Overlay measurement method and system based on soft x-ray scatterometry

Abstract

Methods and systems for performing overlay and edge placement error based on measurement data from soft x-ray (SXR) scatterometry measurements are presented herein. Short wavelength SXR radiation focused on small illumination spot sizes enables measurement of design rule targets and functional device structures in dies. In some embodiments, measurements of SXR scatterometry measurements are performed with SXR radiation having energies in the range from 10 to 5,000 electron volts. Thus, measurements at SXR wavelengths permit target designs at process design rules that closely represent actual device overlay. In some embodiments, measurements of SXR scatterometry measurements of overlay and shape parameters are performed simultaneously from the same metrology target to enable accurate measurement of edge placement error. In another aspect, based on SXR measurements of design rule targets, overlay of non-periodic device structures is estimated by calibrating the SXR measurements to reference measurements of actual device targets.

Description

[0001] Cross Reference to Related Applications

[0002] This patent application claims priority under 35 U.S.C. § 119 from U.S. Provisional Patent Application No. 62 / 958,089, filed January 7, 2020, entitled “System and Method for Measuring Overlay and Edge Placement Error With Soft X-ray Scatterometry,” the subject matter of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The described embodiments relate to metrology systems and methods, and more particularly to methods and systems for improved measurement accuracy. BACKGROUND

[0004] Semiconductor devices such as logic and memory devices are typically fabricated by a sequence of processing steps applied to a sample. Various features and multiple structural levels of the semiconductor devices are formed by these processing steps. For example, lithography, among other processing steps, is a semiconductor fabrication process that involves generating patterns on a semiconductor wafer. Additional examples of semiconductor fabrication processes include, but are not limited to, chemical-mechanical polishing, etch, deposition, and ion implantation. Multiple semiconductor devices can be fabricated on a single semiconductor wafer and then separated into individual semiconductor devices.

[0005] During the semiconductor manufacturing process, metrology processes are used at various steps to detect defects on wafers to facilitate higher yields. Several metrology-based techniques, including scatterometry and reflectometry implementations, and associated analysis algorithms are commonly used to characterize critical dimensions, film thickness, composition, and other parameters of nanoscale structures.

[0006] As devices (e.g., logic and memory devices) progress toward smaller nanometer scale dimensions, characterization becomes more difficult. Devices incorporating complex three-dimensional geometries and materials with disparate physical properties exacerbate the characterization difficulty. For example, modern memory structures are often high aspect ratio three-dimensional structures, which make it difficult for optical radiation to penetrate to the bottom layers. Optical metrology tools utilizing infrared light to visible light can penetrate many layers of a semi-transparent material, but the longer wavelengths that provide good penetration depth do not provide sufficient sensitivity to small anomalies. Similarly, electron-based metrology tools suffer from insufficient penetration depth without damaging the sample. Additionally, the increasing number of parameters required to characterize complex structures (e.g., FinFETs) results in increasing parameter correlation. Thus, the parameters of the characterization target are often not reliably decoupled from the available measurements. For some structure parameters, such as edge placement error (EPE), there are currently no high throughput (e.g., optical) measurement solutions.

[0007] Currently, several techniques are employed to measure overlay and critical dimensions (CDs) with varying degrees of success. CD and overlay measurements are typically performed on specialized metrology targets using optical and electron beam metrology techniques.

[0008] Optical overlay measurements are primarily based on optical imaging or non-imaging diffraction (scatterometry). However, these methods have not reliably overcome the fundamental challenges associated with measurement of many advanced targets (e.g., complex 3D structures, structures smaller than 10 nm, structures employing non- transparent materials) and measurement applications (e.g., line edge roughness and line width roughness measurements).

[0009] Using existing methods, overlay errors are typically evaluated based on measurements of specialized target structures formed on a wafer at various locations by a lithography tool. In some examples, spatially separated gratings are used for imaging-based optical overlay measurements. In some other examples, frame-in-frame structures are used for imaging-based optical overlay measurements. In this form, a frame is formed on one layer of the wafer and a second, smaller frame is formed on another layer. Localized overlay errors are measured by comparing the alignment between the centers of the two frames. Such measurements are made at locations on the wafer where the target structures are available. In some examples, gratings or interleaved gratings of overlay are used for scatterometry-based optical overlay measurements or electron beam overlay measurements.

[0010] Unfortunately, these specialized target structures often do not conform to the design rules of the particular semiconductor manufacturing process used to produce the electronic devices. This results in an estimation error for overlay errors associated with actual device structures manufactured according to the applicable design rules.

[0011] In one example, image-based optical overlay metrology is severely limited by the resolution of imaging at optical wavelengths. Thus, only targets with features much larger than the design rule can be measured. Image-based optical overlay metrology typically requires resolving patterns with critical dimensions much larger than the design rule critical dimension with a light microscope.

[0012] In another example, zeroth-order diffraction based scatterometry based optical overlay metrology has very low sensitivity to small overlay errors because the sensitivity decreases with the pitch of the periodic target. This drives the pitch to be much larger than the design rule of the device. Also, the accuracy of this measurement method is dramatically degraded in the presence of any asymmetry in either of the layers in which the overlay is measured. In addition, this method cannot distinguish between positive and negative overlay errors in a single measurement.

[0013] In another example, higher than zeroth-order diffraction based scatterometry based optical overlay metrology also requires relatively large pitch targets to generate sufficient signal at the non-zero propagating diffraction orders. Overlay asymmetry optical signals are typically generated from periodic targets with a pattern pitch comparable to the optical wavelength of the illuminating light. In some examples, pitch values in the range 500 nm to 800 nm can be used. Meanwhile, the actual device pitch (design rule size) for logic or memory applications is much smaller, e.g., in the range 100 nm to 400 nm or even below 100 nm. In addition, the accuracy of this method is dramatically degraded in the presence of any asymmetry in either of the layers in which the overlay is measured. Reducing the optical wavelength of the illumination into the deep ultraviolet and vacuum ultraviolet ranges does not help because these photons are attenuated and insufficient to penetrate multiple layer structures to reach the underlying patterns needed to assess overlay and edge placement errors.

[0014] Electron-based metrology techniques, such as scanning electron microscopy (SEM) and e-beam metrology, are capable of resolving nanoscale features and measuring non-periodic structures, such as random logic. However, electron-based metrology systems are destructive when used to measure actual devices. Additionally, electron-based metrology systems have low throughput. Measurement times can be on the order of a few seconds per measurement site. Additionally, electron-based metrology systems are top-down imaging systems that provide very limited three-dimensional measurement capability. For example, when employing SEM to measure overlay between overlapping gratings, it loses the ability to measure CD and EPE as the point spread function increases with penetration depth. Further details are described in U.S. Patent No. 10,473,460 to Gutman et al. and assigned to KLA-Tencor Corp., the contents of which are incorporated herein by reference in their entirety. Generally, SEM implementations achieve an intermediate level of resolution, but are not able to penetrate structures to sufficient depth without destroying the sample. Additionally, the required charging of the sample has adverse effects on imaging performance.

[0015] Atomic force microscopy (AFM) and scanning tunneling microscopy (STM) are capable of atomic resolution, but they can only probe the surface of a sample. Additionally, AFM and STM microscopes require long scanning times.

[0016] Transmission electron microscopy (TEM) achieves a high level of resolution and is capable of probing arbitrary depths, but TEM requires destructive sectioning of the sample.

[0017] To overcome some of the limitations of optical and electron-based metrology for overlay and EPE measurement, absolute positioning measurement techniques can be employed. Absolute distance between features is measured using an accurate stage of translation with a range equal to the spatial separation between any two patterns. This result is used to assist in the evaluation of overlay, CD, and EPE. Unfortunately, absolute positioning measurement techniques require accurate stages, adding complexity to the measurement tool and limiting throughput. Also, the techniques can not perform well on patterned targets or device structures that cover the pattern. Further details are described in WIPO Publication No. 2019 / 173171 to Shchegrov et al. and assigned to KLA-Tencor Corp., the contents of which are incorporated herein by reference in their entirety.

[0018] In summary, for logic devices and advanced DRAM as well as vertical or planar NAND devices, semiconductor device yield at device fabrication nodes below 20 nanometers is a complex function of many parameters including film thickness, profile parameters of patterned lines, overlay error, and edge placement error (EPE). Of these, EPE has the highest required process window and requires metrology and control of CD and overlay. Currently, there does not exist a high-throughput optical metrology solution for EPE measurement and many design rule overlay measurement applications. Additionally, the lack of adequate metrology makes it challenging to define control schemes to improve device yield.

[0019] Future metrology applications present challenges to metrology due to smaller resolution requirements, multi-parameter correlation, more complex geometries, and increased use of non- transparent materials. Accordingly, methods and systems for improved overlay and shape measurement are desired. SUMMARY

[0020] Methods and systems are presented herein for performing overlay and edge placement error based on measurement data from soft x-ray (SXR) scatterometry measurements. Short wavelength SXR radiation focused on small illumination spot sizes enables measurement of design rule targets, i.e., targets having the same or approximately the same pitch as nearby in-die structures or in-die structures themselves. In addition to providing overlay metrology capability, the methods and systems described herein also enhance the precision and accuracy of shape parameter measurements by strongly de-correlating the geometric parameters of the measured structures.

[0021] SXR illumination radiation enables penetration into non-transparent regions of a target and into underlying layers. In some embodiments, the measurements of SXR scatterometry measurements are performed with SXR radiation having energies in the range from 10 to 5,000 electron volts. In general, diffraction limits and other optical effects control the minimum possible target size for shape and overlay measurements. Due to the relatively short wavelength of SXR illumination, measurements of SXR scatterometry measurements can be performed on metrology targets having relatively small target regions.

[0022] SXR penetration into underlying layers enables measurements of SXR scatterometry measurements to have relatively high sensitivity to the signals required to estimate overlay, CD, and EPE. Also, overlay measurements of SXR scatterometry measurements of design rule targets more closely represent actual device overlay than traditional overlay targets having much larger pitch. SXR scatterometry enables overlay measurements on design rule targets because the illumination wavelength is shorter than the period of the measured design rule targets. Accordingly, measurements at SXR wavelengths permit target designs at process design rules.

[0023] In some examples, overlay measurements by SXR scatterometry are based on direct measurements of actual device structures, e.g., SRAM.

[0024] In some embodiments, the design rule targets include multiple layers each having an underlying periodicity. SXR scatterometry enables measurement of multi-layer design rule targets with high sensitivity to the underlying pattern. In some of these embodiments, the top layer of the design rule target is a photoresist layer. In this way, SXR scatterometry enables after- development inspection (ADI) process monitoring.

[0025] In one aspect, the SXR scatterometry system is configured to estimate overlay error between different layers of a design rule metrology target or functional device structures in a die from non-zero diffraction orders scattered from the measured structure. Due to the relatively short wavelength of SXR radiation, non-zero diffraction orders and in particular + / - 1 diffraction orders provide relatively high sensitivity to overlay error.

[0026] In some embodiments, overlay and shape parameter measurements by SXR scatterometry are performed simultaneously from measurements of SXR scatterometry collected from the same metrology target. This enables measurement of edge placement error (EPE), e.g., end-of-line shortening, line-to-contact distance, etc. Thus, SXR scatterometry enables edge placement error (EPE) measurement without error due to target biasing that occurs when overlay and CD measurements are performed on different targets. In addition, simultaneous measurement of overlay and CD structure parameters from the same metrology target improves both measurement accuracy and throughput.

[0027] In some embodiments, SXR scatterometry measurements from two cells of metrology targets each having a nominal offset in opposite directions are employed to resolve overlay error.

[0028] In some examples, metrology based on SXR scatterometry involves determining parameters of interest characterizing the sample, e.g., overlay error, shape parameters, etc., by solving an inverse of a predetermined measurement model with measured SXR scatterometry data. In this way, target parameters are estimated by solving for values of a parameterized measurement model that minimize error between measured scattered x-ray intensity and modeled results.

[0029] In some embodiments, overlay error values associated with a measurement target or targets in a die are determined directly from detected intensity within one or more non-zero diffraction orders based on a trained machine learning based measurement model. In these embodiments, a trained machine learning based model extracts overlay error directly from SXR measurement data.

[0030] In some embodiments, overlay measurement based on SXR scatterometry involves illuminating a sample with SXR radiation and detecting the intensity of resulting diffraction orders for a plurality of angles of incidence, a plurality of wavelengths, or both, relative to the sample. Also, overlay error associated with the measurement target is determined based on modulation of a plurality of intensities within each of one or more non-zero diffraction orders at each of a plurality of measurement instances.

[0031] In another aspect, the actual device targets are aperiodic. Based on measuring design rule targets with sufficient periodicity, by calibrating overlay measurements to reference measurements, SXR scatterometry techniques can be used to estimate overlay for aperiodic structures. This effectively overcomes the limitation of scatterometry measurements that require the measured target to be periodic or approximately periodic.

[0032] In some embodiments, a multi-layer overlay metrology target is designed with different pitches at different layers such that diffraction orders generated from one layer constructively interfere with different diffraction orders of another layer. Conversely, intensity measurements detected at different order pairs that are not subject to constructive interference of overlay are dominated by shape parameters. Thus, in some embodiments, metrology overlay targets are designed with specific grating structures to increase sensitivity to overlay at specific grating order pairs and also provide intensity data useful for shape parameter value estimation.

[0033] In some embodiments, a multi-layer overlay metrology target is designed with different pitch orientations at different layers such that diffraction orders generated from one layer constructively interfere with different diffraction orders of another layer. In general, a set of layers with different periodicity (e.g., different grating pitches), different pitch orientations, or any combination thereof, generates a set of scattering vectors, each associated with a different layer. The overlay metrology target is designed such that a predetermined subset of the set of scattering vectors is aligned. In this way, sensitivity to overlay among the layers corresponding to the predetermined subset of scattering vectors is enhanced.

[0034] The foregoing is a summary and thus necessarily contains substantial simplification, generalization and omission of detail; consequently, those skilled in the art will appreciate that the summary is illustrative only and is not limiting in any way. Other aspects, inventive features, and advantages of the devices and / or processes described herein will become apparent in the non-limiting detailed description and drawings set forth herein. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 is a diagram illustrating a hard mask pattern of a line structure 11 fabricated in a static random access memory (SRAM) region 10 of a microelectronic chip.

[0036] Figure 2 is a diagram illustrating a bottom anti-reflective coating (BARC) layer 12 and a hard mask layer 13 disposed on the BARC layer 12.Figure 1 Schematic diagram of a resist layer 13 on top of the pattern of line structures depicted in FIG.

[0037] Figure 3 is a diagram illustrating an embodiment of a soft X-ray (SXR) metrology tool 100 for measuring properties of a sample according to exemplary methods presented herein.

[0038] Figure 4 is a diagram illustrating a top view of a device structure 400 including active fields 401 - 404 , gates 405 - 408 , and contacts 409 - 421 .

[0039] Figure 5A Depicted are cells having stacked grating structures offset from each other by a distance equal to the sum of the nominal offset + OFFSET and the overlay error OVL in one direction.

[0040] Figure 5B Depicts cells having stacked grating structures offset from each other by a distance equal to the distance between Figure 5A The sum of the nominal offset -OFFSET in the direction opposite to the nominal offset depicted and the overlay error OVL.

[0041] Figure 6 is a diagram illustrating a model building and analysis engine 180 configured to estimate overlay based on model fitting analysis as described herein.

[0042] Figure 7 is a diagram illustrating a wafer 101 including a metrology target 120 illuminated by a beam of x-ray radiation at incident and azimuthal angles.

[0043] Figure 8 A layered metrology target 150 comprising two periodic arrays of different layer centerlines 151 and 152 is depicted.

[0044] Figure 9 Describe the diagram Figure 8 Graph 170 of a simulation of the intensities of the -2 and +2 diffraction orders for the angular range of the metrology target illustrated in FIG, and a simplified model fit to the corresponding measured values.

[0045] Figure 10 is a diagram illustrating a semiconductor die including active device structures fabricated in an active device area and design rule metrology targets fabricated in a scribe line area.

[0046] Figure 11 is a flow chart illustrating an exemplary method 200 of estimating overlay based on calibrated SXR measurements as described herein. DETAILED DESCRIPTION

[0047] Reference will now be made in detail to background examples and some embodiments of the application, examples of which are illustrated in the accompanying drawings.

[0048] Figure 1 A hard mask pattern depicting line structures 11 fabricated in a static random access memory (SRAM) region 10 of a microelectronic wafer. A complex layout of the active area is formed by combining multiple patterning techniques with a cut mask. The cut mask selectively removes portions of the hard mask layer used to pattern the substrate into the active area. Figure 2 A bottom anti-reflective coating (BARC) layer 12 and a resist layer 13 disposed on top of the pattern of line structures depicted in Figure 1 The resist layer 13 is used to selectively remove a portion of the hard mask pattern under an opening 14 of the resist layer 13. As depicted in Figure 1 The hard mask pattern of line structures 11 is even buried by the BARC layer 12 within the opening 14 of the resist layer 13, as depicted in

[0049] To provide sufficient yield for the cut mask process, reliable measurements of shape parameters (e.g., CD, HT, SWA, profile parameters, etc.), film thickness, and overlay are required. Overlay calculations reveal that they are dependent on many structure parameters from previous steps of the quadruple patterning process. The distribution of the gap between the edge of the cut and the adjacent line structure and thus the yield of the process depend on a complex interaction of all process parameters.

[0050] In another example, edge placement distance (EPD) and associated edge placement error (EPE) are important parameters that are monitored and controlled after making device electrical contacts. The difference between the desired EPD and the actual EPD is referred to as EPE. EPD and EPE are functions of both overlay and CD error.

[0051] Methods and systems are presented for performing overlay and edge placement error of structures and materials based on measurement data of soft X-ray (SXR) scatterometry. Short wavelength SXR radiation focused on small illumination spot sizes enables measurement of design rule targets, i.e., targets having the same or approximately the same pitch as nearby functional device structures in the wafer or the functional device structures themselves. The methods and systems presented herein are applicable to two- and three-dimensional design rule metrology targets located within or outside of a functional wafer. In addition to providing overlay metrology capability, the methods and systems described herein also enhance the precision and accuracy of shape parameter measurements by strongly de-correlating the geometry parameters of the measured structures.

[0052] SXR illumination radiation enables penetration into non- transparent regions of a target and into underlying layers. Examples of measurable geometric parameters using SXR scatterometry include hole size, hole density, line edge roughness, line width roughness, sidewall angle, profile, critical dimension, overlay, edge placement error, and pitch. Examples of measurable material parameters include electron density, element identification, and composition. In some examples, SXR scatterometry enables measurement of features smaller than 10 nm and advanced semiconductor structures where measurement of geometric and material parameters is required (e.g., spin transfer torque MRAM).

[0053] In some embodiments, measurements of SXR scatterometry measurements are performed with SXR radiation having energies in the range from 10 to 5,000 electron volts. In general, diffraction-limited and other optical effects control the smallest possible target size for shape and overlay measurements. Due to the relatively short wavelength of SXR illumination, measurements of SXR scatterometry measurements can be performed on metrology targets having relatively small target areas. In some embodiments, measurements of SXR scatterometry measurements are performed on metrology targets within areas having maximum extent dimensions greater than 5 microns. In some embodiments, measurements of SXR scatterometry measurements are performed on metrology targets within areas having maximum extent dimensions greater than 2 microns.

[0054] In some embodiments, design rule targets include multiple layers each having an underlying periodicity. SXR scatterometry enables measurement of multi-layer design rule targets with high sensitivity to underlying patterns. In some of these embodiments, the top layer of a design rule target is a photoresist layer. In this way, SXR scatterometry enables after- development inspection (ADI) process monitoring.

[0055] SXR penetration into underlying layers enables measurements of SXR scatterometry measurements to have relatively high sensitivity to the signals required to estimate overlay, CD, and EPE. Also, overlay measurements of SXR scatterometry measurements of design rule targets more closely represent actual device overlay than traditional overlay targets having much larger pitch. In some examples, overlay measurements of SXR scatterometry measurements are based on direct measurements of actual device structures (e.g., SRAM).

[0056] In one aspect, a SXR scatterometry system is configured to estimate overlay error between different layers of a design rule metrology target or an active device structure in a die from non-zero diffraction orders scattered from the measured structure.

[0057] Figure 3 An embodiment of a SXR scatterometry tool 100 for measuring properties of a sample in at least one novel aspect is illustrated. As shown in Figure 3 The system 100 can be used to perform measurements of SXR scatterometry measurements within a measurement area 102 of a sample 101 illuminated by an incident illumination beam spot, as shown in

[0058] In the depicted embodiment, metrology tool 100 includes an x-ray illumination source 110, focusing optics 111, a beam divergence control slit 112, and a slit 113. X-ray illumination source 110 is configured to generate SXR radiation suitable for SXR scatterometry measurements. In some embodiments, x-ray illumination source 110 is a polychromatic, high-brightness, large-etendue source. In some embodiments, x-ray illumination source 110 is configured to generate x-ray radiation in a range between 10 to 5000 electron volts. In general, any suitable high-brightness x-ray illumination source capable of generating high-brightness SXR at flux levels sufficient to enable high-throughput inline metrology is contemplated for application of x-ray illumination for SXR measurements.

[0059] In some embodiments, the x-ray source includes a tunable monochromator that enables the x-ray source to deliver x-ray radiation at different selectable wavelengths. In some embodiments, one or more x-ray sources are employed to ensure that the x-ray source supplies light at wavelengths that allow sufficient penetration into the sample under test.

[0060] In some embodiments, illumination source 110 is a high-harmonic generation (HHG) x-ray source. In some other embodiments, illumination source 110 is an undulator / synchrotron radiation source (SRS). Exemplary undulator / synchrotron SRSs are described in U.S. Patent Nos. 8,941,336 and 8,749,179, the contents of which are incorporated herein by reference in their entirety.

[0061] In some other embodiments, illumination source 110 is a laser-produced plasma (LPP) light source. In some of these embodiments, the LPP light source includes any of xenon, krypton, argon, neon, and nitrogen emission materials. In general, the selection of suitable LPP target materials is optimized for brightness in the resonant SXR region. For example, plasmas emitted by krypton provide high brightness at the silicon K-edge. In another example, plasmas emitted by xenon provide high brightness across the entire SXR region (10 to 5000 eV).

[0062] The LPP target material selection can also be optimized for reliable and long-life light source operation. Inert gas target materials such as xenon, krypton, and argon are inert and can be reused in closed-loop operation with only minimal or no decontamination processing required. Exemplary SXR illumination sources are described in Khodykin et al. and assigned to U.S. Patent Publication No. 2019 / 0215940 to KLA-Tencor Corp., the contents of which are incorporated herein by reference in their entirety.

[0063] In some embodiments, the wavelengths emitted by the illumination source (e.g., illumination source 110) are selectable. In some embodiments, illumination source 110 is an LPP light source that is controlled by computing system 130 to maximize the flux in one or more selected spectral regions. The laser peak intensity at the target material controls the plasma temperature and thus the spectral region of the emitted radiation. The laser peak intensity is varied by adjusting the pulse energy, the pulse width, or both. In one example, a 100 picosecond pulse width is suitable for generating SXR radiation. As Figure 3 As depicted in FIG. 1, computing system 130 communicates command signals 136 to illumination source 110, which causes illumination source 110 to adjust the spectral range of wavelengths emitted from illumination source 110. In one example, illumination source 110 is an LPP light source, and the LPP light source adjusts any of the pulse duration, the pulse frequency, and the target material composition to achieve the desired spectral range of wavelengths emitted from the LPP light source.

[0064] By way of non-limiting example, any of a particle accelerator source, a liquid anode source, a rotating anode source, a stationary solid anode source, a microfocus source, a microfocus rotating anode source, a plasma-based source, and an inverse Compton source can be used as x-ray illumination source 110.

[0065] Exemplary x-ray sources include electron beam sources configured to bombard a solid or liquid target to simulate x-ray radiation. Methods and systems for generating high-brightness liquid metal x-ray illumination are described in U.S. Patent No. 7,929,667, issued to KLA-Tencor Corporation on April 19, 2011, which is incorporated herein by reference in its entirety.

[0066] X-ray illumination source 110 generates x-ray emissions within a source region having a finite lateral dimension (i.e., a non-zero dimension orthogonal to the beam axis). In one embodiment, the source region of illumination source 110 is characterized by a lateral dimension of less than 20 microns. In some embodiments, the source region is characterized by a lateral dimension of 10 microns or less. The small source size enables illumination of small target regions on a sample with high brightness, thus improving measurement precision, accuracy, and throughput.

[0067] Generally, the x-ray optics shape and direct the x-ray radiation to the sample 101. In some examples, the x-ray optics use multilayer x-ray optics to collimate or focus the x-ray beam onto the measurement area 102 of the sample 101 to achieve a divergence of less than 1 milliradian. In some embodiments, the x-ray optics include one or more x-ray collimating mirrors, x-ray apertures, x-ray beam stop, refractive x-ray optics, diffractive optics (e.g., zone plates), Schwarzschild optics, Kirkpatrick-Baez optics, Montel optics, Wolter optics, mirror x-ray optics (e.g., ellipsoidal mirrors), multi-capillary optics (e.g., hollow capillary x-ray waveguides), multilayer optics or systems, or any combination thereof. Further details are described in U.S. Patent Publication No. 2015 / 0110249, the contents of which are incorporated herein by reference in its entirety.

[0068] As Figure 3 Focusing optics 111 focus source radiation onto metrology targets located on sample 101, as depicted in FIG. 1. The finite lateral source size results in a finite spot size 102 on the target defined by rays 116 from the edges of the source and any beam shaping provided by beam slits 112 and 113.

[0069] In some embodiments, focusing optics 111 include focusing optical elements shaped as ellipses. In the embodiment depicted in FIG. 1, focusing optics 111 include focusing optical elements shaped as ellipses with a major axis of 1000 nm and a minor axis of 500 nm. Figure 3 In the embodiment depicted in FIG. 1, the magnification of focusing optics 111 at the center of the ellipse is approximately 1. Thus, the illumination spot size projected onto the surface of sample 101 is approximately the same as the size of the illumination source, adjusted for beam expansion due to the nominal angle of incidence G.

[0070] In yet another aspect, focusing optics 111 collect source emission and select one or more discrete wavelengths or spectral bands, and focus the selected light onto sample 101 at a desired nominal angle of incidence.

[0071] The nominal angle of incidence is selected to achieve a desired penetration of the metrology target to maximize signal information content while remaining within the metrology target boundary. The critical angle for hard x-rays is minimal, but the critical angle for soft x-rays is significantly larger. As a result of this additional measurement flexibility, SXR measurements probe deeper into structures that have less sensitivity to the precise value of the nominal angle of incidence.

[0072] In some embodiments, the focusing optics 111 include a graded multilayer that is selected for a desired wavelength or range of wavelengths to be projected onto the sample 101. In some examples, the focusing optics 111 include a graded multilayer structure (e.g., layers or coatings) that is selected for one wavelength and projects the selected wavelength onto the sample 101 over a range of angles of incidence around a nominal angle of incidence. In some examples, the focusing optics 111 include a graded multilayer structure that is selected for a range of wavelengths and projects the selected wavelengths onto the sample 101 at one angle of incidence. In some examples, the focusing optics 111 include a graded multilayer structure that is selected for a range of wavelengths and projects the selected wavelengths onto the sample 101 over a range of angles of incidence.

[0073] Graded multilayer optics are preferred to minimize light loss that occurs when single layer grating structures are too deep. In general, multilayer optics are selected for reflected wavelengths. The spectral bandwidth of the selected wavelengths is optimized to provide flux to the sample 101, information content in the measured diffraction orders, and to prevent degradation of the signal via angular dispersion at the detector and overlap of diffraction peaks. Additionally, graded multilayer optics are employed to control divergence. Angular divergence at each wavelength is optimized for flux and minimal spatial overlap at the detector.

[0074] In some examples, the graded multilayer optics select wavelengths to enhance contrast and information content of the diffraction signal from a particular material interface or structure size. For example, the selected wavelengths can be chosen to span a resonant region specific to the element (e.g., silicon K-edge, nitrogen K-edge, oxygen K-edge, etc.). Additionally, in these examples, the illumination source can also be tuned to maximize flux in the selected spectral region (e.g., HHG spectral tuning, LPP laser tuning, etc.).

[0075] In some embodiments, the focusing optics 111 include a plurality of reflective optical elements each having an elliptical surface shape. Each reflective optical element includes a substrate and a multilayer coating tuned to reflect a different wavelength or range of wavelengths. In some embodiments, a plurality (e.g., 1 to 5) of reflective optical elements each reflecting a different wavelength or range of wavelengths are arranged at each angle of incidence. In yet another embodiment, a plurality (e.g., 2 to 5) of groups of reflective optical elements each reflecting a different wavelength or range of wavelengths are arranged at different nominal angles of incidence. In some embodiments, the plurality of groups of reflective optical elements simultaneously project illumination light onto the sample 101 during a measurement. In some other embodiments, the plurality of groups of reflective optical elements sequentially project illumination light onto the sample 101 during a measurement. In these embodiments, an active shutter or aperture is employed to control the illumination light projected onto the sample 101.

[0076] In some embodiments, the wavelength range, AOI, azimuthal angle, or any combination thereof projected onto the same metrology area is adjusted by actively positioning one or more mirror elements of the focusing optics. As Figure 3 As depicted in FIG. 1, the computing system 130 communicates a command signal 137 to the actuator system 115, which causes the actuator system 115 to adjust the position, alignment, or both of one or more of the optical elements of the focusing optics 111 to achieve the desired wavelength range, AOI, azimuthal angle, or any combination thereof projected onto the sample 101.

[0077] Generally, the angle of incidence is selected for each wavelength to optimize the penetration and absorption of the illumination light by the measured metrology target. In many examples, multiple layer structures are measured and the angle of incidence is selected to maximize the signal information associated with the desired layer of interest. In examples of overlay metrology, the wavelength and angle of incidence are selected to maximize the signal information resulting from interference between scattering from a previous layer and the current layer. Additionally, the azimuthal angle is also selected to optimize the signal information content. Additionally, the azimuthal angle is selected to ensure angular separation of the diffraction peaks at the detector.

[0078] In some embodiments, the SXR scatterometry system (e.g., metrology tool 100) includes one or more beam slits or apertures to shape the illumination beam 114 incident on the sample 101 and selectively block a portion of the illumination light that would otherwise illuminate the measured metrology target. The one or more beam slits define the beam size and shape such that the x-ray illumination spot fits within the area of the measured metrology target. Additionally, the one or more beam slits define the illumination beam divergence to minimize the overlap of diffraction orders on the detector.

[0079] In some embodiments, the SXR scatterometry system (e.g., metrology tool 100) includes one or more beam slits or apertures to select a set of illumination wavelengths that simultaneously illuminate the measured metrology target. In these embodiments, the one or more slits are configured to pass illumination including multiple illumination wavelengths. Generally, simultaneous illumination of the measured metrology target is preferred to increase signal information and throughput. In practice, however, the overlap of diffraction orders at the detector limits the range of illumination wavelengths. In some embodiments, the one or more slits are configured to pass different illumination wavelengths sequentially. In some examples, sequential illumination at larger angular divergence provides higher throughput because the signal-to-noise ratio of sequential illumination can be higher when the beam divergence is larger compared to simultaneous illumination. When measurements are performed sequentially, the overlap of diffraction orders is not an issue. This increases measurement flexibility and improves signal-to-noise ratio.

[0080] Figure 3A beam divergence control slit 112 is depicted in the beam path between the focusing optics 111 and the beam shaping slit 113. The beam divergence control slit 112 limits the divergence of the illumination provided to the measured sample. A beam shaping slit 113 is located in the beam path between the beam divergence control slit 112 and the sample 101. The beam shaping slit 113 further shapes the incident beam 114 and selects the illumination wavelengths of the incident beam 114. The beam shaping slit 113 is located in the beam path immediately before the sample 101. In some embodiments, the slits of the beam shaping slit 113 are located immediately proximate to the sample 101 to minimize magnification of the incident beam spot size due to the beam divergence defined by the finite source size. As Figure 3 depicted in FIG. 1, the computing system 130 communicates a command signal 138 to the beam divergence control slit 112, which causes the active elements of the beam divergence control slit 112 to adjust the position, alignment, or both of one or more of the optical elements of the beam divergence control slit 112 to achieve the desired beam divergence. Similarly, as Figure 3 depicted in FIG. 1, the computing system 130 communicates a command signal 139 to the beam shaping slit 113, which causes the active elements of the beam shaping slit 113 to adjust the position, alignment, or both of one or more of the optical elements of the beam shaping slit 113 to achieve the desired beam shape projected onto the sample 101.

[0081] In some embodiments, the beam shaping slit 113 includes multiple independently actuated beam shaping slits. In one embodiment, the beam shaping slit 113 includes four independently actuated beam shaping slits. The four beam shaping slits effectively block a portion of the incoming beam and produce an illumination beam 114 with an illumination cross-section having a frame shape.

[0082] The slits of the beam shaping slit 113 are constructed from a material that minimizes scattering and effectively blocks the incident radiation. Exemplary materials include single crystal materials such as germanium, gallium arsenide, indium phosphide, etc. Typically, the slit material is cleaved rather than sawed along the crystal direction to minimize scattering across the structure boundary. Additionally, the slits are oriented with respect to the incoming beam such that interactions between the incoming radiation and the internal structure of the slit material produce a minimal amount of scattering. A crystal is attached to each slit holder made of a high density material (e.g., tungsten) in order to completely block the x-ray beam on one side of the slit.

[0083] In some embodiments, the focusing optics of the SXR scatterometry system project an image of the illumination source onto the sample under test at a demagnification of at least five (i.e., a magnification factor of 0.2 or less). In some embodiments, the SXR scatterometry system as described herein employs an SXR illumination source having a source area characterized by a lateral dimension of 20 microns or less (i.e., a source size of 20 microns or less). In some embodiments, the focusing optics are employed to project the illumination onto a sample having an incident illumination spot size of four microns or less at a demagnification of at least five (i.e., projecting an image of the source onto a wafer that is five times smaller than the source size).

[0084] In some embodiments, the illumination source 110 is an LPP light source having a source size of 10 microns or less, and the focusing optics 111 have a demagnification factor of approximately 10. This enables the SXR scatterometry tool to focus the illumination light onto metrology targets having a size of 1 to 2 microns. The ability to measure targets having a size of 1 to 2 microns reduces the wafer area devoted to specialized metrology targets. In addition, the ability to measure targets having a size of 1 to 2 microns enables direct measurement of device structures rather than specialized metrology targets. Measuring device structures directly eliminates the bias of the targets to the device. This significantly improves the measurement quality. In addition, measurement of targets in a die enables characterization of intra-die parameter variations. Exemplary parameters of interest include critical dimensions, overlay, and edge placement error.

[0085] The x-ray detector 119 collects x-ray radiation 118 scattered from the sample 101 according to the SXR scatterometry measurement and generates an output signal 135 indicative of a property of the sample 101 sensitive to incident x-ray radiation. In some embodiments, the scattered x-rays 118 are collected by the x-ray detector 119 while the sample positioning system 140 positions and orients the sample 101 to produce angularly resolved scattered x-rays.

[0086] In some embodiments, the SXR scatterometry system includes one or more photon counting detectors having a high dynamic range (e.g., greater than 10 5 ) In some embodiments, a single photon counting detector detects the location and number of detected photons.

[0087] In some embodiments, the x-ray detector resolves one or more x-ray photon energies and generates a signal indicative of a property of the sample for each x-ray energy component. In some embodiments, the x-ray detector 119 includes any of a CCD array, a microchannel plate, a photodiode array, a microstrip proportional counter, a gas-filled proportional counter, a scintillator, or a fluorescent material.

[0088] In this manner, X-ray photon interactions within the detector are distinguished by energy in addition to pixel location and number of counts. In some embodiments, the X-ray photon interactions are distinguished by comparing the energy of the X-ray photon interactions to a predetermined upper threshold and a predetermined lower threshold. In one embodiment, this information is communicated to the computing system 130 via the output signal 135 for further processing and storage.

[0089] Due to angular dispersion in diffraction, the diffraction pattern resulting from simultaneous illumination of a periodic target with multiple illumination wavelengths is separated at the detector plane. In these embodiments, an integrating detector is employed. An area detector (e.g., a vacuum compatible backside CCD or hybrid pixel array detector) is used to measure the diffraction pattern. The angular sampling is optimized for Bragg peak integration. If a pixel level model fit is employed, the angular sampling is optimized for signal information content. The sampling rate is chosen to prevent zero order signal saturation.

[0090] In some instances, it can be desirable to perform measurements over a large wavelength range, a range of incident angles, and a range of azimuthal angles to increase the precision and accuracy of the measured parameter values. This approach reduces the correlation among parameters by extending the number and diversity of data sets available for analysis.

[0091] The intensity of diffracted radiation is measured as a function of illumination wavelength and x-ray incident angle relative to the wafer surface normal. The information contained in the multiple diffraction orders is typically unique among each model parameter under consideration. Thus, x-ray scattering yields an estimate of the values of the parameters of interest with small error and reduced parameter correlation.

[0092] In some embodiments, the metrology tool 100 includes a wafer chuck 103 fixedly supporting the wafer 101 and coupled to a sample positioning system 140. The sample positioning system 140 is configured to actively position the sample 101 with six degrees of freedom relative to the illumination beam 114. In one example, the computing system 130 communicates command signals (not shown) to the sample positioning system 140 indicative of a desired position of the sample 101. In response, the sample positioning system 140 generates command signals to various actuators of the sample positioning system 140 to achieve the desired positioning of the sample 101.

[0093] In yet another aspect, an SXR scatterometry system is employed to determine properties (e.g., structural parameter values) of a sample based on one or more diffraction orders of scattered light. As Figure 8 As depicted in FIG. 1, the metrology tool 100 includes a computing system 130 employed to acquire the signals 135 generated by the detector 119 and determine properties of the sample based at least in part on the acquired signals.

[0094] SXR scatterometry enables overlay measurement on design rule targets because the illumination wavelength is shorter than the period of the structures being measured. This provides a significant benefit over prior art where overlay is measured on structures larger than the design rule targets. The use of SXR wavelengths permits target design to process design rules (i.e., no "non-zero offset"). Reduction of non-zero offset is a result of reduced pattern placement error and process optimized for patterning at device pitch. Pattern placement error is dependent on the manufacturing process driven by pattern pitch. As pattern pitch is reduced, pattern placement error is also reduced. Also, as the manufacturing process is optimized to device pitch, local asymmetry distortion of device characteristic patterns (e.g., pattern similar to device) is reduced.

[0095] Overlay metrology targets for SXR measurement can include one-dimensional periodic arrays or two-dimensional periodic arrays. One-dimensional targets exhibit large angular divergence along the plane of incidence, increasing flux and throughput. For two-dimensional targets, the diffraction angular dispersion is not equivalent for the two in-plane axes. Thus, for sample directions parallel to the plane of incidence, an additional super-period can be imposed. In these instances, it can be advantageous to rotate the wafer and perform sequential orthogonal measurements on the same target by a single sub-system.

[0096] In one aspect, SXR scatterometry signals are employed to resolve overlay error based on non-zero diffraction orders. Due to the relatively short wavelength of SXR radiation, non-zero diffraction orders and in particular + / - 1 diffraction orders provide relatively high sensitivity to overlay error.

[0097] In general, target design and associated measurement algorithms differ depending on whether 0thor 1storder scatterometry is employed. For 0thorder scatterometry, each measured region (i.e., cell) of a metrology target produces a single 0thorder signal as a function of wavelength, angle of incidence, and azimuthal angle. However, for 1storder scatterometry, each measured region (i.e., cell) of a metrology target produces two signals, a +1storder signal and a -1storder signal, as a function of wavelength, angle of incidence, and azimuthal angle. Thus, in some instances, fewer cells are needed to extract sufficient signal information to resolve parameters of interest, such as overlay or shape parameters, based on 1storder scatterometry. In these instances, smaller metrology targets can be employed. For efficiency, 1storder scatterometry requires relatively high pupil uniformity. Calibration methods to reduce the effects of pupil non-uniformity are provided in U.S. Patent Publication No. 2004 / 0169861 to Mieher et al. and assigned to KLA-Tencor Corporation, the contents of which are incorporated herein by reference in its entirety.

[0098] In yet another aspect, the SXR scatterometry signals at zeroth, non-zeroth order, or any combination thereof are employed to resolve values characterizing the shape parameters of the structure under measurement, e.g., CD, H, SWA, profile parameters, etc. In some embodiments, SXR scatterometry overlay and shape parameter measurements are performed simultaneously from measurements of SXR scatterometry collected from the same metrology target. This enables measurement of edge placement error (EPE), e.g., end-of-line shortening, line-to-contact distance, etc. Thus, SXR scatterometry enables edge placement error (EPE) measurement without errors due to target biasing that occurs when performing overlay and CD measurements on different targets. In addition, simultaneous measurement of overlay and CD shape parameters from the same metrology target improves both measurement accuracy and throughput.

[0099] In one example, edge placement error between layers is estimated based on overlay measurements as described herein and based on measurements of shape parameters within each x-ray diffraction order measured at multiple different angles of incidence and multiple different azimuthal angles. Edge placement error (EPE) combines overlay and shape parameter (e.g., CD) errors. In one example, EPE is the difference between a CD value (e.g., width W) and an overlay value (e.g., overlay D) as depicted in Figure 8 Figure 4 Thus, by employing computationally efficient overlay measurements as described herein and using the same intensity measurement data to estimate CD parameter values, measurement of EPE is streamlined.

[0100] Figure 4 A top view of a device structure 400 is depicted, including active fields 401-404, gates 405-408, and contacts 409-420. Figure 4 Edge placement distance EPD1 between gate 407 and contact 418 is illustrated. Figure 4 Edge placement distance EPD2 between gate 408 and contact 418 and edge placement distance EPD3 between gate 406 and contact 414 are also illustrated. Edge placement distances must be carefully controlled to ensure high device yield. If edge placement error associated with any of these edge placement distances is too large, the device will fail. As Figure 5A ​As illustrated in the middle, both overlay error and CD error contribute to EPE. For example, if the layer associated with the contact is misaligned from the layer associated with the gate, then EPE results. Similarly, if the CD associated with the contact structure deviates from the nominal size, then EPE results. For example, contacts 413 and 416 are oversized. The result is the overlap between each contact and the corresponding gate structure and device failure. Also, the three-dimensional shape of each measured structure plays a role. In some examples, the sidewall angle cannot be ignored. In some embodiments of these examples, the structure is characterized by a top CD size and a bottom CD size, rather than a single CD size.

[0101] Additional details regarding EPE measurement are described in U.S. Patent Publication No. 2016 / 0003609 by Shchegrov et al., which is incorporated herein by reference in its entirety.

[0102] In some embodiments, the measurement signals from SXR scatterometry of two cells of metrology targets each having a nominal offset in opposite directions are employed to resolve overlay error.

[0103] Figure 5B Cell 160A is depicted with grating structure 161 A stacked above another grating structure 162A. Grating structures 161 A and 162A have the same pitch, but grating structure 161 A is offset from grating structure 162A by a distance equal to the sum of the nominal offset + OFFSET and the overlay error OVL in one direction.

[0104] Figure 3 Cell 160B is depicted with grating structure 161 B stacked above another grating structure 162B. Grating structures 161 B and 162B have the same pitch, but grating structure 161 A is offset from grating structure 162A by a distance equal to the sum of the nominal offset - OFFSET and the overlay error OVL in the opposite direction compared to metrology target 160A. Nominal ly, metrology targets 160A and 160B are identical, with the exception of the direction of the offset of the top grating structure relative to the bottom grating structure.

[0105] The value of the overlay error OVL associated with the measurement target is based on the difference between the detected intensities in the +1 st order and -1 st order associated with cell 160A and the difference between the detected intensities in the +1 st order and -1 st order associated with cell 160B. In one example, computing system 130 determines difference signals Dl and D2 as illustrated by equation (1), where I A +1 is the measured intensity from the +1 st order of cell 160A, I A -1is the measured intensity of the -1 level from unit 160A, I B +1 is the measured intensity of the +1 level from unit 160B, and I B -1 is the measured intensity of -1 level from unit 160B.

[0106]

[0107]

[0108] Assuming a linear relationship between the differential signal value and the actual offset, the overlay error is determined directly from the differential signal as illustrated by equation (2), where the offset is the magnitude of the nominal offset distance as described above.

[0109]

[0110] In some examples, metrology based on SXR scatterometry involves determining parameters of interest that characterize a sample, such as overlay error, shape parameters, etc., by inversely solving a predetermined measurement model with measured SXR scatterometry data. Inverse solution methods include, but are not limited to, model-based regression, tomography, machine learning, or any combination thereof. In this manner, the target parameter is estimated by solving the parameterized measurement model for values ​​that minimize the error between the measured scattered x-ray intensity and the modeled result.

[0111] In some embodiments, the computing system 130 is configured to generate a structural model (e.g., a geometric model, a material model, or a combined geometric and material model) of the measured structure of the sample, generate an SXR response model that includes at least one geometric parameter from the structural model, and resolve at least one sample parameter value by performing a fit analysis of the SXR measurement data with the SXR response model. An analysis engine is used to compare the simulated SXR signal with the measured data, thereby allowing the determination of geometric and material properties of the sample, such as electron density. Figure 6 In the embodiment depicted in , computing system 130 is configured as a model building and analysis engine configured to implement model building and analysis functionality as described herein.

[0112] Figure 6 is a diagram illustrating an exemplary model building and analysis engine 180 implemented by computing system 130. Figure 10As depicted in the middle, the model construction and analysis engine 180 includes a structure model construction module 182 that generates a structure model 181 of the measured structure of the sample. The structure model 182 is received as input to an SXR response function construction module 183. The SXR response function construction module 183 generates an SXR response function model 184 based at least in part on the structure model 182. In some examples, the SXR response function model 184 is based on x-ray form factors, also known as structure factors.

[0113]

[0114] where F is the form factor, q is the scattering vector, and p(r) is the electron density of the sample in spherical coordinates. Then, the x-ray scattering intensity is given by the following equation.

[0115]

[0116] The SXR response function model 184 is received as input to a fitting analysis model 185. The fitting analysis model 185 compares the modeled SXR response with the corresponding measured data 135 to determine the geometry and material properties of the sample.

[0117] In some examples, the fitting of the modeled data to the experimental data is achieved by minimizing a chi-squared value. For example, for SXR measurements, the chi-squared value can be defined as

[0118]

[0119] where, is the measured SXR signal 135 in “channel” j, where the index j describes a set of system parameters, such as diffraction order, energy, angular coordinate, etc. is the modeled SXR signal S L for a set of structure (target) parameters v1,..., v j where these parameters describe geometry (CD, sidewall angle, overlay, etc.) and material (electron density, etc.). s SXR，j is the uncertainty factor associated with the jth channel. N SXR is the total number of channels in x-ray metrology. L is the number of parameters characterizing the metrology target.

[0120] Equation (5) assumes that the uncertainty factors associated with different channels are uncorrelated. In examples where the uncertainty factors associated with different channels are correlated, the covariance between the uncertainty factors can be calculated. In these examples, the chi-squared value for the SXR measurement can be expressed as

[0121]

[0122] where, VSXR is the covariance matrix of the SXR channel uncertainties, and T denotes transpose.

[0123] In some examples, the fitting analysis model 185 resolves at least one sample parameter value by performing a fitting analysis on the SXR measurement data 135 and the SXR response model 184. In some examples, the fitting analysis model 185 is configured to determine a value of a sample parameter value by minimizing a value of a chi-squared statistic. optimization.

[0124] As described above, the fitting of the SXR data is achieved by minimizing a chi-squared value. However, in general, the fitting of the SXR data can be achieved by other functions.

[0125] The fitting of the SXR metrology data is advantageous for any type of SXR technique that provides sensitivity to the geometric and / or material parameters of interest. The sample parameters can be deterministic (e.g., CD, SWA, etc.) or statistical (e.g., rms height of sidewall roughness, roughness correlation length, etc.) as long as the proper model describing the SXR beam interaction with the sample is used.

[0126] In general, the computing system 130 is configured to access the model parameters in real-time using real-time critical dimension (RTCD), or it can access a library of pre-computed models in order to determine a value of at least one sample parameter value associated with the sample 101. In general, some form of CD engine can be used to evaluate the difference between an assigned CD parameter of a sample and the CD parameter associated with the measured sample. Exemplary methods and systems for computing sample parameter values are described in U.S. Patent No. 7,826,071, issued to KLA-Tencor Corporation on November 2, 2010, which is incorporated herein by reference in its entirety.

[0127] In some examples, the model building and analysis engine 180 improves the accuracy of the measured parameters by any combination of side-fed analysis, feed-forward analysis, and parallel analysis. Side-fed analysis refers to taking multiple data sets on different areas of the same sample and transferring a common parameter determined from a first data set to a second data set for analysis. Feed-forward analysis refers to taking data sets on different samples and using a stepwise copy accurate parameter feed-forward method to forward pass common parameters to subsequent analysis. Parallel analysis refers to applying a non-linear fitting method to multiple data sets in parallel or simultaneously, where at least one common parameter is coupled during the fitting.

[0128] Multi-tool and structure analysis refers to feed-forward, side-feed, or parallel analysis based on regression, look-up tables (i.e., "library" matching), or another fitting process of multiple data sets. Exemplary methods and systems for multi-tool and structure analysis are described in U.S. Patent No. 7,478,019, issued to KLA-Tencor Corporation on January 13, 2009, the entirety of which is incorporated herein by reference.

[0129] In yet another aspect, an initial estimate of the value of one or more parameters of interest is determined based on SXR measurements performed at a single orientation of the incident x-ray beam relative to the measurement target. The initial estimated value is implemented as a starting value for the parameter of interest for a regression of a measurement model having measurement data collected from SXR measurements at multiple orientations. In this way, a close estimate of the parameter of interest is determined with relatively small amount of computational effort, and a refined estimate of the parameter of interest is obtained with less overall computational effort by implementing this close estimate as a starting point for a regression of a much larger data set.

[0130] In some embodiments, overlay error values associated with a measurement target or a target in a die are determined directly from detected intensities within one or more non-zero diffraction orders based on a trained machine learning based measurement model. In these embodiments, overlay error is extracted directly from SXR measurement data based on a trained machine learning based model.

[0131] In some embodiments, the targets measured by the SXR metrology system are not periodic, yet a trained machine learning based measurement model is able to extract overlay error from SXR measurements.

[0132] In some embodiments, the machine learning based measurement model is trained based on SXR measurement data collected from a design of experiments (DOE) target having known overlay error values. In some embodiments, the overlay error associated with the target is measured by a trusted reference metrology system, such as an electron beam based metrology system, a scanning electron microscope, etc. In some embodiments, the DOE target is a design rule metrology target. In some embodiments, the DOE target is a device structure in a die.

[0133] In some instances, the machine learning based measurement model is a neural network model, a support vector machine model, etc. Additional details are described in U.S. Patent No. 10,352,876 to Shchegrov et al. and assigned to KLA-Tencor Corporation, the contents of which are incorporated herein by reference in their entirety.

[0134] In some other embodiments, a machine learning based measurement model is trained based on SXR measurement data collected from a design of experiment (DOE) target (e.g., a periodic target) and known overlay error values associated with in-die device structures proximate to the measured DOE target. In these embodiments, the overlay error associated with the in-die device structures is measured by a trusted reference metrology system (e.g., an electron beam based metrology system, a scanning electron microscope, etc.). In this way, the trained machine learning based measurement model estimates the overlay error of the in-die device structures based on measurement data collected from SXR scatterometry measurements of the nearby design rule metrology targets.

[0135] Figure 7 A semiconductor die 190 is illustrated that includes a device region 191 and a scribe region 192 that circumscribes the device region. A number of design rule metrology targets are fabricated within the scribe region 192, such as design rule target 193. Similarly, a number of device structures are fabricated within the device region 191, such as device structure 194. In some embodiments, a machine learning based measurement model estimates the overlay error associated with the device structure 194 based on measurements of SXR scatterometry measurements of the design rule target 193.

[0136] Additional details regarding machine learning based measurement models are described in U.S. Patent Publication No. 2016 / 0003609 by Shchegrov et al., which is incorporated herein by reference in its entirety.

[0137] In some other embodiments, a machine learning based measurement model is trained based on SXR measurement data collected from a design of experiment (DOE) target and shape parameter values determined based on the SXR measurement data. In some examples, the SXR measurement data is employed to estimate shape parameter (e.g., CD, H, SWA, profile parameters, etc.) values as described above. These shape parameter values are employed along with the SXR measurement data and known overlay error values associated with in-die device structures proximate to the measured DOE target to train the machine learning based measurement model. The additional structure information improves the accuracy of the trained model.

[0138] In some embodiments, the overlay measurement based on SXR scatterometry involves illuminating a sample with SXR radiation and detecting the intensity of resulting diffraction orders with respect to a plurality of angles of incidence, a plurality of wavelengths, or both, of the sample. Also, the overlay error associated with the measurement target is determined based on a modulation of a plurality of intensities within each of one or more non-zero diffraction orders at each of a plurality of measurement instances.

[0139] In these embodiments, the measurements of SXR scatterometry measurements of metrology targets are performed at a number of different angles of incidence and azimuthal angles.Figure 7 A wafer 101 containing a metrology target 120 is depicted. An x-ray illumination source illuminates the metrology target 120 with an x-ray radiation beam 114 at an angle of incidence AOI and an azimuth angle Az. The angle of incidence and azimuth angle of the incident x-ray radiation beam are defined with respect to a coordinate system { B X, B Y, B Z} fixed to the sample 101. Figure 3 As depicted in B Y, B X, B Z} of the metrology target 120 contains a grating structure that extends in the B X- B Z plane. In this sense, a change in the angle of incidence can be considered as a rotation of the wafer 101 about the B Y axis that is coplanar with the wafer 101. Similarly, the azimuth angle is defined as the angle of the incident beam projected onto the B X- B Y plane with respect to the B X axis. In this sense, a change in the azimuth angle can be considered as a rotation of the wafer 101 about the B Z axis that is normal to the wafer 101.

[0140] As illustrated in Figure 3 , the metrology tool 100 contains a sample positioning system 140 configured to both align and orient the sample 101 over a large out-of-plane angular orientation range relative to the SXR scatterometer. In other words, the sample positioning system 140 is configured to rotate the sample 101 about one or more rotation axes that are coplanar with and aligned normal to a surface of the sample 101 over a large angular range. In this way, every location on the surface of the sample 101 can be used to make measurements over a rotation range about the axes of a coordinate system { B X, B Y, B Z} fixed to the sample 101.

[0141] In the embodiment depicted in Figure 8 , a single incident x-ray radiation beam is illustrated. The orientation of the single beam relative to the wafer is defined by a single nominal angle of incidence and azimuth angle. For embodiments employing a single illumination beam, x-ray diffraction measurements associated with multiple different angles of incidence and azimuth angles are performed sequentially. However, in general, x-ray diffraction measurements associated with multiple different angles of incidence and azimuth angles can be performed simultaneously. In some embodiments, one or more x-ray sources and one or more sets of x-ray optics can be employed such that the metrology target is illuminated simultaneously from multiple directions discretely or continuously in angle of incidence and azimuth angle.

[0142] When measured at multiple different angles of incidence and multiple different azimuthal angles, the vertical stack of two or more structures in different layers of the metrology target influences the x-ray diffracted signal in a strong and unique way. Therefore, overlay and shape parameter values can be estimated based on the measured intensities.

[0143] In these embodiments, the estimation of overlay involves a parameterization of the intensity modulation of a common level, such that low frequency shape modulation is described by a basis function set or ratio, and high frequency overlay modulation is described by an affine circular function that includes parameters indicative of overlay.

[0144] In one example, the parameterization is derived from an analysis of the exact overlay problem. Figure 9 A layer-type metrology target 150 is depicted, including two periodic line arrays 151 and 152 each having a periodicity P. The line arrays are vertically separated by a distance S, and the line arrays are offset by an overlay distance D. The height and width of each of the line arrays are given by parameters H and W, respectively.

[0145] Additionally, the electron density of the top line is given by a parameter δ0, and the bottom line is given by δ1. The metrology target 150 is illuminated by an x-ray radiation beam having a wavelength λ. When the projection of the light rays is normal to the periodicity of the grating, the incident x-ray radiation beam is at an angle of incidence θ and an azimuthal angle φ Illumination is on the metrology target, where For such a grating, we define the number of wavelengths as k0= 2π / λ, and the number of gratings in the 1D periodic direction as k x = 2π / P, where m is the order. The angle of incidence is measured with respect to an axis normal to the wafer.

[0146] Analysis of the diffracted intensity provides an approximation of the intensity of each level, as explained in equation (7).

[0147]

[0148] Equation (7) explains that there is a modulation of the level intensity not only by changing the angle of incidence θ, but also by changing the azimuthal angle φ In another way of stating, by changing the angle of incidence, we can expect a modulation of the projection of the diffracted level from a grating periodicity dimension aligned with B X to a direction aligned with B Z. Additionally, by changing the azimuthal angle, we can expect a modulation of the projection of the diffracted level from a grating periodicity dimension aligned with B X to a direction aligned with B Y. Additionally, the changes in the angle of incidence and the azimuthal angle can be coordinated to enhance the overlay signal. For example, since the angle of incidence is a scaled factor of the azimuthal angle Changes in the azimuth angle can slow down the shape and overlay modulation. Finally, equation (7) also shows that the modulation in the level strength of the shape parameters W and H generally has low spatial frequencies compared to the overlay modulation described by the last term of equation (7).

[0149] Due to the relatively low spatial frequency modulation due to shape, this modulation can be modeled by a low order polynomial (e.g., a linear or quadratic function). Then, the modulation in the separation distance S and overlay D can be represented by the cosine terms illustrated in equation (7). Thus, the simplified model of the intensity of each level takes the additive and multiplicative form as illustrated by equations (8a) and (8b), respectively.

[0150]

[0151]

[0152] Without explicit knowledge of the shape, the shape function defined by the first term of equation (8a) and the first factor of equation (8b) models the shape modulation as a linear combination of basis functions θjweighted by the parameters aj. As illustrated in equations (8a) and (8b), a monomial basis is employed to describe the shape changes. However, in general, any polynomial, rational, or any kind of basis set can be employed.

[0153] The parameter b defines the modulation depth. The parameters D and S define the overlay. By changing the angle of incidence, the azimuth angle, or both, the resulting data of any order can be fitted with any suitable curve fitting constant to the parameters a j , b, D, and S. The overlay is imparted by the fitting of the parameter D.

[0154] The simplified model for overlay measurement described above illustrates a phenomenological approach to model the intensity variation of the diffraction orders based on changes in the angle of incidence and azimuth angle. In general, the model can be based on other waveforms and non-polynomial basis functions.

[0155] By fitting the measured intensity signal with the phenomenological simple functions, overlay shifts associated with multiple layers can be estimated in an efficient manner. Thus, performed at a relatively low computational cost and without external reference metrology measurements, overcoming the limitations of current methods based on SEM, optical metrology, or other proposed x-ray metrology techniques.

[0156] Figure 8 Depicts the overlay Figure 9A plot 170 of the simulation of the fitting results of the metrology target 150 depicted in FIG. 16B. Plot line 171 depicts the simulation of the normalized intensity of the -2 diffraction order over the angular range. Plot line 172 depicts the simulation of the normalized intensity of the +2 diffraction order over the same angular range. Plot line 173 depicts the results of the fitting of the simulated diffraction intensity by a model of the type described by reference equation (8). As Figure 3 As illustrated in FIG. 16B, the simplified model described by reference equation (8) provides a close fit to the simulated intensity values.

[0157] As illustrated by equation (8), the overlay modulation is an even function of the diffraction order. Therefore, data from both the positive and negative orders can be averaged or jointly fitted. In addition, multiple orders can be jointly fitted. In some examples, different ranges in angular space can be used for each different diffraction order. Additional details are described in Hench et al. and assigned to KLA-Tencor Corp. WIPO Publication No. WO2016176502A1, the contents of which are incorporated herein by reference in their entirety.

[0158] In some embodiments, the actual device targets are non-periodic. By calibrating overlay measurements to reference measurements, SXR scatterometry techniques can be used to estimate overlay of non-periodic structures based on measurements of design rule targets having sufficient periodicity. This effectively overcomes the limitation of scatterometry measurements that require the measured targets to be periodic or approximately periodic.

[0159] In some embodiments, calibrated SXR measurements are employed to estimate overlay error associated with in-die device structures as part of the advanced detection in design (ADI) process monitoring. In some embodiments, the in-die device structures are non-periodic logic devices.

[0160] In some embodiments, a single overlay error calibration value is employed to calibrate SXR scatterometry based overlay measurements. A reference metrology system (e.g., a SEM) is employed to measure overlay error associated with in-die device structures. In addition, an SXR scatterometry system (e.g., SXR scatterometry tool 100) is employed to measure nearby design rule metrology targets. The difference between the overlay error measured by the SXR scatterometry system and the overlay error measured by the reference metrology system is the overlay error calibration value. Subsequent overlay measurements of the design rule metrology targets are adjusted by the overlay error calibration value to estimate overlay error associated with nearby in-die device structures. More specifically, the overlay error calibration value is added to the overlay error measured by the SXR scatterometry to estimate the overlay error associated with the in-die device structures. A limitation of this approach is that it does not compensate for target error induced by the fact that the reference measurements are performed on targets different than the SXR measurements.

[0161] In one example, this limitation is overcome by feeding back the SEM calibration to SXR measurements of in-die device structures. This approach is feasible in cases where the SXR measurements do not impose any physical changes to the measured targets.

[0162] In another example, two overlay error calibration values are employed to calibrate overlay measurements based on SXR scatterometry measurements. In these embodiments, a reference metrology system (e.g., a SEM) is employed to measure overlay error associated with in-die device structures and to measure overlay error associated with nearby design rule metrology targets. Additionally, an SXR scatterometry system (e.g., SXR scatterometry tool 100) is employed to measure overlay error associated with nearby design rule metrology targets. The difference between the overlay error associated with in-die device structures measured by the reference metrology system and the overlay error associated with nearby design rule metrology targets is a first overlay error calibration value. The difference between the overlay error associated with design rule targets measured by the reference metrology system and the overlay error associated with design rule targets measured by the SXR scatterometry system is a second overlay error calibration value. Subsequent overlay measurements of design rule metrology targets are adjusted by both the first overlay error calibration value and the second overlay error calibration value to estimate overlay error associated with nearby in-die device structures. More specifically, the first overlay error calibration value and the second overlay error calibration value are added to the overlay error measured by the SXR scatterometry to estimate the overlay error associated with in-die device structures.

[0163] Reference measurements performed by the e-beam tool can be performed at high (e.g., 10 kV or higher) voltages to image one or more underlying layers. In general, the landing energy of the e-beam tool can be adjusted to maximize imaging performance.

[0164] SXR scatterometry tools as described herein are capable of performing many different types of measurements related to semiconductor manufacturing. For example, SXR scatterometry tools can be employed to measure characteristics of one or more targets, such as critical dimensions, overlay, sidewall angles, film thicknesses, process-related parameters (e.g., focus and / or dose), etc. Measurement targets can include periodic regions of interest, such as gratings in memory dies. Measurement targets can include multiple layers, and the thickness of one or more layers can be measured by the SXR scatterometry tool. Measurement targets can be located within a scribe line or within a die itself. In some embodiments, multiple targets are measured simultaneously or sequentially by one or more metrology tools, as described in US 7,478,019, the contents of which are incorporated herein by reference in their entirety. For example, data from such measurements can be combined and used to feed forward, feed back, and side feed corrections to processes (e.g., photolithography, etching) in a semiconductor manufacturing process.

[0165] The measurement of the parameters of interest often involves several algorithms. In some embodiments, an electromagnetic solver is used to model the interaction of the incident beam with the sample, and algorithms such as RCWA, FEM, moment method, surface integral method, volume integral method, FDTD, Born approximation (BA), Distorted Wave BA (DWBA), and others. For model-based measurements, a geometry engine or in some cases a process modeling engine or a combination of both is often used to model the target of interest. For example, a geometry engine is implemented in the AcuShape software available from KLA-Tencor, Milpitas, California.

[0166] The collected data can be analyzed by several data fitting and optimization techniques and arts, including: libraries; fast degradation models; regression; machine learning algorithms such as neural networks, support vector machines (SVM); dimensionality reduction algorithms such as PCA (principal component analysis), ICA (independent component analysis), LLE (locally linear embedding)); sparse representations such as Fourier or wavelet transforms; Kalman filters; algorithms that facilitate matching between the same or different tool types, and others. The collected data can also be analyzed by algorithms that do not include modeling, optimization, and / or fitting, to extract size and material information about the structure.

[0167] In some embodiments, a multi-layer overlay metrology target is designed such that the set of separation parameters differs between each combination of two layers, and the minimum separation distance between all layer combinations is maximized, subject to a constraint on the overall height of the metrology target.

[0168] In some embodiments, a multi-layer overlay metrology target is designed with different pitches at different layers, such that a diffraction order generated from one layer constructively interferes with a different diffraction order of another layer. In one embodiment, a periodic grating structure located in a first layer has a pitch equal to 2A, where A is an arbitrary positive constant. Another periodic structure located in a different layer has a pitch equal to 3A. In this example, the second diffraction order of the first layer constructively interferes with the third diffraction order of the second layer. Thus, the intensity measurements detected at these order pairs are dominated by overlay between the two layers. Conversely, the intensity measurements detected at different order pairs that are not subject to constructive interference by overlay are dominated by shape parameters. Thus, in some embodiments, metrology overlay targets are designed with specific grating structures to increase sensitivity to overlay at specific grating order pairs, and also provide intensity data useful for shape parameter value estimation.

[0169] Similarly, the multi-layer overlay metrology target is designed to have different pitch orientations at different layers such that the diffracted orders generated from one layer constructively interfere with different diffracted orders of another layer. In general, a set of layers having different periodicities (e.g., different grating pitches), different pitch orientations, or any combination thereof, generates a set of scattering vectors, each scattering vector being associated with a different layer. The overlay metrology target is designed such that a predetermined subset of the scattering vectors is aligned. In this manner, sensitivity to overlay of the layers corresponding to the predetermined subset of scattering vectors is enhanced.

[0170] In general, the overlay metrology target can include a ID periodic structure (i.e., where periodic in one direction and constant in another direction), a 2D periodic structure (i.e., periodic in two directions), or any combination thereof. For a 2D periodic target, the two periodic directions can be perpendicular to each other or can not be perpendicular to each other. Also, the pitch of each of the constituent structures can be the same or different.

[0171] It should be appreciated that the various steps described throughout the present disclosure can be performed by a single computer system 130 or, alternatively, multiple computer systems 130. Also, different subsystems of the system 100 (e.g., the sample positioning system 140) can include a computer system suitable for performing at least a portion of the steps described herein. Thus, the foregoing description should not be interpreted as a limitation on the present disclosure but merely an illustration. Further, the one or more computer systems 130 can be configured to perform any other step(s) of any of the method embodiments described herein.

[0172] Additionally, the computer system(s) 130 can be communicatively coupled to the detector 119 and the illumination optics 115 in any manner known in the art. For example, one or more computer systems 130 can be coupled to a computer system 130 associated with the detector 119 and the illumination optics 115, respectively. In another example, any of the detector 119 and the illumination optics 115 can be directly controlled by a single computer system coupled to the computer system 130.

[0173] The computer system(s) 130 can be configured to receive and / or acquire data or information from subsystems of the system (e.g., the detector 119, the illumination optics 115, etc.) over a transmission medium that can include wired and / or wireless portions. In this manner, the transmission medium can serve as a data link between the computer system 130 and other subsystems of the system 100.

[0174] The computer system 130 of the metrology system 100 can be configured to receive and / or acquire data or information (e.g., measurement results, modeling inputs, modeling results, etc.) from other systems over a transmission medium, which can include wired and / or wireless portions. In this manner, the transmission medium can serve as a data link between the computer system 130 and other systems (e.g., the on-board metrology system 100, external memory, or external systems). For example, the computing system 130 can be configured to receive measurement data (e.g., signals 135) from a storage medium (i.e., memory 132 or 190) via a data link. For example, scatterometry data collected by the detector 119 can be stored in a permanent or semi-permanent memory device (e.g., memory 132 or 190). In this regard, measurement results can be imported from on-board memory or from an external memory system. Also, the computer system 130 can send data to other systems via a transmission medium. For example, overlay values 186 determined by the computer system 130 can be stored in a permanent or semi-permanent memory device (e.g., memory 190). In this regard, measurement results can be exported to another system.

[0175] The computing system 130 can include, but is not limited to, a personal computer system, mainframe computer system, workstation, image computer, parallel processor, or any other device known in the art. In general, the term "computing system" can be broadly defined to encompass any device having one or more processors, which executes instructions from a memory medium.

[0176] Program instructions implementing methods such as those described herein can be transmitted via a transmission medium (e.g., a wire, cable, or wireless transmission link). For example, as Figure 11 illustrated in FIG. 1, program instructions stored in the memory 132 are transmitted to the processor 131 via the bus 133. The program instructions 134 are stored in a computer-readable medium (e.g., the memory 132). Exemplary computer-readable media include read-only memory, random access memory, magnetic or optical disk, or tape.

[0177] In some embodiments, scatterometry analysis as described herein is implemented as part of a fabrication process tool. Examples of fabrication process tools include, but are not limited to, photolithography exposure tools, film deposition tools, implant tools, and etch tools. In this manner, results of SXR scatterometry analysis are used to control the fabrication process. In one example, measurement data of SXR scatterometry collected from one or more targets is sent to a fabrication process tool. The measurement data of the SXR scatterometry is analyzed as described herein and the results are used to adjust the operation of the fabrication process tool.

[0178] Measurements such as scatterometry as described herein can be used to determine characteristics of various semiconductor structures. Exemplary structures include, but are not limited to, FinFETs, low size structures (e.g., nanowires or graphene), structures less than 10 nm, lithography structures, through substrate vias (TSVs), memory structures (e.g., DRAM, DRAM 4F2, flash, MRAM), and high aspect ratio memory structures. Exemplary structure characteristics include, but are not limited to, geometric parameters (e.g., line edge roughness, line width roughness, hole size, hole density, sidewall angle, profile, critical dimension, pitch) and material parameters (e.g., electron density, composition, grain structure, morphology, stress, strain, and element identification).

[0179] ​ A method 200 suitable for implementation by embodiments of the metrology system 100 of the present disclosure is illustrated. In one aspect, it is recognized that data processing blocks of the method 200 can be implemented via preprogrammed algorithms executed by one or more processors of the computing system 130. While the following description is presented in the context of the metrology system 100, it is recognized herein that particular structural aspects of the metrology system 100 are not meant to be limiting and should be interpreted as illustrative only.

[0180] In block 201, a first instance of a design rule target disposed on a substrate is illuminated with a soft x-ray (SXR) radiation beam having an energy in a range between 10 and 5,000 electron volts. The design rule target is a multilayer target.

[0181] In block 202, a first plurality of intensities associated with +1 / -1 diffraction orders of an amount of SXR radiation scattered from the first instance of the design rule target in response to the incident SXR radiation beam are detected.

[0182] In block 203, a first overlay error value associated with the first instance of the design rule target is estimated based on the first plurality of detected intensities within the +1 / -1 diffraction orders.

[0183] In block 204, an overlay error value associated with an in-function device structure in a first die is estimated based on measurements of the in-function device structure by a scanning electron microscope. The first instance of the design rule target and the in-function device structure in the first die are fabricated according to the same fabrication process rules.

[0184] In block 205, an overlay calibration value is determined based on the overlay error value associated with the in-function device structure in the first die and the first overlay error value associated with the first instance of the design rule target.

[0185] As described herein, the term "critical dimension" includes any critical dimension of a structure (e.g., bottom critical dimension, middle critical dimension, top critical dimension, sidewall angle, grating height, etc.), a critical dimension between any two or more structures (e.g., a distance between two structures), and a displacement between two or more structures (e.g., an overlay displacement between overlay grating structures, etc.). A structure can include a three-dimensional structure, a patterned structure, an overlay structure, etc.

[0186] As described herein, the term "critical dimension application" or "critical dimension measurement application" includes any critical dimension measurement.

[0187] As described herein, the term "metrology system" includes any system used at least in part to characterize a sample in any aspect including critical dimension applications and overlay metrology applications. However, these technical terms do not limit the scope of the term "metrology system" as described herein. Additionally, the metrology systems described herein can be configured for measuring patterned wafers and / or unpatterned wafers. The metrology systems can be configured as an LED inspection tool, an edge inspection tool, a backside inspection tool, a large format inspection tool, or a multi-mode inspection tool (involving data from one or more platforms simultaneously), as well as any other metrology or inspection tool that benefits from the imaged or structures being measured.

[0188] Various embodiments are described herein for semiconductor processing systems (e.g., metrology systems or lithography systems) that can be used to process a sample. The term "sample" as used herein refers to a wafer, a reticle, or any other specimen that can be processed (e.g., printed or inspected for defects) by means known in the art.

[0189] As used herein, the term "wafer" generally refers to a substrate formed of semiconductor or non-semiconductor material. Examples include, but are not limited to, single crystalline silicon, gallium arsenide, and indium phosphide. These substrates can typically be found in and / or processed in a semiconductor fabrication facility. In some cases, a wafer can include only a substrate (i.e., a bare wafer). Alternatively, a wafer can include one or more layers of different materials formed on a substrate. One or more layers formed on a wafer can be "patterned" or "unpatterned." For example, a wafer can include a plurality of dies having repeatable patterned features.

[0190] A "reticle" can be a reticle at any stage of a reticle fabrication process or a finished reticle that can or can not be released for use in a semiconductor fabrication facility. A reticle or "mask" is generally defined as a substantially transparent substrate having substantially non-transparent areas formed thereon and configured to pattern. For example, the substrate can include a glass material such as amorphous S1O2. The reticle can be deposited over a wafer of resist overlay during an exposure step of a lithography process such that the pattern on the reticle can be transferred to the resist.

[0191] One or more layers formed on the wafer can be patterned or unpatterned. For example, the wafer can include a plurality of dies each having repeatable patterned features. The formation and processing of these layers of materials can ultimately result in a completed device. Many different types of devices can be formed on a wafer, and the term wafer as used herein is intended to encompass wafers on which any type of device known in the art has been fabricated.

[0192] In one or more exemplary embodiments, the functions described can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media can be any available media that can be accessed by a general purpose or special purpose computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code means in the form of instructions or data structures and that can be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, include compact discs (CDs), laser discs, XRF discs, digital versatile discs (DVDs), floppy disks, and Blu-ray discs where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0193] Although certain specific embodiments are described above for instructional purposes, the present teachings are open to interpretation and variation known to those skilled in the art and use of the content of this patent document is not limited to any specific embodiment described. Thus various modifications, changes, and combinations of the described embodiments are possible and will be apparent to the skilled artisan in the art upon consideration of the content of this patent document.

Claims

1. A metrology system comprising: a soft x-ray (SXR) illumination source configured to illuminate a first instance of a measurement target disposed on a substrate with a beam of SXR radiation having an energy in a range between 10 and 5,000 electron volts, wherein the measurement target includes: a first structure disposed in a first layer fabricated at a first height above the substrate; and a second structure disposed in a second layer fabricated at a second height above the substrate; an x-ray detector configured to detect a plurality of intensities each associated with one or more non-zero diffraction orders of an amount of x-ray radiation scattered from the measurement target in response to an incident x-ray radiation beam; one or more beam slits or apertures configured to select a set of illumination wavelengths to simultaneously illuminate a metrology target under test; and a computing system configured to: estimate overlay error values associated with the measurement target or an active device structure in a corresponding die based on a plurality of the detected intensities within each of the one or more non-zero diffraction orders; wherein determination of the overlay error values associated with the active device structure in the corresponding die involves summation of the estimated overlay error values associated with the measurement target based on a plurality of the detected intensities within each of the one or more non-zero diffraction orders and correction values.

2. The metrology system of claim 1, the x-ray detector is further configured to detect an intensity associated with a zero diffraction order of the amount of x-ray radiation scattered from the measurement target in response to the incident x-ray radiation beam, and the computing system is further configured to: estimate values of one or more parameters characterizing a shape of the measurement target based on the detected intensity within the zero diffraction order, a plurality of the detected intensities within each of the one or more non-zero diffraction orders, or any combination thereof; and estimate a value of an edge placement error of the measurement target based on the estimated overlay error values and the estimated values of the one or more parameters characterizing the shape of the measurement target. ​ 3. The metrology system of Claim 1, the SXR illumination source further configured to illuminate a second instance of the measurement target disposed on the substrate with the SXR radiation beam, wherein the first structure of the first instance of the measurement target is offset from the second structure of the first instance of the measurement target by an offset distance in a direction aligned with the first layer, wherein the first structure of the second instance of the measurement target is offset from the second structure of the second instance of the measurement target by the offset distance in a direction opposite the direction aligned with the first layer, the x-ray detector configured to detect a plurality of intensities each associated with one or more non-zero diffraction orders of an amount of x-ray radiation scattered from the second instance of the measurement target in response to the incident x-ray radiation beam, and wherein the estimation of the overlay error value associated with the measurement target is based on a difference between the detected intensities within +1 and -1 diffraction orders associated with the first instance of the measurement target and a difference between the detected intensities within the +1 and -1 diffraction orders associated with the second instance of the measurement target.

4. The metrology system of Claim 1, wherein the estimation of the overlay error value associated with the measurement target is based on a fitting analysis of the detected intensities within the one or more non-zero diffraction orders with a physics-based measurement model.

5. The metrology system of Claim 1, wherein the SXR radiation beam is incident on the measurement target at different nominal angles of incidence, different nominal angles of azimuth, or both, at a plurality of measurement instances.

6. The metrology system of Claim 5, wherein the estimation of the overlay error value associated with the measurement target is based on a modulation in the plurality of intensities within each of the one or more non-zero diffraction orders at each of the plurality of measurement instances.

7. The metrology system of Claim 1, wherein the overlay error value associated with the measurement target or a target in the corresponding die is determined directly from the detected intensities within the one or more non-zero diffraction orders by a trained machine-learned measurement model.

8. The metrology system of Claim 7, wherein the measurement target is not periodic.

9. The metrology system of claim 7, the soft x-ray (SXR) illumination source further configured to illuminate a plurality of design of experiment measurement targets with the SXR radiation beam having an energy in a range between 10 and 5,000 electron volts, the x-ray detector further configured to detect a plurality of intensities each associated with one or more non-zero diffraction orders of an amount of x-ray radiation scattered from each of the plurality of design of experiment measurement targets in response to the incident x-ray radiation beam, the computing system further configured to train the trained machine learning based measurement model based on a plurality of the detected intensities associated with each of the design of experiment measurement targets or corresponding die-in-target and known overlay error values.

10. The metrology system of claim 9, wherein the known overlay error values are determined from measurements of the design of experiment measurement targets or corresponding die-in-target by a reference metrology system.

11. The metrology system of claim 1, wherein the first layer is a resist layer.

12. The metrology system of claim 1, wherein the measurement target is a design rule target disposed within a scribe or die-in-place device structure.

13. The metrology system of claim 1, wherein the corresponding die-in-place device structure is not periodic.

14. A method comprising: illuminating a first instance of a design rule target disposed on a substrate with a soft x-ray (SXR) radiation beam having an energy in a range between 10 and 5,000 electron volts, wherein the design rule target is a multi-layer target; detecting a first plurality of intensities associated with +1 / -1 diffraction orders of an amount of SXR radiation scattered from the first instance of the design rule target in response to an incident SXR radiation beam; estimating a first overlay error value associated with the first instance of the design rule target based on the first plurality of detected intensities within the +1 / -1 diffraction orders; estimating an overlay error value associated with a first die-in-place device structure based on a measurement of the first die-in-place device structure by a scanning electron microscope, wherein the first instance of the design rule target and the first die-in-place device structure are fabricated according to a same fabrication process recipe; and determining an overlay calibration value based on the overlay error value associated with the first die-in-place device structure and the first overlay error value associated with the first instance of the design rule target.

15. The method of claim 14, further comprising: illuminating a second instance of the design rule target with a SXR radiation beam having an energy in a range between 10 and 5,000 electron volts; detecting a second plurality of intensities each associated with the +1 / -1 diffraction orders of an amount of SXR radiation scattered from the second instance of the design rule target in response to the incident SXR radiation beam; estimating an overlay error value associated with the second instance of the design rule target based on a plurality of detected intensities within the +1 / -1 diffraction orders; ​ estimating an overlay error value associated with a second in-die device structure based on a sum of the overlay error value associated with the second instance of the design rule target and the overlay calibration value.

16. The method of claim 14, wherein the overlay calibration value is a difference between the overlay error value associated with the first in-die device structure and the first overlay error value associated with the first instance of the design rule target.

17. The method of claim 14, further comprising: estimating a second overlay error value associated with the first instance of the design rule target based on a measurement of the first instance of the design rule target by the scanning electron microscope, wherein the determination of the overlay calibration value is based on a difference between the first overlay error value associated with the first instance of the design rule target and the second overlay error value and a difference between the overlay error value associated with the first in-die device structure and the second overlay error value associated with the first instance of the design rule target.

18. The method of claim 15, further comprising: detecting an intensity associated with a zero diffraction order of an amount of x- ray radiation scattered from the second instance of the design rule target in response to the incident SXR radiation beam; estimating a value of one or more parameters characterizing a shape of the design rule target based on the detected intensity within the zero diffraction order, the plurality of detected intensities within the +1 / -1 diffraction orders, or any combination thereof; and estimating a value of an edge placement error of the second instance of the design rule target based on the overlay error value associated with the second instance of the design rule target and the value of one or more parameters characterizing the shape of the second instance of the design rule target.

19. The method of claim 14, wherein the design rule target is disposed within a scribe or is an in-die device structure.

20. The method of claim 14, wherein the in-die device structure is not periodic.

21. A metrology system, comprising: a soft x-ray (SXR) illumination source configured to illuminate a first instance of a measurement target disposed on a substrate with a SXR radiation beam having an energy in a range between 10 and 5,000 electron volts, an x-ray detector configured to detect a plurality of intensities each associated with one or more non-zero diffraction orders of an amount of x-ray radiation scattered from the measurement target in response to an incident x-ray radiation beam; one or more beam slits or apertures configured to select a set of illumination wavelengths to simultaneously illuminate a metrology target and a non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to: estimate an overlay error value associated with the measurement target or a corresponding in-die device structure based on a plurality of the detected intensities within each of the one or more non-zero diffraction orders; wherein the determination of the overlay error values associated with the device feature structures of interest in the corresponding dies involves a summation of estimated overlay error values associated with the measurement targets within a plurality of the detected intensities in each of the one or more non-zero diffraction orders and a correction value.

Images