A system including a photomask for providing EUV light

Abstract

To provide an effective and convenient system capable of performing in-situ EUVTDI sensor linearity correction.SOLUTION: A system includes a photomask including a plurality of distinctly patterned regions capable of responding to irradiation by an extreme ultraviolet (EUV) beam to provide extreme ultraviolet (EUV) light having different intensities. The plurality of distinctly patterned regions includes a plurality of regions including an EUV absorber line and a distinct line space lattice pattern having an alternated EUV reflective multilayered coating; and the distinct line space lattice patterns have an independent absorber duty ratio being a ratio of an EUV absorber linewidth to a lattice pitch.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to image sensors, and more particularly to the calibration of time delay integration (TDI) image sensors in the extreme ultraviolet (EUV). [Background technology]

[0002] [Related Applications] This application claims priority to U.S. Provisional Patent Application No. 60 / 864,313, filed June 20, 2019, the entire contents of which are incorporated herein by reference for all purposes.

[0003] Optical inspection tools for inspecting photomasks (i.e., reticles) use TDI image sensors (abbreviated as TDI sensors). To record photomask inspection images with almost no intensity distortion, the linearity (and nonlinearity) of the TDI sensor must be precisely calibrated down to the pixel level. To calibrate the linearity of a TDI sensor in an optical inspection tool using 193 nm light, various intensity levels of 193 nm light are generated and images are recorded at those various intensity levels. For this calibration process, a polarizer is used to control the light intensity, which can generate various intensity levels, without changing the beam profile. A small, patternless mask area is imaged using the full TDI sensor. A calibrated reference intensity detector is placed downstream of the intensity control. The linearity of the TDI sensor is calibrated by comparing the pixel-by-pixel TDI signal from the TDI sensor with the reference signal from the reference intensity detector. This calibration process relies on a transmission optical system.

[0004] TDI sensor linearity calibration is also required for extreme ultraviolet (EUV) photomask inspection tools (e.g., tools using light at 13.5 nm or other EUV wavelengths). However, all known materials strongly absorb EUV light (to varying degrees), so linearity calibration of EUV TDI sensors cannot be performed successfully using similar transmission methods as those used in 193 nm inspection tools. One possible approach is to control the intensity of the incident EUV light using multiple neutral density (ND) filters with different attenuation levels. While materials for such ND filters can be found (e.g., polysilicon), the thickness of these ND filters would be extremely thin, on the order of tens of nanometers. This thinness makes the use of ND filters for TDI sensor linearity calibration impractical.

[0005] Traditionally, TDI sensor linearity calibration has been performed on a bench setup using a visible-wavelength light source. A homogenizing sphere creates a nearly uniform illumination on the TDI sensor. By varying the light intensity and using a well-calibrated reference detector, the TDI sensor linearity (i.e., TDI response nonlinearity) can be calibrated. However, this traditional approach has several drawbacks. First, it does not account for the wavelength-dependent nonlinearity of the sensor. Second, it is difficult to vary the overall light intensity without perturbing the illumination profile on the TDI sensor. Third, EUV TDI sensor linearity calibration should be performed in situ (i.e., inside the inspection tool with the TDI sensor installed within the inspection tool). In situ calibration is desirable for convenience; the complexity of EUV inspection systems makes it impractical to remove the TDI sensor for bench calibration. In situ calibration is also desirable to reduce calibration costs; bench calibration requires expensive resources, such as a specialized EUV light source, vacuum conditions, and room space. Over its lifetime, a TDI sensor may accumulate several dead pixels (i.e., pixels that become defective and cease to function). These dead pixels create the need for periodic in-situ calibration of the TDI sensor's scan-averaging nonlinearity, which cannot be determined by bench calibration. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] U.S. Patent Application Publication No. 2013 / 0244142 Summary of the Invention [Problem to be solved by the invention]

[0007] Therefore, there is a need for an effective and convenient method and system for performing in situ EUV TDI sensor linearity calibration. This need can be met with a test photomask that allows for the in situ generation of EUV light of various intensities. [Means for solving the problem]

[0008] An image sensor calibration test structure according to certain embodiments includes a photomask with a plurality of distinct patterned regions that, in response to illumination with an extreme ultraviolet (EUV) beam, produce distinct intensities of EUV light.

[0009] In certain embodiments, a calibration method includes loading a photomask with a plurality of distinct patterned areas into a time delay and integration (TDI) inspection tool. The plurality of distinct patterned areas are sequentially illuminated with an EUV light beam. While illuminating each distinct patterned area of ​​the plurality of distinct patterned areas, a TDI sensor in the TDI inspection tool is used to perform separate imaging instances of each distinct patterned area. While performing each imaging instance, a reference intensity of EUV light collected from the photomask is measured using a reference intensity detector. The linearity of the TDI sensor is determined based on the results of the separate imaging instances and the reference intensity of EUV light measured by the reference intensity detector.

[0010] According to certain embodiments, a system includes a TDI inspection tool having an EUV light source and a TDI sensor. The system also includes a photomask loaded into the TDI inspection tool. The photomask has a plurality of distinct patterned areas that produce distinct intensities of EUV light in response to illumination by an EUV beam generated by the EUV light source. The system further includes a reference intensity detector implemented within the TDI inspection tool to measure the intensity of the EUV light collected from the photomask.

[0011] For a better understanding of the various implementations described, reference should be made to the detailed description below in conjunction with the following drawings, which may not be to scale. [Brief explanation of the drawings]

[0012] [Figure 1] 1 illustrates an off-axis image of a photomask being inspected in an EUV photomask inspection tool, size and angle not to scale, in accordance with certain embodiments. [Figure 2] FIG. 2 is a plan view of a line-space grating pattern located within an area on the surface of a test photomask in accordance with certain embodiments. [Figure 3] 1 is a cross-sectional view of a line-space grating pattern located within an area on the surface of a test photomask according to certain embodiments. [Figure 4] 1 illustrates off-axis imaging of a test photomask in accordance with certain embodiments, size and angle not to scale. [Figure 5A] 1A and 1B are cross-sectional views of individual multi-layer coatings that act as EUV reflective areas with a degree of reflectivity on the surface of a test photomask, according to certain embodiments. [Figure 5B] 10A-10C are cross-sectional views of different discrete multi-layer coatings that act as EUV reflective areas with different degrees of reflectivity on the surface of a test photomask, according to certain embodiments. [Figure 6] FIG. 2 is a line plot illustrating the calculated reflectance of multi-layer coatings versus the number of bilayers in those coatings according to certain embodiments. [Figure 7] 1 is a cross-sectional view of a graded multilayer coating on the surface of a test photomask according to certain embodiments. [Figure 8A] 1 is a cross-sectional view of a discrete region on a surface of a test photomask according to certain embodiments, the discrete region having a thickness of a discrete EUV absorber area disposed on a discrete EUV-reflective multilayer coating. [Figure 8B]FIG. 10 is a cross-sectional view of another discrete region on the surface of a test photomask according to certain embodiments, the photomask having a discrete EUV absorber area disposed on a discrete EUV-reflective multilayer coating and having a different thickness. [Figure 9] FIG. 8C is a line plot illustrating the calculated reflectivity of the structure of FIGS. 8A and 8B versus the thickness of its EUV absorber according to certain embodiments. [Figure 10] 1 is a flowchart illustrating a method for calibrating and using an EUV photomask inspection tool according to certain embodiments. [Figure 11] FIG. 1 is a block diagram of a photomask inspection system according to certain embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0013] Like reference characters refer to corresponding parts throughout the drawings and specification.

[0014] Reference will now be made in detail to various embodiments, examples of which are illustrated in the accompanying drawings. In the following detailed description, numerous specific details are set forth to provide a consistent understanding of the various described embodiments. However, those skilled in the art will appreciate that the various described embodiments may practice without these specific details. In other instances, well-known methods, procedures, components, circuits, and networks have been omitted so as not to unnecessarily obscure aspects of the embodiments.

[0015] Extreme ultraviolet (EUV) photomask (i.e., reticle) inspection tools are typically fully reflective. EUV is a common, well-known, and well-understood technical term that refers to light with wavelengths in the 124 nm to 10 nm range. For example, the EUV light used in an EUV photomask inspection tool may be 13.5 nm light. The imaging system in an EUV photomask inspection tool has several EUV reflective mirrors and no transmissive optics. The illumination path (i.e., the optical path that delivers EUV light to the inspected photomask) and the imaging path (i.e., the optical path that collects light from the inspected photomask) are spatially separated. Such inspection tools therefore have an off-axis imaging design.

[0016] FIG. 1 illustrates off-axis imaging of a photomask 100 being inspected in an EUV photomask inspection tool (e.g., inspection tool 1130 in FIG. 11 ) in accordance with certain embodiments. The photomask 100 has a patterned mask surface 102 that is to be inspected for defects. An illumination EUV light cone 104 produced by an EUV light source (not shown) (e.g., light source 402 in FIG. 4 ) is focused onto the patterned mask surface 102 to illuminate a small-sized area (e.g., 200 μm × 200 μm). An imaging EUV light cone 108 emanating from the illuminated patterned mask surface 102 is collected and imaged by a TDI sensor 116. The illumination EUV light cone 104 is spatially separated from the imaging EUV light cone 108. The illumination EUV light cone 104 makes a chief ray angle (CRA) 106 with respect to a surface normal 112 (i.e., normal direction) of the patterned mask surface 102. The imaging EUV light cone 108 forms a CRA 110 with respect to the surface normal 112. The numerical aperture (NA) of the imaging path determines the size of the imaging EUV light cone 108. On the imaging path (i.e., the path of the imaging EUV light cone 108) are EUV imaging optics 114, which focus the imaging EUV light cone 108 onto the TDI sensor 116, thereby illuminating all TDI pixels. The EUV imaging optics 114 include an EUV mirror. The illumination path (i.e., the path of the illumination EUV light cone 104) may also include EUV optics with an EUV mirror that are used to direct the illumination EUV light cone 104 to the patterned mask surface 102. For simplicity, these mirrors are not shown in FIG. 1 .

[0017] EUV photomask inspection tools typically perform time delay integration (TDI) and therefore have a TDI image sensor (abbreviated as TDI sensor). TDI sensors must be calibrated to ensure accurate inspection results (e.g., images of inspected photomasks). Calibrating a TDI sensor involves determining its linearity. Linearity refers to how accurately the TDI sensor measures incident light intensity at various intensity levels. Quantifying the inaccuracy indicates the degree of corresponding nonlinearity in the TDI sensor. Therefore, determining the linearity of a TDI sensor implies quantifying any nonlinearity associated with the TDI sensor. The TDI sensor linearity (or nonlinearity) determined through calibration can be stored and used to correct subsequent inspection results (e.g., images of photomasks subsequently inspected by the inspection tool).

[0018] Thus, calibration of a TDI sensor involves generating different intensity levels of EUV light incident on the TDI sensor. Such calibration can be performed in situ in an EUV photomask inspection tool by using an EUV test photomask (i.e., calibration photomask) with multiple distinct patterned regions on its surface and providing the TDI sensor with distinct intensities of EUV light in response to illumination of the distinct regions by an EUV beam (e.g., illumination EUV light cone 104). The distinct patterned regions generate (e.g., reflect) different intensities of EUV light when illuminated with the same EUV beam (i.e., when the same EUV beam profile is incident on each patterned region).

[0019] In certain embodiments, the separate patterned regions include regions with individual line-space grid patterns (i.e., gratings) of alternating EUV absorber lines and EUV reflective areas. The line-space grid patterns of the separate regions include different absorber duty ratios. In certain embodiments, the separate patterned regions include EUV reflective areas that exhibit varying degrees of reflectivity. In certain embodiments, the separate patterned regions comprise EUV reflective areas that have a gradient in thickness and therefore varying reflectivity. In certain embodiments, the separate patterned regions include EUV absorber areas of varying thicknesses, with each EUV absorber area above an EUV reflective area.

[0020] [Line-space lattice pattern with alternating EUV absorber lines and EUV reflective areas] FIG. 2 illustrates a plan view of a line-space grating pattern 200 located within a region on the surface of a test photomask according to certain embodiments. The line-space grating pattern 200 includes alternating EUV-reflective multilayer coatings 202 (i.e., mirror-like coatings) and EUV absorber lines 204. The period of the EUV absorber lines 204 (i.e., the distance from the beginning of one EUV absorber line 204 to the beginning of the next EUV absorber line 204) is the grating pitch 206. The ratio of the width of the EUV absorber lines 204 (i.e., the EUV absorber line width) to the pitch is referred to as the absorber duty ratio (this width is along the x-direction in FIG. 2). In certain embodiments, the EUV absorber lines 204 are oriented during calibration to be perpendicular to the plane formed by the illumination and imaging CRAs 106 and 110 (FIG. 1).

[0021] 3 illustrates a cross-sectional view of a line-space grating pattern 300 within a region on the surface of a test photomask according to certain embodiments. The line-space grating pattern 300 is an example of the line-space grating pattern 200 (FIG. 2) and includes alternating EUV-reflective multilayer coatings 316 and EUV-absorber lines 318 at a grating pitch 206. The EUV-reflective multilayer coating 316 and the EUV-absorber lines 318 are examples of the EUV-reflective multilayer coating 202 and the EUV-absorber lines 204 (FIG. 2), respectively.

[0022] The EUV-reflective multilayer coating 316 (which is essentially a single multilayer coating divided into various EUV-reflective areas by EUV absorber lines 318) resides above a substrate (e.g., a blank photomask) 302 and includes alternating molybdenum (Mo) layers 304 and silicon (Si) layers 306 capped by a capping layer 308. The capping layer 308 may be ruthenium (Ru) or boron (B). Each pair of adjacent Mo and Si layers 304 and 306 is referred to as a MoSi bilayer 314. In certain embodiments, the Mo layer 304 is 2.8 nm thick, the Si layer 306 is 4.2 nm thick, and the capping layer 308 is 2.5 nm thick (thicknesses are along the z-direction in FIG. 3 ). The EUV-reflective multilayer coating 316 is partially, rather than completely, reflective to EUV light (the reflectivity of the multilayer coating 316 is discussed below in connection with FIG. 6). The number of MoSi bilayers 314 is selected to provide a relatively high degree of reflectivity (e.g., at least 60% or at least 65% reflectivity). In certain embodiments, the number of MoSi bilayers 314 in the line-space grating pattern 300 is 40, or 40-45, or 35-40, or 50.

[0023] An EUV absorber line 318 is disposed above the multilayer coating 316. The EUV absorber line 318 includes a tantalum boron nitride (TaBN) layer 310 above the capping layer 308 and a tantalum boron oxide (TaBO) capping layer 312 above the TaBN layer 310. In certain embodiments, the TaBO capping layer 312 is 2 nm thick. TaBN is a highly EUV absorbing material. The thickness of the TaBN layer 310 is variable and is selected to absorb substantially all incident EUV light (e.g., in accordance with FIG. 9, discussed below). In certain embodiments, the thickness of the TaBN layer 310 is 70-80 nm.

[0024] Various regions on the surface of the test photomask have individual line-space grating patterns 200 (FIG. 2) (e.g., line-space grating pattern 300 in FIG. 3) with individual absorber duty ratios (i.e., individual absorber duty ratios). For example, the individual line-space grating patterns 200 (FIG. 2) (e.g., line-space grating pattern 300 in FIG. 3) on the surface of the test photomask have EUV absorber lines 318 with individual widths, but exhibit the same grating pitch 206 (the width is along the x-direction in FIGS. 2 and 3). That is, the width of the EUV absorber lines 318 can vary between regions and be constant within each region, while the pitch 206 can be constant throughout the regions.

[0025] FIG. 4 illustrates off-axis imaging of a test photomask 400 having a line-space grating pattern 200 (FIG. 2) (e.g., line-space grating pattern 300 in FIG. 3) on its surface 402, according to certain embodiments. The grating pattern 200 lies in an xy plane perpendicular to the plane of the paper in FIG. 4 and is not visible in FIG. 4. A small area (e.g., 200 μm × 200 μm) of the grating pattern 200 is illuminated with an illumination EUV light cone 104 (shown in FIG. 4 as a single arrow for simplicity) generated by an EUV light source 402. The grating pattern 200 generates multiple diffraction order beams, including a zeroth order beam 406 (m=0) (i.e., a simple reflection of the illumination EUV light cone 104 from the grating surface), a positive first order diffraction beam 408 (m=+1), and a negative first order diffraction beam 410 (m=−1). An opaque aperture 412 is implemented on the imaging path to block the first-order diffracted beams 408 and 410 and select the zeroth-order beam 406. The zeroth-order beam 406 is focused by the EUV imaging optics 114 onto the TDI sensor 116 and imaged by the TDI sensor 116. In certain embodiments, the zeroth-order beam 406 illuminates all of the TDI pixels in the TDI sensor 116 and is used to image the image. A calibrated reference intensity detector 416 is implemented on the imaging path (e.g., at or near the edge of the zeroth-order beam 406) to detect the intensity of the zeroth-order beam 406 (i.e., the intensity of the portion of the zeroth-order beam 406 profile that is incident on the reference intensity detector 416). In certain embodiments, the reference intensity detector 416 is located at the edge of the aperture 412. For example, a reference intensity detector 416 can be mounted to the aperture 412 (eg, extending from the edge of the aperture 412 into the zero order beam 406).

[0026] In embodiments in which the aperture 412 is detachable from the inspection tool, the aperture 412 can be installed within the inspection tool to perform calibration, and after calibration is complete, the aperture 412 can be removed from the inspection tool and used to inspect production photomasks in sequence. In embodiments in which the aperture 412 is permanently installed within the inspection tool, the aperture 412 can be movable within the inspection tool, allowing it to be moved into the imaging path during calibration and out of the imaging path for inspection of production photomasks. In embodiments in which the reference intensity detector 416 is detachable from the inspection tool, the reference intensity detector 416 can be installed within the inspection tool to perform calibration, and after calibration is complete, the reference intensity detector 416 can be removed from the inspection tool and used to inspect production photomasks in sequence. A single detachable reference intensity detector 416 can also be used to calibrate multiple inspection tools. In certain embodiments, the reference intensity detector 416 is certified or calibrated by an official standards body (eg, the National Institute of Standards and Technology (NIST) or a similar government standards organization).

[0027] Different line-space grating patterns 200 (FIG. 2) (e.g., line-space grating pattern 300 in FIG. 3) with different absorber duty ratios can be placed in different regions of the test photomask 400. To calibrate the TDI sensor 116, these different regions can be illuminated and then imaged using off-axis imaging as shown in FIG. 4. The imaging results obtained using the TDI sensor 116 are compared with the intensity detected by the reference intensity detector 416. This comparison determines the linearity of the TDI sensor 116.

[0028] The TDI sensor 116 can be operated in two modes: frame imaging mode and scan mode. In frame imaging mode, all pixels of the TDI sensor 116 simultaneously capture their light intensity, a so-called frame, over a short period of time (e.g., 0.001 ms to several ms), without pixel-by-pixel integration. In scan mode, pixel-by-pixel light intensity is integrated along the scan direction (e.g., the x-direction). Scan mode is commonly used in production photomask inspection. During TDI linearity calibration, frame imaging mode may be used to calibrate the response linearity of pixels (e.g., each pixel). The TDI response linearity in scan mode can be calculated based on the pixel-level linearity measured in frame imaging mode.

[0029] When line-space grating patterns 200 with various absorber duty ratios are illuminated, in embodiments where the pitch 206 is the same, the diffraction angles of the first-order diffracted beams 408 and 410 are the same for each pattern (although the angles of the +1st-order diffracted beam 408 and the −1st-order diffracted beam 410 will generally be different). However, due to differences in the width of the EUV absorber lines 204 (e.g., differences in the area of ​​the multilayer coating 316 covered by the EUV absorber lines 318), the intensity of the zeroth-order diffracted beam 406 will differ. The beam profile of the zeroth-order diffracted beam 406 remains identical to the illumination beam because it is essentially a simple reflection of the incident beam (i.e., the illumination EUV light cone 104). Therefore, by selecting a particular line-space grating pattern 200, the intensity of EUV light delivered to the TDI sensor 116 can be controlled without changing the incident beam profile for TDI linearity calibration.

[0030] In order for the opaque aperture 412 to select a clean zeroth-order diffracted beam 406, the illumination parameter σ should be small enough so that there is also little overlap between the zeroth-order diffracted beam 406 and the two first-order diffracted beams 408 and 410. Table 1 shows the calculation results for the maximum value of σ for various grating pitches 206. In these calculations, the imaging NA is assumed to be 0.2, and the CRA for both the illumination and imaging paths is 8.15°.

[0031] [Table 1]

[0032] The three columns in the middle of Table 1 show the CRAs of the three diffraction orders calculated using the grating equation assuming an illumination CRA of 8.15°. The maximum σ is calculated by taking the smaller of the two half angles between the CRAs for m = 0 and m = ±1 and dividing it by the solid angle half angle associated with the NA. For example, with an 80 nm pitch, the CRA difference between the m = 0 and m = +1 diffraction waves is 9.7°, and the CRA difference between the m = 0 and m = -1 diffraction waves is 9.94°. The smaller CRA difference, 9.7°, is selected. If the half angle of the illumination EUV light cone 104 is 9.7° / 2 = 4.85°, there will be no overlap between the three diffraction orders. This condition corresponds to illumination σ=4.85° / 11.5°=0.42, where 11.5° is the half solid angle associated with NA=0.2. If σ is greater than 0.42, a smaller aperture is required to select the pure zeroth order diffraction beam 406.

[0033] In Table 2, the zeroth-order effective reflectance of the line-space grating pattern 300 (FIG. 3) is calculated for various absorber duty ratios and line-space grating pitches 206. The effective reflectance is normalized relative to the reflectance of the pure multilayer coating 316 (i.e., the absence of any EUV absorber lines 318 in FIG. 3, resulting in an absorber duty ratio of 0) at a wavelength of 13.5 nm and an angle of incidence of 8.15°. The calculations assume that the EUV light is unpolarized and use an absorber thickness (i.e., the thickness of the TaBN layer 310) of 70 nm. For each value of pitch 206 listed in the first column, the zeroth-order normalized effective reflectance is calculated for absorber duty ratios of 0.8, 0.65, 0.5, 0.35, 0.2, and 0. The absorber duty cycle has two extreme cases: a pure absorber (e.g., the multilayer coating 316 in FIG. 3 is completely covered with EUV absorber) and a pure EUV-reflective multilayer coating (e.g., the multilayer coating 316 in FIG. 3 is completely exposed with no EUV absorber), which respectively provide minimum and maximum light intensity values. Since intensity corresponds to the reflectance for a given intensity of incident light (i.e., a given illumination EUV light cone 104), the various reflectances shown in Table 2 clearly indicate the ability to generate EUV at different intensities. During this calibration, focus differences caused by different absorber duty cycles should be corrected during imaging by the TDI sensor 116.

[0034] [Table 2]

[0035] The narrower the line width of the EUV absorber lines 204 (FIG. 2) (e.g., EUV absorber lines 318 in FIG. 3), the more difficult it is to create the line-space grating pattern 200 (FIG. 2) (e.g., line-space grating pattern 300 in FIG. 3). Using a larger grating pitch 206 facilitates creation, but reduces the maximum achievable illumination σ, as shown in Table 1. Empirically, the narrowest absorber lines that can be created on an EUV photomask are one-third the absorber thickness. For an absorber thickness (i.e., absorber line height) (e.g., the thickness of the TaBN layer 310) of 70 nm, the narrowest achievable absorber lines are 23.3 nm wide, which corresponds to a duty ratio of 0.29 for an 80 nm pitch 206. When the line-space grating pattern has a small pitch, the absorber thickness (absorber line height) can be reduced to optimize the zero-order effective reflectance and facilitate the creation of the grating pattern.

[0036] [EUV reflective areas with different degrees of reflectivity] 5A and 5B illustrate cross-sectional views of discrete multilayer coatings 500-1 and 500-2, which, according to certain embodiments, serve as EUV-reflective areas with different degrees of reflectivity (i.e., different reflectivities) in discrete regions on the surface of a test photomask. Multilayer coatings 500-1 and 500-2 can have the same structure and / or layer thickness as multilayer coating 316 (FIG. 3). Multilayer coating 500-2 (FIG. 5B) has more Mo layers 304 and Si layers 306, and therefore more MoSi bilayers 314, than multilayer coating 500-1 (FIG. 5A). Other regions on the surface of the test photomask can have multilayer coatings 500 with different (i.e., different) numbers of Mo layers 304 and Si layers 306, and therefore different numbers of MoSi bilayers 314. Generally, the EUV reflective multi-layer coating in different areas on the surface of the test photomask will have different numbers of layers.

[0037] The EUV reflectivity of the multilayer coatings 500-1 and 500-2 is a function of the number of MoSi bilayers 314. FIG. 6 is a plot 600 illustrating the calculated reflectivity 604 of the multilayer coating 500 versus the number of bilayers 314 602. The calculations behind this plot 600 assumed unpolarized 13.5 nm light at an angle of incidence of 8.15° and the layer thicknesses described with respect to FIG. 3. As shown in FIG. 6, a theoretical reflectivity of over 70% can be achieved. In practice, with a sufficient number 602 of bilayers 314 (e.g., more than 40 bilayers 314), a reflectivity of at least 65% can be achieved. Table 3 summarizes the calculated reflectivity for several bilayers, normalized to the reflectivity of the multilayer coating 500 with 40 bilayers.

[0038] [Table 3]

[0039] A test photomask having a multilayer coating 500 with different numbers of layers (i.e., different numbers of bilayers 314) in distinct regions on its surface can be used in the off-axis imaging arrangement of Figure 4 to calibrate the TDI sensor 116. Because there is no grating and therefore no diffraction of the illuminating EUV light cone 104, the aperture 412 can be omitted from the off-axis imaging arrangement. By selecting different regions on the surface of the test photomask for illumination and imaging, the intensity of the EUV light delivered to the TDI sensor 116 can be varied in a controlled manner for a given illuminating EUV light cone 104 during calibration. Focus offsets caused by differences in the number of bilayers should be corrected during imaging by the TDI sensor 116.

[0040] [EUV reflective area with gradient thickness] A variation on using a multilayer coating with a varying number of bilayers (e.g., multilayer coatings 500-1 and 500-2 in FIGS. 5A and 5B) is to create a graded multilayer coating on the surface of a test photomask. This gradeing, while keeping the interlayer thickness ratio of each bilayer the same (i.e., constant), results in differences in the (average) bilayer thicknesses across different regions of the surface of the test photomask. Different bilayer thicknesses in a graded multilayer coating correspond to different peak reflectance wavelengths. Therefore, for the same wavelength and angle of incidence, different bilayer thicknesses result in different reflectivities. When different areas of a graded multilayer coating are imaged with the TDI sensor 116 (e.g., using the off-axis configuration of FIG. 4 without aperture 412), differences in the intensity of EUV light reflected by the coating (e.g., under a given illumination EUV light cone 104) will be apparent.

[0041] FIG. 7 illustrates a cross-sectional view of a graded multilayer coating 700 on the surface of a test photomask according to certain embodiments. The multilayer coating 700 can be similar in structure to the multilayer coating 316 (FIG. 3) and include a specific number of bilayers 314 (e.g., 40 bilayers 314). In certain embodiments, the thickness of the MoSi bilayers 314 (e.g., both the alternating Mo layers 304 and Si layers 306) varies along a first direction (e.g., the x-direction) and is uniform along a second direction perpendicular to the first direction (e.g., the y-direction perpendicular to the plane of the page in FIG. 7). The thickness ratio between the Si layers 306 and the Mo layers 304 remains constant throughout the multilayer coating 700 (e.g., with the same thickness ratio as that of the multilayer coatings 500-1 and 500-2 in FIGS. 5A and 5B).

[0042] The bilayer thickness variations in the graded multilayer coating 700 are small enough that the thickness can be treated as a constant within the field of view (FOV) (e.g., 200 μm×200 μm) of the TDI sensor 116. That is, for a given area on the surface of the test photomask, the reflectivity (i.e., intensity scaling factor) is substantially the same for every pixel of the TDI sensor 116. Focus offsets caused by thickness variations in the graded multilayer coating 700 should be corrected during imaging by the TDI sensor 116.

[0043] [EUV absorber area with different thickness above the EUV reflective area] Another in situ EUV light intensity control technique is to use pure absorber areas of various absorber thicknesses in distinct regions on the surface of a test photomask. These absorber areas are positioned above an EUV-reflective multilayer coating. Figures 8A and 8B are cross-sectional views of distinct regions 800-1 and 800-2 on the surface of a test photomask, according to certain embodiments. Regions 800-1 and 800-2 have distinct EUV absorber areas and distinct EUV-reflective multilayer coatings, with the EUV absorber area of ​​each region 800 positioned above the EUV-reflective multilayer coating area of ​​that region 800. In certain embodiments, the EUV-reflective multilayer coating of regions 800-1 and 800-2 has the same structure and / or layer thickness as multilayer coating 316 (Figure 3), i.e., a set of MoSi bilayers 314 disposed on substrate 302 and covered by capping layer 308. The Mo layer 304 may be 2.8 nm thick, the Si layer 306 may be 4.2 nm thick, and the capping layer 308 may be 2.5 nm thick (thicknesses are along the z-direction in FIGS. 8A and 8B). Each region 800 (e.g., both regions 800-1 and 800-2) may have the same number of MoSi bilayers 314 (e.g., 40 bilayers 314, or 40-45 bilayers 314, or 35-40 bilayers 314). The EUV absorber area may have the same structure as the EUV absorber line 318 (FIG. 3), i.e., a TaBN layer 310 covered by a TaBO capping layer 312 (e.g., 2 nm thick) positioned above the capping layer 308. Different regions 800 have different thicknesses of the TaBN layer 310. For example, the TaBN layer 310 in region 800-2 (FIG. 8B) is thicker than the TaBN layer 310 in region 800-1 (FIG. 8A). The difference in thickness of the TaBN layer 310 results in a difference in the reflectivity of the entire structure.

[0044] Figure 9 is a line plot 900 illustrating the calculated reflectivity 904 of the structure of Figures 8A and 8B versus the thickness 902 of the TaBN layer 310, according to certain embodiments. The calculations behind Figure 9 assume unpolarized 13.5 nm light with an angle of incidence of 8.15°, 40 MoSi bilayers 314, and a fixed thickness of the TaBO capping layer 312. As depicted in Figure 9, the reflectivity curve exhibits an etalon effect with a period of approximately 7.2 nm (≈λ / 2), where the reflectivity decreases with increasing thickness 902, then increases somewhat, and then decreases again. Table 4 summarizes the normalized reflectivity (normalized relative to the reflectivity of an EUV-reflective multilayer coating without the TaBN layer 310) versus the thickness 902 of the TaBN layer 310.

[0045] [Table 4]

[0046] A test photomask having region 800 (e.g., including regions 800-1 and 800-2) with distinct EUV absorber areas of different thicknesses can be used to calibrate TDI sensor 116 in the off-axis imaging arrangement of FIG. 4. Because there is no grating and therefore no diffraction of the illuminating EUV light cone 104, aperture 412 can be omitted from the off-axis imaging arrangement. By selecting different regions for illumination and imaging, the intensity of EUV light delivered to TDI sensor 116 can be varied in a controlled manner for calibration. Focus offsets caused by differences in absorber thickness should be corrected when imaging region 800 with TDI sensor 116.

[0047] [Method Flowchart] FIG. 10 is a flowchart illustrating a method 1000 for calibrating and using an EUV photomask inspection tool (e.g., inspection tool 1130 of FIG. 11 ) according to certain embodiments. In method 1000, a photomask having a plurality of distinct patterned regions (e.g., test photomask 400 of FIG. 4 ) is loaded into a TDI inspection tool (1002). In certain embodiments, the photomask has a plurality of regions exhibiting distinct line-space grating patterns (e.g., line-space grating pattern 200 of FIG. 2 or line-space grating pattern 300 of FIG. 3 ) of alternating EUV absorber lines and EUV-reflective multilayer coatings (1004). The distinct line-space grating patterns have distinct absorber duty ratios (i.e., ratios of EUV absorber line width to grating pitch). For example, the distinct line-space grating patterns can have distinct EUV absorber line widths but the same grating pitch. In certain embodiments, the photomask includes a plurality of regions (1006) having individual EUV-reflective multilayer coatings (e.g., multilayer coatings 500-1 and 500-2 in FIGS. 5A and 5B). The individual EUV-reflective multilayer coatings include a different number of layers (e.g., a different number of MoSi bilayers 314 in FIGS. 5A and 5B). In certain embodiments, the plurality of distinct patterned regions form a graded EUV-reflective multilayer coating (e.g., graded multilayer coating 700 in FIG. 7) (1008). In certain embodiments, the photomask includes a plurality of regions (1010) having individual EUV-absorber areas and individual EUV-reflective multilayer coatings (e.g., regions 800-1 and 800-2 in FIGS. 8A and 8B). The individual EUV-absorber areas are positioned above the individual EUV-reflective multilayer coatings and have different thicknesses.

[0048] The plurality of distinct patterned areas are sequentially illuminated 1012 with an EUV light beam (e.g., illumination EUV light cone 104 of FIG. 4 ) (e.g., 13.5 nm light, e.g., unpolarized), thereby illuminating each distinct patterned area with the EUV beam. While each distinct patterned area of ​​the plurality of distinct patterned areas is illuminated, a separate imaging instance of the distinct patterned area is performed 1012 using a TDI sensor (e.g., TDI sensor 116 of FIG. 4 ) within the TDI inspection tool. The separate imaging instances may be performed in a frame imaging mode. While the separate imaging instances are performed, a reference intensity of EUV light collected from the photomask is measured 1012 using a reference intensity detector (e.g., reference intensity detector 416 of FIG. 4 ).

[0049] In certain embodiments (e.g., where the photomask has multiple regions with a discrete line-space grating pattern of alternating EUV absorber lines and EUV-reflective multilayer coatings (1004)), an aperture (e.g., aperture 412 in FIG. 4) is positioned to select the zeroth order diffracted beam (e.g., zeroth order beam 406 in FIG. 4) as the light to be collected from the photomask.

[0050] The linearity of the TDI sensor is determined based on the results of the individual imaging instances and the reference intensity measurements of EUV light (1014). In certain embodiments, the per-pixel linearity of the TDI sensor is determined based on the results of the individual imaging instances and the reference intensity measurements (1016). The per-pixel linearity can be determined by performing a pixel-by-pixel comparison of the signal from the TDI sensor with the signal from the reference intensity detector. In certain embodiments, the TDI integrated intensity linearity (e.g., scan-averaged intensity linearity) of the TDI sensor is determined based on the per-pixel linearity by integrating the per-pixel calibration results (e.g., integrating the calibration results for the entire two-dimensional pixel array of the TDI sensor along the TDI scan direction) (1018).

[0051] In certain embodiments, the EUV beam is pulsed 1012 as it illuminates the plurality of distinct patterned areas. As the image moves with the movement of the photomask during inspection, this pulsing can have the effect of selecting a subset of pixels within the TDI sensor. The linearity of this pixel subset can be determined in step 1014.

[0052] At this point in method 1000, the TDI sensor is calibrated and the TDI inspection tool is ready for use. A production photomask (e.g., a reticle that has been fabricated but not yet used to manufacture a semiconductor device) is inspected (1020) using the TDI inspection tool. This inspection can be performed to check the production photomask for defects. The inspection results of the production photomask are corrected (1022) based on the results of the TDI sensor linearity determination. For example, the image generated in inspection step 1020 is corrected based on the results of the TDI sensor linearity determination. After the TDI sensor is calibrated, steps 1020 and 1022 can be repeated to inspect multiple production photomasks. The calibration process of steps 1002-1014 can be periodically repeated to ensure accurate operation of the TDI inspection tool.

[0053] [System block diagram] 11 is a block diagram of a photomask inspection system 1100 (i.e., a reticle inspection system) according to certain embodiments. The photomask inspection system 1100 includes an EUV photomask inspection tool 1130 and a computer system having one or more processors 1102 (e.g., CPUs), a user interface 1106, memory 1110, and communication bus(es) 1104 interconnecting these components. Alternatively, the computer system may be communicatively coupled to the EUV photomask inspection tool 1130 via one or more networks. The computer system may also include one or more network interfaces (wired and / or wireless, not shown) for communication with the EUV photomask inspection tool 1130 and / or remote computer systems.

[0054] According to certain embodiments, photomask inspection system 1100 can be configured to perform the off-axis imaging of Figure 4. Aperture 412 can be implemented in inspection tool 1130 (e.g., permanently or temporarily for calibration). Reference intensity detector 416 can be implemented in inspection tool 1130 (e.g., temporarily for calibration). A test photomask can be loaded into inspection tool 1130 to perform calibration. A production photomask can be loaded into the calibrated inspection tool 1130 for inspection.

[0055] User interface 1106 may include a display 1107 and one or more input devices 1108 (e.g., a keyboard, a mouse, a touch-sensitive surface of display 1107, etc.) that may display the results of method 1000 (FIG. 10).

[0056] The memory 1110 may include volatile and / or nonvolatile memory. The memory 1110 (e.g., nonvolatile memory within the memory 1110) may include a non-transitory computer-readable storage medium. The memory 1110 may optionally include one or more storage devices located remotely from the processor 1102 and / or non-transitory computer-readable storage media inserted or removed from the system 1100. In certain embodiments, the memory 1110 (e.g., a non-transitory computer-readable storage medium of the memory 1110) stores the following modules and data, or a subset or superset thereof: an operating system 1112 that includes procedures for handling various basic system services and performing hardware-dependent tasks; a calibration module 1114 for calibrating an inspection tool 1130 using a patterned test photomask (e.g., patterned according to Figures 2, 3, 5A, 5B, 7, 8A, and / or 8B) (e.g., test photomask 400 of Figure 4); a TDI inspection module 1116 for inspecting the photomask; a result correction module 1118 for correcting inspection results (e.g., images) based on the TDI sensor linearity determined through calibration; and a notification module 1120 for notifying calibration and / or inspection results. The memory 1110 (eg, a non-transitory computer-readable storage medium of the memory 1110) is thus embodied with instructions for performing all or part of the method 1000 (FIG. 10).

[0057] Each module stored in memory 1110 corresponds to a set of instructions for performing one or more functions described herein. Separate modules need not be implemented as separate software programs. Modules and various subsets of those modules may be combined or reconfigured. In certain embodiments, memory 1110 stores a subset or superset of the modules and / or data structures described above.

[0058] 11 is intended as a functional description of various features that may appear within a photomask inspection system, rather than a structural overview. For example, the functionality of the computer system within photomask inspection system 1100 may be distributed among multiple devices. Portions of the modules stored within memory 1110 may instead be stored within one or more other computer systems that are communicatively coupled to the computer system of photomask inspection system 1100 via one or more networks.

[0059] The foregoing description is for purposes of illustration and has been described with reference to specific embodiments. However, the illustrative discussion above is not intended to limit or exclude the scope of the claims to the precise forms disclosed. Many modifications and variations are possible in light of the teachings above. The embodiments have been selected to best explain the principles underlying the claims and their practical application, and to enable one skilled in the art to best utilize the embodiments, with various modifications commensurate with the specific use envisioned.

Claims

1. 1. A system comprising: Extreme ultraviolet (EUV) test photomask including a plurality of distinct patterned regions that provide different intensities of extreme ultraviolet (EUV) light in response to illumination by an EUV beam wherein the plurality of distinct patterned regions comprise a gradient EUV reflective multi-layer coating.

2. the plurality of distinct patterned regions comprising a plurality of regions having a discrete line-space grating pattern of alternating EUV absorber lines and EUV reflective multilayer coatings; the individual line-space grating patterns have different absorber duty ratios, wherein the absorber duty ratios are a ratio of the EUV absorber line width to a grating pitch; The system of claim 1 .

3. the individual line-space grating patterns have individual, distinct EUV absorber line widths but the same grating pitch; The system of claim 2 .

4. the EUV absorber line contains tantalum boron nitride (TaBN); the EUV-reflective multilayer coating comprises alternating molybdenum (Mo) and silicon (Si) layers; The system of claim 2 .

5. the plurality of distinct patterned regions comprising a plurality of regions having individual EUV reflective multilayer coatings; the individual EUV-reflective multi-layer coatings have individual and distinct numbers of layers; The system of claim 1 .

6. the individual EUV reflective multilayer coatings comprise individual and discrete numbers of alternating Mo and Si layers; The system of claim 5.

7. the gradient EUV-reflective multilayer coating comprises a plurality of alternating Mo and Si layers; the alternating Mo and Si layers have a graded thickness along a first direction and a uniform thickness along a second direction perpendicular to the first direction; the alternating Mo and Si layers have a constant thickness ratio; The system of claim 1 .

8. the plurality of distinct patterned regions comprising a plurality of regions each comprising a distinct EUV absorber area and a distinct EUV reflective multilayer coating, the distinct EUV absorber area being above the distinct EUV reflective multilayer coating; the individual EUV absorber areas have individual, distinct thicknesses; The system of claim 1 .

9. the individual EUV absorber areas comprise TaBN and a TaBO capping layer above the TaBN; the individual EUV-reflective multilayer coatings comprise the same number of alternating Mo and Si layers; The system of claim 8.

10. an extreme ultraviolet (EUV) light source that generates EUV light; a time delay integration (TDI) sensor; and wherein the photomask is mounted to the TDI inspection tool; a reference intensity detector implemented within the TDI inspection tool to measure the intensity of EUV light collected from the photomask; The system of claim 1 further comprising:

Images