1.35, λ=193 nm, external σ value: 1.0, internal σ value: 0.88, exposure setting: Disar, polarization: y-direction, pattern: L&S (Line and Space) MoSi with 6% absorption at the chemical line wavelength. The electromagnetic radiation incident on the mask (1800) is assumed to be coherent.

[0154] The central part image (1835) of FIG. 18 shows the variation in CD for seven optically transparent strips 0 through 6 that follow the strip direction in the simulation, i.e., perpendicular to the plane of the drawing. Additionally, the shift of the center of gravity (CoG) of the optical intensity distribution is confirmed using the simulation. The dashed horizontal line (1870) represents a predefined target CD for the mask (1800). The curve in the central part image (1835) describes the variation in critical dimensions, i.e., ΔCD, along the pattern element (1820). The table in the central part image (1835) summarizes the ΔCD and ΔCoG for optically transparent strips 0, 1, and 2. The table shows that the variation in CD for strip 0 is 3.1 nm, which is significantly larger than the error budget of 2.5% of the mask (1800). The effect of the defect (1840) on the optically transparent strip 2 is the result of exposure of the mask (1800) by coherent radiation. The simulation was performed by the precise optical imaging program DrLitho at the Fraunhofer Institute for Integrated Device Technology (IISB) Erlangen.

[0155] The enlarged segment (1825) represents the pattern element (1820) around the defective transparent strip 1 after the repair of the defect (1840) by implementing the corresponding repair shape identified for the defect. When implementing the repair shape for local etching of the defect (1840), it is assumed that the local etching process generates a sidewall angle (1850) having a negative angle of -20°. The negative sidewall angle (1860) causes an amplitude defect in the aerial image or during the exposure of the wafer. A magnification vector of MEEF=1.4 is considered. The negative sidewall angle (1850) is compensated by a (positive) shift of the reference point (1860) of the edge (1830) so that the width or breadth of strip 1 corresponds to approximately half the size of the nominal width of the optically transparent strip. For an angle range of about ±20°, the error in the sidewall angle can be substantially fully compensated by a corresponding shift of the reference point of the edge (1830) of the repaired pattern element (1820). The lateral shift required for this is within a range of about ±3.3nm when λ=193nm.

[0156] The bottom partial image (1865) shows the simulated CD variation for optically transparent strips 0 through 6 along the strip direction. Additionally, the CoG shift of the optical intensity distribution is determined using simulation. The dashed horizontal line (1870) describes the target value of CD on the wafer. From a comparison of the sets of curves in partial images (1835 and 1865), it is immediately apparent that the repair of the defect (1840) significantly reduces the CD variation. The table in the bottom partial image (1865) summarizes the ΔCD and ΔCoG for optically transparent strips 0, 1, and 2—in a manner similar to the table in the middle partial image (1865). Compared to the table in the middle partial image (1835), the repair of the defect (1840) reduced the CD variation by more than one order of magnitude.

[0157] Repairing a defect (1840) of excess material was discussed in relation to FIG. 18, said defect adjacent to the edge (1820) of a pattern element without completely covering the transparent strip. It goes without saying that it is also possible to repair a defect of excess material that completely bridges two pattern elements. Additionally, the method described in the context of FIG. 18 can also be used to repair a defect of missing material.

[0158] The flowchart (1900) of FIG. 19 illustrates an exemplary first example of the method discussed in the present application for repairing at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800). The method begins at step (1910). In the next step (1920), at least one parameter of a repair shape for at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) is identified, and the step of identifying said parameter includes the step of assigning a numeric value deviating from a numeric value predetermined by at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) to the corresponding at least one parameter. The method ends in step (1930).

[0159] Additionally, the flowchart (2000) of FIG. 20 illustrates the steps of an exemplary second example of a method for repairing at least one defective pattern element (220, 740, 840, 860, 920, 1130, 1230, 1320, 1620, 1820) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1600, 1800). The method begins at step (2010). In the next step (2020), at least one repair element (410, 510, 610, 770, 790, 850, 870, 1010, 1170, 1190, 1270, 1290, 1440, 1540, 1640) of the lithographic mask that does not image the lithographic mask during exposure is determined, and said repair element changes the imaging behavior of at least one defective pattern element.

[0160] Then, in step (2030), at least one repair element is created on a lithographic mask by a focused particle beam and at least one precursor gas. The method ends in step (2040).

[0161] FIG. 21 schematically illustrates a section through a device (2100) designed to verify the parameters of a repair shape for at least one defect (240). To this end, the device (2100) includes an optical mask inspection device (2110). FIG. 22 illustrates the principle of an optical mask inspection device designed to record an aerial image of a transmissive mask (200). Some components of the scanner are schematically illustrated in the left partial image (2205) in FIG. 22. The exposure system focuses electromagnetic radiation of chemical wavelengths onto the photolithographic mask. A projection optical unit or projection lens has a large numerical aperture (NA). W Imaging radiation passing through a photomask with a photoresist distributed on a wafer having ) or a reduction on the wafer (typically 1:4 or 1:5).

[0162] The right-hand portion image (2255) of FIG. 22 shows some components of an optical mask inspection device (2110) designed for the chemical line wavelength of the scanner from the left-hand portion image (2205). The exposure systems of the scanner and the optical mask inspection device (2110) are substantially the same. This means that image generation is substantially the same in both systems. Accordingly, the optical mask inspection device (2110) images a segment of the optical intensity distribution of the mask, such as that incident on the photoresist arranged on the wafer. However, unlike in the case of the scanner, in the case of the optical mask inspection device (2110), the lens images a small segment of the optical intensity distribution of the photomask at a large magnification on a CCD (charge-coupled device) camera. As a result, defects that the photomask has at the chemical line wavelength appear in its aerial image, and it becomes possible to detect said defects using a CCD sensor or a CCD camera.

[0163] The optical mask inspection device (2110) of the device (2100) may provide measurement data of one or more aerial images to the computer system (2130) of the device (2100) through the connection part (2120). The computer system (2130) of the device (2100) may determine the parameters of the repair shape assigned to the defect in the aerial image from the measurement data of the aerial image(s). To this end, the computer system (2130) may include a coprocessor (2140) specifically designed to efficiently execute an algorithm for determining the parameters of the repair shape assigned to the defect from the aerial image(s) of the optical mask inspection device. Additionally, the computer system (2130) of the device (2100) may have a second algorithm designed to assign a value or numerical value that deviates from a numerical value predetermined by the defect to one or more parameters of the repair shape. The second algorithm may likewise be executed by the coprocessor (2140).

[0164] However, the computer system (2130) may also have a dedicated hardware component (2150) that executes one or both of the algorithms described above. The hardware component (2150) of the computer system may be implemented in the form of an ASIC (Application-Oriented Integrated Circuit), a composite programmable logic circuit (CPLD, composite programmable logic device), and / or a field programmable gate array (FPGA).

[0165] Additionally or alternatively, the computer system (2130) may include a dedicated graphics processor (2160) designed to implement a trained machine learning model. The machine learning model may be trained in at least two ways, or the graphics processor may implement two different trained machine learning models designed for each formulated problem. First, the machine learning model may be trained to identify parameters of one or more repair shapes implemented to repair or compensate for defect(s), i.e., to generate one or more repair elements (410, 510, 610), from the measurement data of the optical mask inspection device (2110), the design data of the lithographic mask (200), the exposure system, and optionally the settings of the RET structure created on the mask (200).

[0166] Alternatively or additionally, in the case of a mathematical shape that has already been parameterized, the machine learning model may assign different numerical values ​​to one or more parameters for the purpose of identifying the mathematical elements (410, 510, 610) described above according to the present invention. However, a currently preferred embodiment is for the machine learning model to directly predict the parameters of the mathematical shape from the input data described above for the purpose of forming one of the mathematical elements (410, 510, 610) described in this application. The process of training the machine learning model is not discussed in this application.

[0167] Additionally, the computer system (2130) may include non-volatile memory (2170) in which algorithm(s), machine learning model(s), and / or trained machine learning model(s) are stored. The non-volatile memory (2170) may include solid-state memory (SSD, solid-state drive).

[0168] Additionally, the computer system (2130) may include a control device (2180) designed to control the optical mask inspection device (2110).

[0169] Additionally, the device (2100) may include a scanning particle microscope, a scanning probe microscope, and / or a confocal microscope, which is designed to scan defects (240) of the photolithographic mask (200) and generate a pictorial representation of the measurement data. If the device (2100) includes one or more of these measuring instruments, the control device (2180) can likewise control these measuring instruments.

[0170] FIG. 23 schematically illustrates a section through a device (2300) capable of repairing at least one defective pattern element (220). To this end, the device (2300) includes an optical mask inspection device (2310). This type of measuring device has already been described in detail in the context of the discussion in FIG. 22.

[0171] The optical mask inspection device (2310) of the device (2300) may provide measurement data of one or more aerial images or aerial image stacks to the computer system (2330) of the device (2300) through the connection part (2320). The computer system (2330) of the device (2300) may be similar to the computer system (2130) of the device (2100) from FIG. 21. To avoid lengthy descriptions, the description of the computer system (2330) will be omitted. Rather, the discussion of FIG. 21 is referenced.

[0172] The device (2300) further includes a particle beam source (2350) capable of providing a focused particle beam. The focused particle beam of the particle beam source (2350) can first be used to analyze defects (240) of a photolithographic mask (200). Based on the measurement data of the focused particle beam and / or the measurement data of the optical mask inspection device (2310), the repair shape for the defect can be determined using one or more algorithms or one or more machine learning models. Next, the focused particle beam of the particle beam source (2350) combined with the gas supply system (2370) of the device (2300) can be used to repair the analyzed defects (240). Both the particle beam source (2450) and the gas supply system (2370) can exchange data with a computer system (2330) through connections (2340 and 2360). Additionally, the control unit (2370) of the computer system (2330) can control the optical mask inspection device (2310), the particle beam source (2350), and the gas supply system (2370).

[0173] FIG. 24 shows a schematic section through a device (2400) that combines a particle beam source (2350) and a gas supply system (2370) of a device (2300). Repair elements (410, 510, 610) for repairing a defect (240) can be generated by the device (2400). The exemplary device (2400) of FIG. 24 includes a scanning particle microscope (2410) modified in the form of a scanning electron microscope (SEM) (2410). The device (2400) includes a particle beam source (2350) in the form of an electron beam source (2405), which generates an electron beam (2415) as a particle beam (2415) having mass. The electron beam (2415) can be focused into a spot significantly smaller than the focusing diameter of a photon beam. Due to the small de Broglie wavelength of the electron, the electron beam (2415) can be focused to a spot diameter in the range of several nanometers. Thus, as an analysis or measurement tool, the electron beam (2415) has a very large lateral resolution capability.

[0174] Additionally, the electron beam (2415) has the advantage that—compared to an ion beam—electrons incident on a sample (2425), for example, a photolithographic mask (200) cannot substantially damage the sample (2425) or the photomask (200). However, it is also possible to use an ion beam, an atomic beam, or a molecular beam (not shown in FIG. 24) in the device (2400) for the purpose of processing the sample (2425).

[0175] A scanning particle microscope (2410) is composed of a column (2420) in which a beam optical unit (2413) in the form of an electron optical unit of, for example, the SEM (2410) is arranged, and an electron beam source (2405). In the SEM (2410) of FIG. 24, the electron beam source (2405) generates an electron beam (2415) that is sent as a focused electron beam (2415) onto a sample (2425) that may include a photolithographic mask (200) at a location (2422) by an imaging element arranged in the column (2420), said imaging element is not shown in FIG. 24. Thus, the beam optical unit (2413) forms an imaging system (2413) of the electron beam source (2405) of the device (2400).

[0176] Additionally, the imaging element of the column (2420) of the SEM (2410) can scan the electron beam (2415) across the sample (2425). The sample (2425) can be inspected using the electron beam (2415) of the device (2400). Typically, the electron beam (2415) is incident perpendicularly on the sample (2425).

[0177] Backscattered electrons and secondary electrons generated by the electron beam (2415) in the interaction region or scattering cone of the sample (2425) are recognized by the detector (2417). The detector (2417) arranged in the electron column (2420) is referred to as an "in lens detector." The detector (2417) may be installed in the column (2420) in various embodiments. The detector (2417) converts the secondary electrons generated by the electron beam (2415) at the measurement point (2422) and / or the backscattered electrons from the sample (2425) into an electrical measurement signal and transmits the latter to the evaluation unit (2480) of the device (2400). The evaluation unit (2480) analyzes the measurement signal from the detectors (2417 and 2419) and generates an image of the sample (2425) therefrom, and the image is displayed on the display (2495) of the evaluation unit (2480). The detector (2417) may additionally include a filter or filter system to distinguish electrons in terms of energy and / or solid angle (not reproduced in FIG. 24).

[0178] An exemplary device (2400) may include a second detector (2419). The second detector (2419) may be designed to detect electromagnetic radiation, particularly in the X-ray range. As a result, the detector (2419) enables the analysis of the material composition of the radiation generated by the sample (2425) during inspection. The detectors (2417 and 2419) may be controlled by a control unit (2370) of a computer system (2330). In an alternative embodiment, the device (2400) includes a dedicated control unit (not shown in FIG. 24).

[0179] Additionally, the device (2400) may include a third detector (not shown in FIG. 24). The third detector may be implemented in the form of an Eberhart-Tonley detector and is typically arranged outside the column (2420). It is generally used to detect secondary electrons.

[0180] The device (2400) may include an ion source that provides ions having low kinetic energy in an area of ​​the sample (2425) (not shown in FIG. 24). The ions having low kinetic energy can compensate for the charging of the sample (2425).

[0181] The sample (2425) is arranged on a sample stage (2430) or a sample holder (2430) for inspection purposes. The sample stage (2430) is also known in the art as a "stage." As indicated by the arrow in FIG. 24, the sample stage (2430) can be moved in three spatial directions relative to the column (2415) of the SEM (2410), for example, via a micromanipulator not shown in FIG. 24.

[0182] In addition to translational motion, the sample stage (2430) may be rotated about an axis oriented parallel to the beam direction of at least the particle beam source (2405). Additionally, it is possible to implement the sample stage (2430) so as to be rotatable about one or two additional axes, and these axes or these axes are arranged in the plane of the sample stage (2430). The two or three rotation axes preferably form a rectangular coordinate system.

[0183] The sample (2425) to be examined may be any arbitrary microstructured component or device requiring analysis and, suitably, subsequent processing, for example, repair of local defects (240) of the pattern elements (220) of the photolithographic mask (200).

[0184] Additionally, the device (2400) of FIG. 24 may include one or more scanning probe microscopes, such as an atomic force microscope (AFM) (not shown in FIG. 24), which can be used to analyze and / or process a sample (2425).

[0185] The scanning electron microscope (2410) illustrated in FIG. 24 is operated in a vacuum chamber (2470). To generate and maintain the required reduced pressure in the vacuum chamber (2470), the SEM (2410) of FIG. 24 has a pump system (2472).

[0186] A gas supply system (2370) realized by the device (2400) is discussed below. As already described above, a sample (2425) is arranged on a sample stage (2430). An imaging element of the column (2420) of the SEM (2410) can focus an electron beam (2415) and scan the latter across the sample (2525). The electron beam (2415) of the SEM (2410) can be used to induce a particle beam-induced deposition process (EBID, electron beam-induced deposition) and / or a particle beam-induced etching process (EBIE, electron beam-induced etching). The exemplary device (2400) of FIG. 24 has three different supply containers (2440, 2450, and 2460) for storing various precursor gases for the purpose of performing these processes.

[0187] The first supply container (2440) stores a precursor gas, for example, a metal carbonyl, for example, chrome hexacarbonyl (Cr(CO)6), or a main group metal alkoxide, for example, TEOS. Using the precursor gas stored in the first supply container (2440), missing material from the photolithographic mask (200) can be deposited thereon, for example, within the range of a local chemical deposition reaction. The missing material of the mask (200) may include missing absorber material, for example, chrome; missing substrate material (210), for example, quartz; missing material of the OMOG mask, for example, molybdenum silicide; or missing material of the multilayer structure of the reflective photomask, for example, molybdenum and / or silicon.

[0188] The electron beam (2415) of the SEM (2410) acts as an energy supplier for splitting the precursor gas stored in the first supply container (2440) at the site where material is intended to be deposited on the sample (2425). This means that the combined supply of the electron beam (2415) and the precursor gas leads to an EBID process performed for the local deposition of missing material, for example, missing material from the photomask (20). The modified SEM (2410) of the device (2400) may include a device for creating repair elements (410, 510, 610) on a photolithographic mask in combination with the precursor gas stored in the first supply container (2440).

[0189] As previously described, the electron beam (2415) can be focused to a spot diameter in the range of several nanometers. The interaction region or scattering cone where the electron beam (2415) generates secondary electrons depends first on the energy of the electron beam (2415) and second on the composition of the material that the electron beam (2415) strikes. The diameter of the interaction region obtains a value in the low single-order nanometer range. Accordingly, the diameter of the scattering cone of the electron beam (2415) limits the resolution limit achievable during the generation of the repair elements (410, 510, 610) by implementing the corresponding repair shape. The resolution limit is currently within the single-order nanometer range.

[0190] In the apparatus (2400) illustrated in FIG. 24, a second supply container (2450) stores an etching gas that enables a local electron beam-induced etching (EBIE) process to be performed. By using the electron beam-induced etching process, excess material can be removed from the sample (2425), and, for example, excess material of the pattern element (860) can be removed from the photolithographic mask (800). For example, the etching gas may include xenon difluoride (XeF2), a halogen, or nitrosil chloride (NOCl). Thus, the particle beam source (2350) is combined with the gas supply system (2370) to form the apparatus (2400) for producing repair elements (410, 510, 610).

[0191] An additive or additive gas may be stored in the third supply container (2460), and if necessary, said gas may be added to the etching gas kept available in the second supply container (2450) or the precursor gas stored in the first supply container (2440). Alternatively, the third supply container (2460) may store the second precursor gas or the second etching gas.

[0192] In the device (2400) illustrated in FIG. 24, each of the supply containers (2440, 2450, and 2460) has its own control valve (2442, 2452, and 2462) to monitor or control the amount of gas provided per unit time, i.e., the volumetric flow rate of the gas at the incident site (2422) of the electron beam (2415) for the sample (2425). The control valves (2442, 2452, and 2462) can be controlled and managed by the control unit (2370) of the computer system (2330). Thus, it is possible to set a wide range of gas or partial pressure ratios of gases provided at the processing location (2422) to perform the EBID and / or EBIE process.

[0193] Additionally, in the exemplary device (2400) of FIG. 24, each of the supply containers (2440, 2450 and 2460) has its own gas supply line system (2445, 2455 and 2465) ending in a nozzle (2447, 2457 and 2467) near the point of incidence (2422) of the electron beam (2415) for the sample (2425).

[0194] The supply containers (2440, 2450, and 2460) may have their own temperature setting elements and / or control elements that allow both cooling and heating of the corresponding supply containers (2440, 2450, and 2460). This makes it possible to store, in particular, the precursor gas at its respective optimal temperature (not shown in FIG. 24). A control unit (2370) can control the temperature setting elements and temperature control elements of the supply containers (2440, 2450, and 2460). During the EBID and EBIE processing, the temperature setting elements of the supply containers (2440, 2450, and 2460) may be further used to set the vapor pressure of the precursor gas stored inside by selecting an appropriate temperature.

[0195] The device (2400) may include two or more supply containers (2440) to store two or more precursor gases. Additionally, the device (2400) may include two or more supply containers (2450) to store two or more etching gases (not shown in FIG. 24).

Claims

Claim 1 A method (1900) for repairing at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800), comprising the step of identifying at least one repair shape parameter for the at least one defect, wherein the step of identifying the parameter comprises the step of assigning at least one numerical value to the parameter—the numerical value deviates from a numerical value predetermined by the at least one defect for the parameter, and the deviation of the numerical value of the at least one parameter is selected from a range in which the lower limit is greater than 0 and the upper limit is smaller than the resolution limit of the lithographic mask— and the at least one defect (240, 750, 940, A method (1900) comprising the step of recording at least one aerial image of 1140, 1240, 1330, 1630, 1840), wherein the resolution limit of the lithographic mask is determined by the chemical line wavelength of the lithographic mask, the numerical aperture of the projection lens of the exposure system suitable for projecting the pattern of the lithographic mask onto the photoresist, and the exposure setting suitable for projecting the pattern of the lithographic mask onto the photoresist arranged on the wafer, and wherein the lithographic mask comprises one of a binary mask, a phase-shifting mask, or a template for nanoimprint lithography. Claim 2 Method (1900), wherein the at least one parameter comprises at least one lateral dimension of the at least one repair shape, the height dimension of the at least one repair shape, the distance between the at least one repair shape and the at least one defect, the material composition of the at least one defect, the geometric shape of the at least one repair shape, and at least one element from the group of surroundings of the at least one defect on the lithographic mask. Claim 3 A method (1900) in which at least one parameter deviates from a numerical value predetermined by at least one defect for the parameter by a predetermined absolute value. Claim 4 delete Claim 5 In claim 1, the resolution limit of the lithographic mask includes the mask-side resolution limit of an exposure system suitable for exposing a wafer using the lithographic mask, and the mask-side resolution limit Determined by, where λ in the formula is the chemical line wavelength of the lithographic mask, and NA M A method (1900), wherein is the mask-side numerical aperture of the projection lens of the exposure system, and σ is the exposure setting of the exposure system suitable for exposing the lithographic mask. Claim 6 Method (1900) of claim 1, wherein the deviation of the numerical value of the corresponding at least one parameter comprises a range of 2% to 80% of the resolution limit of the lithographic mask. Claim 7 The method (1900) of claim 1, wherein the dimension size of at least one dimension of at least one repair shape comprises a range of 10% to 90% of the dimension size of the corresponding dimension of at least one defect. Claim 8 The method (1900) of claim 1, wherein the at least one defect comprises: an edge placement error (240) of a pattern element (220); at least one of a broken connection of a pattern element and a bridged connection of a pattern element; an outlier of the edge roughness of a pattern element; particles attached to the lithographic mask; a lateral defect having only low printability; a defect residue (340) of a defect repair performed; a sidewall angle error (1850) of a pattern element (1820); and at least one element from at least one group of a distance range between two pattern elements and a center of gravity error of a pattern element. Claim 9 A method (1900) according to claim 1, wherein the step of recording at least one aerial image among the step of verifying the parameters of at least one repair shape comprises at least one of the step of recording at least one aerial image of at least one defect at the chemical line wavelength of the lithographic mask, and the step of recording an aerial image focus stack of at least one defect. Claim 10 The method (1900) of claim 1, wherein the step of assigning the numerical value comprises applying a machine learning model trained to determine the deviation of the numerical value of the at least one parameter from the numerical value predetermined by the at least one defect for the parameter. Claim 11 The method (1900) of claim 1, wherein the step of verifying the parameter of the at least one mathematical shape comprises the step of applying a machine learning model trained to verify the parameter of the at least one mathematical shape. Claim 12 The method of claim 1 further comprises the step of generating at least one repair element (410, 510, 610, 770, 790, 850, 870, 1010, 1170, 1190, 1270, 1290, 1440, 1540, 1640) on the lithographic mask by the identified repair shape, wherein the step of generating the at least one repair element comprises at least one step of performing at least one local etching process and at least one step of performing at least one local deposition process by at least one focused particle beam and at least one precursor gas. Claim 13 A method (1900) according to claim 1, comprising: a. generating at least one repair element (310) by at least one repair shape defined by at least one defect (240) by the parameter; and b. identifying the parameter of the repair shape for the residual defect residue (340) - the step of identifying the parameter of the repair shape for the residual defect residue includes assigning to the parameter at least one numeric value that deviates from a numeric value predetermined by the residual defect residue for the parameter. Claim 14 A method (2000) for repairing at least one defective pattern element (220, 740, 840, 860, 920, 1130, 1230, 1320, 1620, 1820) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800), a. A step of determining at least one repair element (410, 510, 610, 770, 790, 850, 870, 1010, 1170, 1190, 1270, 1290, 1440, 1540, 1640) of the lithographic mask that does not image the lithographic mask during exposure - the at least one repair element is configured to change the imaging behavior of the at least one defect pattern element, and the step of determining the at least one repair element includes the step of recording at least one aerial image of the at least one defect pattern element - ; and b. A method (2000) comprising the step of generating at least one repair element on a lithographic mask using at least one repair shape by means of at least one focused particle beam (2415) and at least one precursor gas, wherein the deviation of the numerical value of at least one parameter of the at least one repair shape is selected from a range in which the lower limit is greater than 0 and the upper limit is less than the resolution limit of the lithographic mask, wherein the lithographic mask comprises one of a binary mask, a phase-shifting mask, or a template for nanoimprint lithography. Claim 15 A method (2000) according to claim 14, wherein the at least one repair element has at least one dimensional size including a range of 10% to 90% of the resolution limit of the lithographic mask. Claim 16 A method (2000) according to claim 15, wherein the distance (420, 520, 620) between the at least one repair element (410, 510, 610) and the at least one defective pattern element (220) comprises a range of 2% to 80% of the resolution limit of the lithographic mask. Claim 17 A computer program stored on a computer-readable storage medium, comprising instructions that, when executed by a computer system, cause the computer system to perform a step of the method described in any one of claims 1 to 3 and claims 5 to 16. Claim 18 An apparatus (2100) for repairing at least one defect (240, 750, 940, 1140, 1240, 1330, 1630, 1840) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800), comprising means for verifying at least one repair shape parameter for said at least one defect, wherein the means for verifying the parameter comprises means for assigning to the at least one parameter a numeric value that deviates from a numeric value predetermined by said at least one defect for said parameter - wherein the deviation of the numeric value of said at least one parameter is selected from a range in which the lower limit is greater than 0 and smaller than the resolution limit of said lithographic mask - ; The apparatus (2100) includes means for recording at least one aerial image of at least one defect, wherein the resolution limit of the lithographic mask is determined by the chemical line wavelength of the lithographic mask, the numerical aperture of the projection lens of the exposure system suitable for projecting the pattern of the lithographic mask onto the photoresist, and the exposure setting suitable for projecting the pattern of the lithographic mask onto the photoresist arranged on the wafer, and wherein the lithographic mask comprises one of a binary mask, a phase-shifting mask, or a template for nanoimprint lithography. Claim 19 An apparatus (2300) for repairing at least one defective pattern element (220, 740, 840, 860, 920, 1130, 1230, 1320, 1620, 1820) of a lithographic mask (200, 400, 500, 600, 700, 800, 900, 1100, 1200, 1300, 1600, 1800), comprising: a. means for recording at least one aerial image of the at least one defective pattern element; b. Means for determining at least one repair element (410, 510, 610, 770, 790, 850, 870, 1010, 1170, 1190, 1270, 1290, 1440, 1540, 1640) of the lithographic mask that is not imaged during exposure using the at least one aerial image - the at least one repair element is configured to change the imaging behavior of the at least one defective pattern element -; and c. An apparatus (2300) comprising means for providing a focused particle beam (2415) configured to generate at least one repair element on a lithographic mask using at least one repair shape, and at least one precursor gas, wherein the deviation of the numerical value of at least one parameter of the at least one repair shape is selected from a range in which the lower limit is greater than 0 and the upper limit is smaller than the resolution limit of the lithographic mask, and the lithographic mask comprises one of a binary mask, a phase-shifting mask, or a template for nanoimprint lithography. Claim 20 A lithographic mask repaired according to the steps of the method described in any one of claims 1 to 3 and claims 5 to 16 used in an exposure system. Claim 21 An exposure system using a lithographic mask repaired according to the steps of the method described in any one of claims 1 to 3 and claims 5 to 16. Claim 22 An exposure system according to claim 21, wherein the resolution limit of the exposure system is determined by the wavelength of the light source of the exposure system, the numerical aperture of the projection lens of the exposure system, and the exposure setting of the exposure system. Claim 23 An exposure system according to claim 22, wherein the resolution limit includes the wafer-side resolution limit of the projection lens. Claim 24 In claim 23, the wafer-side resolution limit is Determined by, where λ in the formula is the chemical line wavelength of the lithographic mask, and NA W is the wafer-side numerical aperture of the projection lens of the exposure system, and σ is an exposure setting of the exposure system suitable for exposing the lithographic mask to project the pattern of the lithographic mask onto the photoresist arranged on the wafer. Claim 25 delete Claim 26 delete Claim 27 delete Claim 28 delete

Images