X-ray method and system for semiconductor substrate alignment

Abstract

X-rays are directed to first and second substrates in a bonded configuration for bonding together. The X-rays are directed to first and third alignment marks in the first substrate and to second and fourth alignment marks in the second substrate. Fluorescent X-rays are detected as emitted from the first and second alignment marks to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first and second alignment marks. The X-rays are transmitted through the first and second substrates using X-ray Talbot-Lau interferometry to measure a second misalignment of the first and second substrates based on a second detected misalignment of the third and fourth alignment marks.

Description

Cross-references to related applications

[0001] This application claims priority and benefit to the filing date of U.S. non-provisional patent application No. 18 / 620,463, filed on March 28, 2024, which is incorporated herein by reference in its entirety. Technical Field

[0002] This disclosure generally relates to semiconductor manufacturing, and in specific embodiments, to X-ray methods and systems for aligning semiconductor substrates. Background Technology

[0003] Semiconductor integrated circuits (ICs) are typically manufactured by sequentially depositing conductive layers, dielectric layers, and semiconductor layers on a semiconductor substrate to form IC devices. Semiconductor processing involves patterning layers using photolithography and etching to form electronic components and interconnects, such as transistors, resistors, capacitors, metal lines, contacts, and vias, in a monolithic structure.

[0004] The semiconductor industry has traditionally followed Moore's Law, which was initially based on the observation that the number of transistors on a chip roughly doubles every two years, resulting in a rhythmic reduction in feature size (also known as "scaling"), along with performance improvements and cost reductions. However, as transistor features approach atomic dimensions, maintaining this pace has become increasingly challenging. Therefore, the scaling rhythm has evolved from a strict focus on feature size reduction to a more complex process incorporating innovations in 3D structures, new materials, and integration methods.

[0005] The miniaturization of semiconductor technology has been a driving force behind the development of advanced 3D integration technologies, such as wafer-to-wafer (W2W), die-to-die (D2D) bonding, die-to-wafer (D2W) bonding, and multi-die stacking, such as in dynamic random access memory (DRAM) with up to 16 layers or more. The success of the 3D integration process can depend on the precise alignment of components to provide good electrical performance and mechanical anchoring, such as to support high-density interconnect schemes. Such precise alignment using conventional optical alignment methods can become increasingly difficult to perform, especially when dealing with opaque materials characterized by thick substrates of doped silicon (Si) and multiple layers of metallized copper. Summary of the Invention

[0006] In one aspect, a first method for measuring misalignment between substrates is disclosed. The first method includes guiding X-rays to a first substrate and a second substrate in a bonded configuration but not yet bonded together, the guiding including guiding the X-rays to a first alignment mark and a third alignment mark in the first substrate, and to a second alignment mark and a fourth alignment mark in the second substrate. The first method further includes detecting fluorescent X-rays emitted from the first alignment mark and from the second alignment mark to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first alignment mark relative to the second alignment mark, and detecting at least some of the X-rays transmitted through the first substrate and through the second substrate using X-ray Talbot-Lau interferometry to measure a second misalignment of the first substrate relative to the second substrate based on a second detected misalignment of the third alignment mark relative to the fourth alignment mark.

[0007] In another aspect, a second method for measuring misalignment between substrates is disclosed. The second method includes guiding X-rays to a first substrate and a second substrate in a bonded configuration but not yet bonded together, the guiding comprising guiding the X-rays to a first alignment mark in the first substrate and a second alignment mark in the second substrate, detecting fluorescent X-rays emitted from the first substrate and the second substrate in response to X-ray irradiation of the first alignment mark and the second alignment mark, and measuring a first misalignment of the first alignment mark relative to the second alignment mark based at least in part on the wavelength of the fluorescent X-rays corresponding to metals in the first alignment mark and the second alignment mark.

[0008] In another aspect, a third method for measuring misalignment between substrates is disclosed. The third method includes guiding X-rays to a first and a second substrate that are bonded together for subsequent bonding, and transmitting the X-rays through a first alignment mark in the first substrate and a second alignment mark in the second substrate using an X-ray Talbot-Laurent interferometer. In the third method, the first and second alignment marks comprise a first pair of moiré interference gratings. The third method also includes measuring a first misalignment of the first substrate relative to the second substrate based on detecting a first detected misalignment of the first alignment mark relative to the second alignment mark using a first interference pattern and a second interference pattern associated with the first moiré interference grating pair. Detection of the first detected misalignment includes measuring a first displacement of the first interference pattern in a first direction and a second displacement of the second interference pattern in a second direction opposite to the first direction.

[0009] In any of the disclosed embodiments, the third method may include transmitting X-rays through a third alignment mark in the first substrate and a fourth alignment mark in the second substrate using an X-ray Talbot-Lau interferometer. In the third method, the third and fourth alignment marks may include a second moiré interference grating pair. The third method may also include measuring a second misalignment of the first substrate relative to the second substrate based on measuring a second detected misalignment of the third alignment mark relative to the fourth alignment mark using a third and fourth interference pattern associated with the second moiré interference grating pair, wherein the measurement of the second detected misalignment includes measuring a third displacement of the third interference pattern in a first direction and a fourth displacement of the fourth interference pattern in a second direction.

[0010] In another aspect, a fourth method for measuring misalignment between semiconductor substrates is disclosed. The fourth method includes transmitting a first X-ray through a first and second substrate bonded together for subsequent bonding using an X-ray Talbot-Lau (TL) interferometer, including transmitting the first X-ray through a first alignment mark in the first substrate and a second alignment mark in the second substrate to generate a second X-ray, and transmitting the second X-ray through a TL phase grating and a TL analyzer grating to generate a third X-ray. The fourth method also includes receiving the third X-ray at an X-ray detector to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment detected by the X-ray detector relative to the first alignment mark relative to the second alignment mark. Attached Figure Description

[0011] To gain a more complete understanding of this disclosure and its advantages, reference is now made to the following description taken in conjunction with the accompanying drawings, in which:

[0012] Figure 1 This is a description of a dual X-ray measurement system in some embodiments;

[0013] Figure 2 This is a depiction of an X-ray transmission measurement system that uses the Talbot-Laur (TL) interferometry method in some embodiments;

[0014] Figure 3 This is a depiction of composite alignment marks used for semiconductor substrates in some embodiments;

[0015] Figure 4A , Figure 4B and Figure 4C The process for detecting misalignment using X-ray fluorescence (XRF) in some embodiments is described;

[0016] Figure 5A , Figure 5B and Figure 5CThe X-ray transmission misalignment detection process in some embodiments is described;

[0017] Figure 6A , Figure 6B and Figure 6C Moiré-fringe alignment marks for semiconductor substrates are depicted in some embodiments;

[0018] Figure 7A and Figure 7B Moiré-stripe X-ray transmission alignment images are depicted in some embodiments.

[0019] Figure 7C It is the calibration curve of the moiré interference pattern in some implementations;

[0020] Figure 8A , Figure 8B and Figure 8C Moiré-fringe alignment marks for semiconductor substrates are depicted in some embodiments;

[0021] Figure 8D It is the calibration curve of the moiré interference pattern in some implementations;

[0022] Figure 9 It is a flowchart depicting a method for aligning two semiconductor substrates using an X-ray alignment method and system in some embodiments;

[0023] Figure 10A and Figure 10B It is a depiction of multilayer moiré-fringe alignment marks in some embodiments;

[0024] Figure 11 It is a flowchart depicting an alignment measurement method in some embodiments;

[0025] Figure 12 It is a flowchart depicting an alignment measurement method in some embodiments; and

[0026] Figure 13 It is a flowchart depicting an alignment measurement method in some embodiments. Detailed Implementation

[0027] This disclosure describes X-ray methods and systems for aligning semiconductor substrates in various embodiments, such as for aligning two or more semiconductor substrates for a 3D bonding process.

[0028] In the following description, details are illustrated by way of example to facilitate the discussion of the disclosed subjects. However, it will be apparent to those skilled in the art that the disclosed implementations are exemplary and not an exhaustive list of all possible implementations.

[0029] Throughout this disclosure, the hyphenated form of the reference numerals refers to a specific instance of an element, and the unhyphenated form of the reference numerals generally or collectively refers to an element. Thus, as an example (not shown in the figures), device "12-1" refers to an instance of a category of devices that can be collectively referred to as device "12," and any device within that category can generally be referred to as device "12." In the figures and specification, the same numbers are intended to denote the same elements.

[0030] As mentioned above, the semiconductor industry has embraced 3D packaging to provide hybrid devices, such as stacking bonded dies together and mixing different technology nodes in a single final product for cost-effectiveness. In some applications, such 3D ICs are manufactured using W2W bonding, which, for economic reasons, produces multiple 3D ICs or chips in a single operation, which can then be diced from the bonded wafer. Therefore, the W2W bonding process also includes aligning the wafers to each other, such that W2W bonding bonds each 3D IC formed on the wafer together in an aligned manner within specified tolerances. In other applications, D2W bonding is used to bond individual dies to a wafer, which also involves precise alignment for successful 3D IC manufacturing, similar to D2D bonding that bonds individual dies together.

[0031] Accordingly, W2W, D2W, and D2D bonding technologies can be used to form surface bonds between two semiconductor portions. The bonding surfaces can be pretreated to ensure sufficient bond strength, for example, by planarizing each surface to be bonded. In various embodiments, CMP and other surface treatments can be used to pretreat the surfaces of the portions to be bonded, among other processing steps. In various embodiments, the surfaces of the portions on a wafer to be bonded to another wafer can include metals, semiconductors, dielectrics, polymers, or other materials.

[0032] Specifically, when using W2W, D2W, or D2D bonding to form 3D ICs, some or all of the layers of the IC device formed on each wafer can be completed before bonding is performed. For example, certain back-end process (BEOL) layers, which may include conductors, barrier layers, and dielectric insulators, may include many layers. These layers, combined with multiple layers in the front-end process (FEOL) portion of the IC, can subject different types of adjacent materials to the alignment process and thus potentially interfere with the alignment process, especially when using visible light or IR light.

[0033] Typical alignment systems and processes often use near-infrared (NIR) radiation, which may struggle to penetrate thick or heavily doped Si and may not be able to pass through metal layers. Furthermore, NIR can be depth-of-field (DoF) limited when used in high-resolution and high-magnification optical path systems. For example, some typical NIR alignment systems can provide magnification levels of 10x to 50x and correspondingly align semiconductor substrates to tolerances of approximately ±3 μm to ±6 μm. At such magnifications, focal plane adjustment for optimal focus may not correspond to the location of the device features to be aligned, potentially leading to the accumulation of system offset errors that could ultimately cause the total measurement uncertainty (TMU) to exceed acceptable tolerances in process metrology. Therefore, it is desirable to be able to measure 3D bonded substrate alignment through materials opaque to the light wavelength and through metallization layers that are not part of the alignment mark but rather part of the active device, which may be difficult or impossible using conventional NIR radiation and associated alignment techniques.

[0034] Recent technological advancements in semiconductor optical alignment systems have validated the value of moiré fringe-based alignment techniques, which can provide higher alignment resolution by amplifying misalignments with the moiré grating in the optical path. In this way, the moiré fringe-based alignment interference pattern provides X-ray amplification, which can be used to detect misalignments smaller than those detectable with conventional image-based coverage (IBO) metrology tools.

[0035] As will be disclosed in further detail herein, X-ray methods and systems for semiconductor substrate alignment that overcome the potential limitations of using NIR light are disclosed. Furthermore, methods and systems for integrating moiré fringe-based alignment techniques with small-angle scattering (dark field) and phase-contrast-based X-ray techniques for W2W, D2W, or D2D bonds are disclosed. In certain embodiments, talbot-Laur (TL) grating-based X-ray interferometry can be used in conjunction with X-ray fluorescence (XRF) to provide a dual measurement strategy using a single substrate alignment system, such as for D2D and D2W bonding. Integrating dual TL gratings and XRF methods into a single system can improve alignment accuracy and provide process versatility in semiconductor manufacturing for 3D integration in advanced packaging. For example, both TL gratings and XRF methods can be used within the same field of view (FOV) to simplify D2D and D2W bonding alignment using different methods for different spatial resolutions. This simplifies typical methods that can use different types of optics and focusing arrangements, which involve certain reconfiguration and setup operations for different FOVs.

[0036] Some embodiments of the X-ray methods and systems disclosed herein for semiconductor substrate alignment provide alignment mark designs tailored to X-ray analysis methods. In some embodiments, both TL grating and XRF methods can be used within a single FOV, for example, for high-precision coverage and high-sensitivity misalignment measurements in D2D and D2W bonding processes. In some embodiments, the X-ray TL grating method can simultaneously generate different imaging modes for example, for W2W, D2W, and D2D bonding, including transmission absorption imaging (IBO), dark-field or small-angle scattering (IBO) in real space, and phase-contrast imaging (IBO). Alignment marks can be formed using copper (Cu) or other metals and can define finely spaced alignment marks corresponding to metal (e.g., Cu) pads in wafer bonding. Therefore, some embodiments can provide improved accuracy, lower limits for misalignment detection, and linearity of magnification for alignment mark detection and measurement. Some embodiments can provide different locations for in-situ integration with the bonding machine and bonding process, such as for process-integrated metrology to generate local feedback for pre-alignment and post-alignment checks. Some implementations can be used for bond inspection as a standalone metrology tool, as well as other applications.

[0037] Accordingly, some embodiments provide an integral alignment mark design that is assembled into a common FOV and functions in both coarse and fine alignment steps. In some embodiments, the integral alignment mark design can save substrate area by eliminating duplicate or different types of alignment marks used for coarse and fine alignment steps. Due to the single alignment mark design compatible with dual X-ray measurement techniques (TL grating and XRF methods), some embodiments can reduce or eliminate reference errors and refocus to adjust lateral errors that might otherwise add unwanted TMUs, such as for D2D and D2W bonding, which is desirable. Some embodiments can support the fine alignment step by including a high-precision target design with an integral alignment mark design, thereby implementing alignment accuracy of at least 20 nm and target accuracy of 10% or less, with linearity up to 99% over the measurement range.

[0038] Now turn to the attached image. Figure 1 It is a depiction of a dual X-ray measurement system 100 (or simply system 100). Figure 1 This is an illustrative representation and is not necessarily drawn to scale or in 3D. For clarity, Figure 1 Some components have been omitted.

[0039] like Figure 1As shown, system 100 includes an X-ray source 110 that can output an X-ray beam 118 toward a substrate pair 112. The X-ray beam 118 may include incoherent or coherent X-rays having defined frequencies, wavelengths, and phases. The substrate pair 112 may represent a test sample or test object and includes two semiconductor substrates to be bonded, such as a first substrate and a second substrate. In various embodiments, system 100 can be used to analyze and measure the substrate pair 112 when they are misaligned, as described in further detail herein. In a particular embodiment, the substrate pair 112 may be used with system 100 during a pre-bonding process, such as when the first and second substrates are held in a fixture, close to each other, and in the process of pre-bonding alignment.

[0040] Accordingly, system 100 also includes a first detector 114 that receives a transmitted X-ray beam 120 from substrate pair 112, and a second detector 116 that receives backscattered X-rays 122 from substrate pair 112. As shown, the backscattered X-rays 122 may include fluorescent X-rays emitted from substrate pair 112 in response to X-ray beam 118 irradiating substrate pair 112. When atoms in substrate pair 112 absorb energy from the irradiating X-ray beam 118, their electrons are ejected from the inner shell (typically a K-shell or L-shell). Ultimately, electrons from higher energy levels (outer shells) fall into lower energy vacancies. As electrons transition from higher to lower energy levels, energy is released as fluorescent X-rays. The emitted fluorescent X-rays have characteristic energies specific to each element. There are two main methods for measuring these X-rays: energy-dispersive X-ray fluorescence (EDXRF) and wavelength-dispersive X-ray fluorescence (WDXRF). EDXRF uses a semiconductor detector to directly measure the energy of incident fluorescent X-rays, thus distinguishing different elements. WDXRF uses a crystal to disperse X-rays onto a detector according to their wavelength, and the detector then measures the intensity of these X-rays.

[0041] Accordingly, in various embodiments, the second detector 116 may be an X-ray fluorescence (XRF) detector, such as a silicon drift detector (SDD), which measures the energy (wavelength) and intensity of incident X-ray photons in the backscattered X-rays 122. The silicon drift detector (SDD) is an energy-dispersive X-ray fluorescence (EDXRF) detector. The second detector 116 may comprise a high-purity silicon wafer serving as the detection material. When X-rays enter the silicon wafer, they interact with silicon atoms and generate electron-hole pairs proportional to the energy of the X-rays. The wafer has a series of annular electrodes or drift rings on its surface. These annular electrodes or drift rings are concentrically arranged around a small collecting anode at the center. When a voltage is applied, the rings generate a potential gradient, which ensures that the generated charge carriers (electrons) drift towards the center. Due to the presence of the electric field, the charge carriers generated by the interaction of the X-rays with the silicon wafer drift towards the collecting anode. A field-effect transistor (FET) is coupled to the collecting anode at the center of the silicon wafer. Once the charge reaches the collecting anode, the FET amplifies the signal generated by the incident X-rays. After initial amplification, the signal undergoes further processing stages, during which it is shaped, amplified, and converted into a digital signal. The energy of each incident X-ray photon is highly proportional to the charge pulse generated by the detector, thereby enabling energy measurement.

[0042] In various embodiments, system 100 may be able to simultaneously provide output signals from first detector 114 and second detector 116 in response to the interaction of X-ray beam 118 with substrate pair 112.

[0043] As noted, for clarity, Figure 1 Various other elements and components for the X-ray measurement system 100 are omitted. For example, the system 100 can operate in a vacuum environment, such as in a vacuum chamber during semiconductor manufacturing. The X-ray source 110 may include various means for filtering or directing the X-ray beam 118 toward the substrate pair 112. In a particular embodiment, the X-ray beam 118 is generated using a copper (Cu) target, for example at energies of about 8 keV to 9 keV, and at other potential targets and energy bands in various embodiments. When the X-ray beam 118 is generated using a Cu target at about 8 keV to 9 keV, XRF can be performed through a Si substrate of about 50 nm to about 100 μm thickness in the substrate pair 112 to avoid excessive self-absorption of X-rays, such as for D2D and D2W bonding in a face-to-face or back-to-face arrangement, in which the first die to be bonded (e.g., the first substrate 112-1) can receive the incident X-ray beam 118 at the back or front surface.

[0044] Figure 2This is a depiction of an X-ray measurement system 200 (or simply system 200) using TL interferometry in some embodiments. For example... Figure 2 As shown, system 200 represents Figure 1 The partial configuration of the dual X-ray measurement system 100 depicts three gratings used for TL interferometry. A typical TL interferometer setup includes the three gratings shown in system 100: a source modulation grating G0, a beam splitter grating G1, and an analyzer grating G2. In some embodiments, when the X-ray beam 118 has a long coherence length, such as when the X-ray source 110 is a synchrotron, the source modulation grating G0 can be omitted. The source modulation grating G0 (also called an X-ray mask) can be placed close to the X-ray source. The source modulation grating G0 is designed to create a line source array by blocking portions of the X-ray beam, thereby effectively generating partially coherent radiation from an incoherent source, much like a conventional X-ray tube.

[0045] The beam splitter grating G1 (also called the phase grating G1) produces a periodic interference pattern that can oscillate with maximum intensity. The phase grating G1 is located downstream of the source modulation grating G0, for example, at a specific distance. The phase shift caused by the phase grating G1 results in an interference pattern known as the Talbot carpet at a distance, even in the absence of a sample between gratings G0 and G1. Therefore, the periodicity of the Talbot carpet can be a characteristic of system 200 itself, regardless of the sample used.

[0046] The third component in system 200 is the analyzer grating G2 (also called the absorption grating G2), which is positioned at one of the self-imaging planes of the Talbot carpet, typically corresponding to the Talbot fractional distance. The analyzer grating G2 has periodic absorption structures that can translate slight variations in the interference fringes into intensity variations at the first detector 114. The periodicity of the Talbot carpet in system 200 can be matched to the spacing of the analyzer grating G2 to optimize the sensitivity of sample displacement measurements (e.g., measurements of substrate misalignment at 112). In some embodiments, the analyzer grating G2 can be omitted, for example, when the first detector 114 has a spatial or pixel resolution significantly smaller than the interference fringes. Various types of X-ray detectors can be used as the first detector 114. In particular, semiconductor X-ray detectors can be used as the first detector 114, for example, by direct detection using a flat-panel X-ray detector with good spatial resolution and X-ray absorption characteristics (e.g., a semiconductor flat-panel imaging array that can generate image data from received X-rays).

[0047] exist Figure 2In the system 200 shown, the analyzer grating G2 converts the phase shift into an intensity change (amplitude modulation), making the phase shift detectable, for example, by a first detector 114. Therefore, the pattern at the plane of the first detector 114, carried by the X-ray beam 120, can include information about the sample (e.g., in...) Figure 2 The diagram shows information on absorption, phase shift, and small-angle scattering caused by the substrate pair 112 (first substrate 112-1 and second substrate 112-2). When the sample (substrate pair 112) is placed between the phase grating G1 and the analyzer grating G2, the interaction between the X-rays and the sample alters the interference pattern due to the modulation of the phase of the incident X-ray beam 118. Local changes in absorption, phase, and small-angle scattering result in corresponding modifications to the intensity pattern in the X-ray beam 120, which are measured by the first detector 114 located behind the analyzer grating G2. In practice, a series of images can be captured while one of the phase grating G1 or the analyzer grating G2 is laterally displaced, for example, across one or several cycles. In addition to standard transmission images, this step-scan process also allows for the reconstruction of differential phase-contrast images and dark-field images.

[0048] In operation, system 200 can employ TL interferometry and can be targeted at objects undergoing transmission analysis, such as substrate pair 112. For example, first detector 114 can be used to detect transmission signals of alignment marks located on one or more surfaces of first substrate 112-1 and second substrate 112-2 (see also...). Figures 5A to 5C Furthermore, when the first substrate 112-1 and the second substrate 112-2 also include interference alignment marks (such as moiré interference patterns) that interact with the X-ray beam 118, the first detector 114 can be used to detect interference patterns associated with misalignment of the moiré interference pattern. In a particular embodiment, the displacement of the moiré interference pattern detectable by the first detector 114 can be directly linear with the misalignment of the first substrate 112-1 relative to the second substrate 112-2, as will be described in further detail. This linear relationship can represent the effective magnification that the system 200 can obtain from the moiré interference pattern to increase sensitivity to actual misalignment of the substrate pair 112. The increased sensitivity to actual misalignment can improve the accuracy of subsequent alignment of the substrate pair 112.

[0049] Furthermore, the moiré interference pattern forming the moiré interference grating pair can have a first grating orientation that can be aligned with a second grating orientation of the beam splitter grating G1 and the analyzer grating G2 to improve sensitivity or achieve maximum sensitivity for detecting displacement of the moiré interference pattern relative to each other (e.g., detected misalignment). To detect and measure misalignment along different axes of the substrate pair 112, the substrate pair 112 can be rotated by an appropriate angle corresponding to the grating orientation of different sets of moiré interference grating pairs formed in the first substrate 112-1 and the second substrate 112-2, such as rotations of 45°, 90°, 135°, 180°, 225°, and 315° in various embodiments (see also...). Figures 6A to 6C , Figures 8A to 8C and Figure 9 ).

[0050] In a particular embodiment, the X-ray beam 118 may have sufficient energy to penetrate a thick Si substrate (including a highly doped Si substrate) to perform TL interferometry using system 200. Accordingly, system 200 can be used to measure substrate misalignment 112 using X-ray TL interferometry in various applications, such as for D2D, D2W, and W2W bonding. Furthermore, as... Figure 2 As shown, when the substrate pair 112 includes a thick or highly doped Si substrate, the X-ray beam 118 can use TL interferometry to measure the misalignment of the substrate pair 112, allowing measurement of various relative semiconductor surface orientations of the substrate pair 112, including face-to-face, back-to-face, face-to-back, and back-to-back arrangements.

[0051] Figure 3 This is a depiction of a composite alignment mark 300 for a semiconductor substrate in some embodiments. The composite alignment mark 300 comprises different individual alignment marks, and is shown from a top view, illustrating the general location and shape of the different alignment marks, which will be described in further detail. In particular, the composite alignment mark 300 may be carefully formed correspondingly on both the first and second substrates of the substrate pair 112 to provide alignment using both TL interferometry and XRF methods, as disclosed herein. Figure 3 The document also shows a coordinate diagram 302, which illustrates that the z-axis emerges from the page, and the x and y axes lie in the plane of the page; this diagram is used for reference herein. In various embodiments, complementary alignment mark sets may be located at or near the top surface (e.g., facet) of the first substrate and at or near the top surface (e.g., facet) of the second substrate.

[0052] As shown and described in the following figures, various alignment marks can be made of the metals disclosed herein in the X-ray methods and systems for semiconductor substrate alignment. In a particular embodiment, the alignment marks disclosed herein can be made of copper (Cu) and can be formed at specific locations on the semiconductor substrate. In various embodiments, the alignment marks can be made of another suitable material, such as another metal selected from nickel (Ni), tungsten (W), cobalt (Co), chromium (Cr), ruthenium (Ru), molybdenum (Mo), or various combinations or alloys of these metals.

[0053] Furthermore, the alignment marks in the composite alignment mark 300 can have various sizes in the semiconductor substrate. For example, the thickness of the composite alignment mark 300 can be from about 100 nm to about 20 μm. In some embodiments, the width of the alignment mark can be from about 100 nm to about 20 μm. When the alignment mark is a periodic structure (such as a moiré-fringe alignment mark), the spacing between the alignment marks can be from about 200 µm to about 40 μm. In certain embodiments, the alignment marks in the composite alignment mark 300 can be formed in prior process steps of semiconductor manufacturing, such as by deposition and photolithography, and other processes.

[0054] like Figure 3 As shown, the composite alignment mark 300 can correspondingly represent a region on a semiconductor substrate that can be used in X-ray methods and systems for semiconductor substrate alignment, as disclosed herein. Because the composite alignment mark 300 includes different types of alignment marks that can be used for different types of alignment techniques, as will be described in further detail, it is desirable that different alignment techniques can be performed without moving the substrate or adjusting the X-ray system (e.g., a single FOV). In other words, the composite alignment mark 300 can provide the ability to quickly and accurately determine the misalignment of the substrate pair 112 using different techniques over a wide range of misalignments and with varying degrees of accuracy. For example, within the same FOV of the incident X-rays, using the same equipment and configuration, the composite alignment mark 300 can be used for coarse alignment (lower spatial resolution) and then for fine alignment (higher spatial resolution), as is desired. Furthermore, the composite alignment mark 300 can include multiple or redundant alignment marks to improve accuracy, precision, and overall reliability, such as by increasing the sample size of the alignment marks being measured, or, in certain embodiments, when some individual alignment marks are defective or inoperable among other operable alignment marks.

[0055] like Figure 3As shown, the composite alignment mark 300 is an exemplary layout of individual alignment marks that can be rearranged or reconfigured in different embodiments. As illustrated, the composite alignment mark 300 includes XRF marks 310, which may be square and located at the corners of the composite alignment mark 300. The XRF marks 310 may be located on the top surface of the substrate (e.g., a facet), or may be covered by a silicon (Si) top layer with a thickness ranging from about 50 nm to 100 μm, which still provides a sufficient amount of fluorescent X-ray photons without excessive self-absorption by the silicon (Si) top layer. In various embodiments, the XRF marks 310 may be used for D2D or D2W bonding, as described above, to measure a first coarse misalignment and subsequently perform a first coarse alignment of the first substrate 112-1 with the second substrate 112-2. Using XRF marker 310 to perform a first coarse alignment of the first substrate 112-1 with the second substrate 112-2 allows for subsequent coarse or fine alignment with other composite alignment markers in the composite alignment marker 300, such as allowing further measurement of remaining fine misalignment by reducing the misalignment to a lower value, as described below.

[0056] As shown in the figure, the composite alignment mark 300 also includes first TL marks 312, which can be used with TL interferometry, for example, by using X-ray measurement system 200 (see Figure 200). Figure 2 The second coarse misalignment is measured using XRF marker 310. In a particular embodiment, the second coarse misalignment may be smaller than the first coarse misalignment performed using XRF marker 310. As shown, the composite alignment marker 300 also includes second TL markers 314, which may be square regions comprising moiré grating elements for moiré fringe-based interferometric alignment. In a particular embodiment described below, the first substrate 112-1 and the second substrate 112-2 may each include a composite alignment marker 300 having second TL markers 314, which comprise one of two complementary moiré patterns that together form a moiré interferometer grating pair when placed adjacent to or close to each other (e.g., in substrate pair 112). Furthermore, while XRF marker 310 and first TL marker 312 may be used with the X-ray beam 118 in different orientations, the moiré pattern located in the second TL marker 314 can advantageously be aligned with the beam splitter grating G1 such that the respective grating axes are parallel to each other to improve the signal-to-noise ratio and increase sensitivity. As will be described below, the second TL mark 314 can be used for a first fine alignment and a second fine alignment, which can have overlapping measurement ranges and different magnification levels, and thus can provide a wide detection range for misaligned distances or displacements.

[0057] Figure 4A , Figure 4B and Figure 4C The XRF misalignment detection process in some implementations is described. Figure 4A This is a depiction of the XRF alignment mark 400, which can represent... Figure 3 An example of XRF marker 310 is shown. A sampling line 410 is shown together with the XRF alignment marker 400, which indicates where XRF measurements of fluorescence X-rays are performed for detection. Figure 4B The alignment marks 401 and 402 shown herein generate Figure 4C The corresponding measurement signal is shown below. To obtain XRF measurements along the line profile of sampling line 410, substrate pair 112 can be moved relative to X-ray beam 118 along the Y-axis while recording XRF measurements. An SDD configured as a point detector can be used to record the XRF measurements. The following information is shown in XRF alignment mark 400. Figure 4B Fields A and B are described. Furthermore, when the substrate pair 112 is to be aligned along two axes (e.g., along the Y-axis and along the X-axis), sampling line 412 can be used along the X-axis, and alignment marks along sampling line 412 can be oriented for X-axis alignment.

[0058] Figure 4B Alignment marks 401 and 402 are depicted representing metal pads (such as copper (Cu) pads) that form XRF alignment marks 400 for a first substrate 112-1 and a second substrate 112-2 used to form a substrate pair 112. Dashed lines between the metal pads indicate the surface interface between the first substrate 112-1 and the second substrate 112-2. In alignment mark 401, for field A, three metal pads are located at the same positions in the first substrate 112-1 and the second substrate 112-2 and are shown as aligned. In alignment mark 401, for field B, the first substrate 112-1 is shown with the same alignment marks as in field A, while the second substrate 112-2 has two alignment marks offset from the alignment marks in substrate 112-1, but is also depicted as aligned. In alignment mark 402, the same metal pads as shown in alignment mark 401 are depicted, but slightly misaligned is shown for both field A and field B.

[0059] Figure 4C Curves 403 and Figure 404 depict the variation of XRF signal intensity with distance along sampling line 410. The XRF signal intensity on the Y-axis of curves 403 and 404 can correspond to the distance from the X-ray beam 118 during irradiation. Figure 4BThe intensity of the fluorescent X-rays detected by the emission from the metal pads shown. Each point in curves 403 and 404 corresponds to a point along sampling line 410, where the left portion of the curve corresponds to field A and the right portion of the curve corresponds to field B, as indicated.

[0060] Figure 4C Curve 403 in Figure 403 illustrates the XRF signal intensity of alignment marks 401 in an aligned state. In curve 403, the amplitude of the signal intensity is approximately twice that corresponding to alignment marks 401 in field A, which are located at the same positions in the first substrate 112-1 and the second substrate 112-2, thus resulting in approximately twice the fluorescence photon intensity. In curve 403, for field B, the constant amplitude corresponding to a single alignment mark corresponds to an offset alignment mark in the first substrate 112-1 and the second substrate 112-2.

[0061] Figure 4C Curve 404 illustrates the XRF signal intensity of a slightly misaligned alignment mark 402. In the left portion of curve 404, the double-amplitude XRF signal intensity with a double-step peak corresponds to a decreasing distance along sampling line 410, where alignment marks 402 overlap in field A; while the single-amplitude signal portion corresponds to a small shoulder, where alignment marks 402 in field A do not overlap in the first substrate 112-1 and the second substrate 112-2. In the right portion of curve 404, corresponding to alignment marks 402 in field B, the double amplitude corresponds to overlapping alignment marks 402 in the first substrate 112-1 and the second substrate 112-2, while the single amplitude corresponds to one alignment mark measured in either the first substrate 112-1 or the second substrate 112-2. Furthermore, a zero signal (shown as a minimum in curve 404) corresponds to no alignment mark measured from either the first substrate 112-1 or the second substrate 112-2 at the corresponding distance where a gap exists in the alignment marks 402 in field B.

[0062] In operation, misalignment of the first substrate 112-1 and the second substrate 112-2, also known as detected misalignment, can be detected when alignment mark 402 is observed based on the signal pattern in curve 404. A library of curves similar to 403 or 404 can be created to record different alignment positions and used to estimate detected misalignment. To perform alignment, the lateral position of the substrates in substrate pair 112 can be adjusted (e.g., along line 410) until the signal pattern in curve 403 is observed, indicating that alignment mark 401 is in an aligned state and that the first substrate 112-1 is aligned with the second substrate 112-2.

[0063] Figure 5A , Figure 5B and Figure 5CThe X-ray transmission misalignment detection process in some embodiments is described. Figure 5A Includes a depiction of TL alignment mark 500, which can represent Figure 3 The alignment state of the first TL mark 312 shown. Figure 5A It also includes a depiction of TL alignment mark 501, which can represent Figure 3 The misaligned state of the first TL mark 312 is shown. Each of the alignment marks 500 and 501 is in the form of a ring, with the outer ring located in the first substrate 112-1 and the inner ring located in the second substrate 112-2 (see also...). Figure 5B Sampling line 510 is shown together with TL alignment marks 500 and 501. This sampling line indicates where TL transmission measurements are performed through substrate pair 112 to detect... Figure 5B The alignment marks 502 and 503 shown herein generate Figure 5C The corresponding measurement signal is shown. To obtain the line profile TL transmission measurement value along the sampling line 510, the substrate pair 112 can be moved relative to the X-ray beam 118 along the Y-axis, while, for example, by using... Figure 2 The X-ray measurement system 200 records TL transmission measurements. In some embodiments, TL transmission measurements can be recorded from an image captured by the first detector 114. TL transmission measurements along the sampling line 510 can also be obtained by moving the analyzer grating G2 relative to the first detector 114 to perform interferometric lateral sampling.

[0064] Figure 5B Alignment marks 502 and 503 are depicted, which indicate alignment with... Figure 5A The cross-section of circular alignment marks 500 and 501 intersecting the sampling line 510. Alignment marks 502 and 503 are depicted as metal pads, such as copper (Cu) pads, which form TL alignment marks 500 and 501 respectively for the first substrate 112-1 and the second substrate 112-2 forming the substrate pair 112. Figure 5B The dashed lines between the metal pads indicate the surface interface between the first substrate 112-1 and the second substrate 112-2. In alignment marks 502 and 503, corresponding to... Figure 5A Two metal pads are shown above and below the surface interface of the outer and inner ring sections.

[0065] Figure 5C Curve 504 in the figure shows the TL signal strength of the alignment mark 502 in the alignment state. In curve 504, the signal strength shows the relationship with... Figure 5AThe alignment marks 502 of the inner and outer rings of the alignment are shown as single amplitude values. In curve 504, the alignment marks in the first substrate 112-1 and the second substrate 112-2 do not overlap, thereby producing a uniform maximum intensity of fluorescent X-ray photons, and the uniformly spaced gaps between the alignment marks 502 are shown, which correspond to the minimum signal intensity.

[0066] Figure 5C Curve 505 in the figure shows the TL signal strength of a slightly misaligned alignment mark 503. In curve 505, the signal strength is shown in relation to... Figure 5A The misaligned inner and outer rings of the alignment marks 503 correspond to single and double amplitude values. In curve 505, the alignment marks 503 in the first substrate 112-1 and the second substrate 112-2 do indeed slightly overlap, resulting in double the TL signal intensity at this distance along the sampling line 510. Curve 505 also shows the unevenly spaced gaps between the alignment marks 503, which correspond to the minimum intensity.

[0067] In operation, when alignment marks 501 and 503 are observed based on the signal pattern in curve 505, a misalignment of the first substrate relative to the second substrate can be detected, also referred to as a detected misalignment. A library of reference curves similar to curves 504 or 505 can be created to record different alignment positions and to estimate the detected misalignment. For example, the stored library of reference curves can be indexed to calibrated misalignment values ​​to match the observed signal intensity map 504 or 505 to the detected misalignment of alignment marks 502 or 503. In various embodiments, different methods can be used to generate curves 504 or 505. In one embodiment, one of the phase grating G1 or the analyzer grating G2 can be moved to detect the signal intensity from X-rays received at the first detector 114, such as when the first detector 114 is an SDD or other small-area X-ray detector (e.g., a beam detector). The movement of the phase grating G1 or the analyzer grating G2 can effectively generate a line scan that produces a signal intensity curve 504 or 505 that can be captured and recorded by the SDD. In another embodiment, when the first detector 114 is a flat-panel X-ray detector that outputs image data, curves 504 or 505 can be distinguished or generated from the image data. In a particular embodiment, instead of storing a library of reference curves 504 or 505, a stored reference image database can be used to match observed image data with detected misalignments of alignment marks 502 or 503. It should be noted that the reference image data can also be used with alignment marks 600 and 800, including moiré fringe elements, as described in further detail below, to match observed image data to detected misalignments of moiré interference patterns (see [link to relevant documentation]). Figure 7A and Figure 7BTo perform alignment, the lateral position of the substrates in substrate pair 112 can be adjusted until a signal pattern in curve 504 is observed, indicating that alignment marks 500 and 502 are aligned and that the first substrate 112-1 is aligned with the second substrate 112-2.

[0068] Figure 6A , Figure 6B and Figure 6C Moiré-fringe alignment marks for semiconductor substrates are depicted in some embodiments. Figure 6A Alignment marks 600 are depicted, and these alignment marks can be Figure 3 An example of the second TL mark 314. Specifically, the alignment mark 600 is shown as being in a square field divided into two rectangles, which correspond to the positions of complementary moiré grating elements. and like Figure 6B and Figure 6C As shown. In alignment mark 600, two complementary moiré grating elements. and Shown as corresponding to a first substrate 112-1 and a second substrate 112-2 in substrate pair 112, a moiré interference grating pair is formed when the first substrate and the second substrate are placed close together. When the first substrate 112-1 and the second substrate 112-2 are close to each other, for example in substrate pair 112, each moiré grating element (e.g., single row) or ) and complementary moiré grating elements (e.g., single row) or The elements are paired to form a moiré interference grating by superposition. Since each layer corresponding to the first substrate 112-1 and the second substrate 112-2 has two moiré grating elements, the alignment marks 600 form a moiré interference grating pair.

[0069] exist Figure 6B In the diagram, a pair of single-row moiré grating elements 601 corresponding to the first substrate 112-1 are shown oriented along the X-axis on the left and along the Y-axis on the right. Figure 6C In the diagram, a pair of single-row moiré grating elements 602 corresponding to the second substrate 112-2 are shown oriented along the X-axis on the left and along the Y-axis on the right. Each single-row moiré grating element 601 in the first substrate 112-1 is paired with a complementary single-row moiré grating element 602 in the second substrate 112-1 to form a moiré interference grating. For the double-row moiré grating elements 601, a spacing is provided. and The elements are paired together so that the two rows form a moiré interference grating pair. The spacing relationship is given by Equation 1, in which... This represents the spacing difference. Then, the resulting moiré interference grating can be defined by the moiré period given in Equation 2 and the moiré magnification given in Equation 3. From the moiré pattern generated from the moiré interference gratings corresponding to the pair, the misalignment direction is orthogonal to the orientation of the moiré interference gratings, while (e.g., inverted in the grating pair) each complementary moiré interference grating causes a misalignment offset in the opposite direction of the moiré pattern (see also...). Figure 7A and Figure 7B Although the moiré grating elements 601 and 602 are shown as rotated 90° to correspond to the X and Y axes, different orientations, such as 45° or 135°, can be used in different embodiments. Figure 6B and Figure 6C The moiré interference grating element oriented along the X-axis is labeled as and The moiré interference grating element oriented along the Y-axis is labeled as and .

[0070] In a particular embodiment, alignment marks 600 in the form of a moiré interference grating pair can be used in conjunction with the X-ray measurement system 200 for a TL transmission arrangement having a substrate pair 112 (see [link]). Figure 2 In this context, because the measurement sensitivity is maximized when the moiré interference grating pair is aligned with the orientation of the beam splitter grating G1 and / or the analyzer grating G2, in various embodiments, it can be used by rotating the substrate pair 112 relative to the X-ray beam 118 by 90° in system 200. Figure 6B and Figure 6C The orthogonal grating is shown. In this way, for example using system 200, the substrate pair 112 can be aligned along both the X and Y axes (see also...). Figure 9 ).

[0071] Figure 7A and Figure 7B Moiré-fringe X-ray transmission alignment images are depicted in some embodiments. Figure 7A In the image, the moiré-fringe X-ray transmission alignment shows a pair of interference patterns corresponding to the moiré interference grating pair, such as those mentioned above. Figures 6A to 6C As shown. Specifically, in Figure 7A The image shown is a transmission image 700 and a dark field image 701, while... Figure 7B The image shown is a phase-contrast image 702.

[0072] Transmission image 700, dark field image 701, and phase contrast image 702 can be used Figure 2 The X-ray measurement system 200 uses a first detector 114 to capture X-ray beams 118 that irradiate a pair of moiré interference gratings formed in a pair of substrates 112. In each of the transmission image 700, dark-field image 701, and phase-contrast image 702, a pair of moiré interference patterns (corresponding to the moiré interference grating pair) are shown for four (4) reference misalignment values ​​of the first substrate 112-1 relative to the second substrate 112-2: moiré interference pattern 710 shows alignment or zero (0) misalignment; moiré interference pattern 712 shows 125 nm misalignment; moiré interference pattern 714 shows 250 nm misalignment; and moiré interference pattern 716 shows 375 nm misalignment. As noted, each moiré interference pattern in the pair (generated by the complementary moiré interference grating pair) results in a translation in a direction opposite to the centerline 720. It can be seen that the corresponding translation in the interference pattern pair (e.g., detected misalignment) increases with the actual misalignment of the substrate pair 112. This corresponding translation can be recorded and analyzed, for example, using digital image processing, to determine the actual misalignment. Specifically, in Figure 7B In the phase-contrast image 702, the human eye can clearly distinguish the displacement of each interference pattern from the center line 720. In operation, any or all of the transmission image 700, dark-field image 701, and phase-contrast image 702 can be used to determine the actual misalignment, depending on the material properties of the moiré interference grating pair used, as well as other factors and variables.

[0073] In operation, when the moiré interference pattern 716 is observed, a misalignment (e.g., of substrate pair 112) of the first substrate 112-1 relative to the second substrate 112-2 can be detected, also referred to as a detected misalignment. In various embodiments, the actual misalignment of the substrate pair 112 is measured as follows: the actual misalignment is calculated using a linear relationship based on the detected misalignment (see also...). Figure 7C To perform alignment, the lateral positions of the substrates in substrate pair 112 can be displaced relative to each other until a moiré image 710 is observed, thereby indicating that the first substrate 112-1 is aligned with the second substrate 112-2.

[0074] Figure 7CThis is calibration curve 703 for the moiré interference pattern in some embodiments. In calibration curve 703, a positive linear relationship is determined using linear regression corresponding to the misalignment values ​​of moiré images 710, 712, 714, and 716, representing the displacement of the moiré interference pattern indicating detected misalignment, and the actual misalignment of the first substrate 112-1 relative to the second substrate 112-2. The detected misalignment obtained from the moiré interference pattern displacement is represented as the sum of the misalignments of the two patterns in a pair. Calibration curve 703 shows that the detected misalignment amplifies the actual misalignment with a magnification factor C, which is expressed as a slope value of approximately 27. This indicates that moiré-fringe TL X-ray transmission measurements, as described herein, provide increased sensitivity for measuring the actual misalignment of the substrate relative to 112, which is desirable.

[0075] Figure 8A , Figure 8B and Figure 8C Moiré-fringe alignment marks for semiconductor substrates are depicted in some embodiments. Figure 8A Alignment marks 800 are depicted; these alignment marks can be Figure 3 An example of the second TL mark 314. Specifically, alignment mark 800 is shown as a square field divided into four rectangles, which correspond to the positions of two moiré interference grating pairs. The moiré grating elements in alignment mark 800 are marked for high magnification. and And for low magnification, it is marked as and ,like Figure 8B and Figure 8C As shown. In alignment mark 800, two layers of complementary moiré grating elements are shown corresponding to the first substrate 112-1 and the second substrate 112-2 in substrate pair 112, which together form two moiré interference grating pairs, one for high magnification ( / ), and one for low magnification ( / When the first substrate 112-1 and the second substrate 112-2 are close to each other, for example in substrate pair 112, two moiré grating elements (e.g., two rows of marks) are paired with two complementary moiré grating elements to form a pair of moiré interference gratings by superposition. The two moiré grating elements including the two rows of complementary marks can be collectively referred to as an alignment mark of each of the first substrate 112-1 and the second substrate 112-2, which together form a pair of moiré interference gratings.

[0076] In certain embodiments, each field or moiré grating element in the alignment mark 800 can be simultaneously irradiated by an incident X-ray beam (such as X-ray beam 118 in system 200), such that the overall size of the alignment mark 800 can be smaller than or similar to the cross-sectional area of ​​the incident X-ray beam. In some embodiments, this allows for the performance of various misalignment measurements described below without readjusting or reconfiguring the X-ray beam, which is desirable.

[0077] exist Figure 8B In the diagram, two pairs of single-row moiré grating elements 801 corresponding to the first substrate 112-1 are shown oriented along the X-axis on the left and along the Y-axis on the right. Figure 8C In the diagram, two pairs of single-row moiré grating elements 802 corresponding to the second substrate 112-2 are shown oriented along the X-axis on the left and along the Y-axis on the right. Each moiré grating element 801 in the first substrate 112-1 is paired with a complementary moiré grating element 802 in the second substrate 112-1. For the moiré interference grating pairs formed using moiré grating elements 801 and 802, a spacing is specified. and The grating elements are paired together to achieve high magnification, while having a spacing and The grating elements are paired together to obtain a low magnification, such that the spacing relationship is given by Equation 1.

[0078] Figure 8D This is calibration curve 803 of the moiré interference pattern in some embodiments. Calibration curve 803 is similar to... Figure 7C Calibration curve 703 illustrates the actual misalignment of the first substrate 112-1 relative to the second substrate 112-2 in the substrate pair 112, based on the detected misalignment using alignment mark 800. Calibration curve 803 shows two distinct linear curves corresponding to the moiré interference grating pair formed by alignment mark 800, as described above. The slope of the high magnification curve is represented by A (e.g., for...). and ), while the slope of the low magnification curve is represented by B (e.g., for and As shown in calibration curve 803, the amplification factor A is approximately 57, while the amplification factor B is approximately 41, such that B < A. In various embodiments, for specific orientations and configurations, pairs of moiré interference gratings can be formed on various substrates, which can be used to achieve amplification factors of up to 100 or greater in this manner.

[0079] More generally, for grating elements with corresponding spacing and For each pair of moiré interference gratings, the lower detection limit (LOD) is given by Equation 4, and the upper detection limit (LOD) is given by Equation 5. In Equation 4, the talbot fringe period corresponds to the spacing of the analyzer grating G2, which can match the periodicity of the talbot carpet of the TL measurement system (e.g., System 200). High-magnification moiré interference grating pairs are used to minimize the lower limit LOD, while low-magnification moiré interference grating pairs are used to maximize the upper limit LOD. Furthermore, the lower limit LOD at low magnification is smaller than the upper limit LOD at high magnification to provide continuous coverage of actual misalignment without gaps. Therefore, the total misalignment range that can be measured using alignment marks 800 can extend from the lower limit LOD at high magnification to the upper limit LOD at low magnification, and can be detected and measured in a single FOV using System 200, which is desirable.

[0080] Figure 9 This is a flowchart depicting a method 900 for aligning two semiconductor substrates using an X-ray alignment method and system in some embodiments. It should be noted that in various embodiments, certain operations in method 900 may be rearranged or omitted. In certain embodiments, method 900 may be performed using X-ray measurement systems 100 and / or 200.

[0081] Method 900 may begin at step 902 by positioning a first substrate and a second substrate close to each other in a pre-bonded arrangement aligned with the X-ray beam along a first axis. The pre-bonded arrangement may involve positioning the first substrate and the second substrate close to each other, for example, as substrate pair 112. At step 904, coarse alignment is performed using first alignment marks on the first and second substrates. In some embodiments, the coarse alignment in step 904 may include measuring a first coarse misalignment using XRF, for example for D2D or D2W bonding, such that the first alignment mark includes XRF mark 310. The coarse alignment in step 904 may further include measuring a second coarse misalignment using a TL method, such that the first alignment mark includes a first TL mark 312. At step 906, it is determined whether the coarse alignment is within tolerance. If the result of step 906 is "no", method 900 loops back to step 904. If the result of step 906 is "yes", at step 908, a first fine alignment is performed using second alignment marks on the first and second substrates. The second alignment mark in step 908 may be a second TL mark 314 representing alignment mark 800 (see...) Figure 3 , Figures 8A to 8C This allows the first fine alignment to be performed at a low magnification (e.g., using...). and ).

[0082] In method 900, at step 910, it is determined whether the first fine alignment is within tolerance. If the result of step 910 is "no", method 900 loops back to step 908. If the result of step 910 is "yes", at step 912, a second fine alignment is performed using a third alignment mark on the first substrate and the second substrate. The third alignment mark in step 912 may be a second TL mark 314 representing alignment mark 800 (see...). Figure 3 , Figure 8A -C), allowing the second fine alignment to be performed at high magnification (e.g., using...). and In a particular embodiment, a system 200 with the same FOV as system 200 can be used, and steps 908 and 912 can be performed using the same X-ray beam. At step 914, it is determined whether the second fine alignment is within tolerance. If the result of step 914 is "No", method 900 returns to step 912. If the result of step 914 is "Yes", at step 916, it is determined whether both the X-axis and Y-axis are within tolerance. The X-axis and Y-axis in step 916 refer to the X-axis and Y-axis in the plane relative to the first and second substrates. If the result of step 916 is "Yes", method 900 ends at step 920. If the result of step 916 is "No", at step 918, the first and second substrates are rotated 90° relative to the X-ray beam in the pre-bonding arrangement to align along a second axis perpendicular to the first axis. After step 918, method 900 returns to step 908.

[0083] Figure 10A and Figure 10B This is a depiction of multilayer moiré-fringe alignment marks in some embodiments. Figure 10A and Figure 10B The alignment marks indicating the position of the moiré interference grating pair can correspond to alignment marks 600 or 800, as described above. Figure 10A A side view 1000 of multiple layers along the Z-axis up to any number of layers N is shown. For each pair of adjacent layers, the moiré interference grating pairs can be placed at different locations in a spatially separated manner to achieve... Figure 10A Alignment is provided between the corresponding two layers indicated in the diagram. Multiple layers can represent multiple semiconductor substrates used for multilayer 3D integration and bonding. Alignment markings for layer N relative to layer 1 are also provided. Figure 10B The layout arrangement 1001 of each alignment pair is shown in top view. In one embodiment, the arrangement of corresponding moiré interference grating pairs in the XY plane with arbitrary orientation is shown as viewed from side view 1000.

[0084] Figure 11 This is a flowchart depicting a method 1100 for aligning two semiconductor substrates using an X-ray alignment method and system in some embodiments. It should be noted that in various embodiments, certain operations in method 1100 may be rearranged or omitted. In certain embodiments, method 1100 may be performed using X-ray measurement systems 100 and / or 200.

[0085] Method 1100 may begin at step 1102 by directing X-rays to a first substrate and a second substrate that are bonded but not yet bonded together, including directing the X-rays to a first alignment mark and a third alignment mark in the first substrate, and to a second alignment mark and a fourth alignment mark in the second substrate. At step 1104, fluorescent X-rays emitted from the first alignment mark and from the second alignment mark are detected to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first alignment mark relative to the second alignment mark. At step 1106, at least some of the X-rays transmitted through the first substrate and through the second substrate are detected using X-ray Talbot-Laur interferometry to measure a second misalignment of the first substrate relative to the second substrate based on a second detected misalignment of the third alignment mark relative to the fourth alignment mark.

[0086] Figure 12 This is a flowchart depicting a method 1200 for aligning two semiconductor substrates using an X-ray alignment method and system in some embodiments. It should be noted that in various embodiments, certain operations in method 1200 may be rearranged or omitted. In certain embodiments, method 1200 may be performed using X-ray measurement systems 100 and / or 200.

[0087] Method 1200 may begin at step 1202 by directing X-rays to a first substrate and a second substrate that are bonded but not yet bonded together, including directing the X-rays to a first alignment mark in the first substrate and to a second alignment mark in the second substrate. At step 1204, in response to X-ray irradiation of the first and second alignment marks, fluorescent X-rays emitted from the first and second substrates are detected. At step 1206, a first misalignment of the first alignment mark relative to the second alignment mark is measured, at least in part, based on the wavelength of the fluorescent X-rays corresponding to the metals in the first and second alignment marks.

[0088] Figure 13This is a flowchart depicting a method 1300 for aligning two semiconductor substrates using an X-ray alignment method and system in some embodiments. It should be noted that in various embodiments, certain operations in method 1300 may be rearranged or omitted. In certain embodiments, method 1300 may be performed using X-ray measurement systems 100 and / or 200.

[0089] Method 1300 may begin at step 1302, where X-rays are directed to a first substrate and a second substrate that are bonded together for subsequent bonding. At step 1304, X-rays are transmitted using an X-ray Talbot-Laur interferometer through a first alignment mark in the first substrate and through a second alignment mark in the second substrate, wherein the first and second alignment marks comprise a first pair of moiré interference gratings. At step 1306, a first misalignment of the first substrate relative to the second substrate is measured based on a first detected misalignment of the first alignment mark relative to the second alignment mark, using a first interference pattern and a second interference pattern associated with the first moiré interference grating pair. This detection of the first detected misalignment includes measuring a first displacement of the first interference pattern in a first direction and a second displacement of the second interference pattern in a second direction opposite to the first direction.

[0090] At step 1308, X-rays are transmitted through a third alignment mark in the first substrate and a fourth alignment mark in the second substrate using an X-ray Talbot-Laur interferometer, wherein the third and fourth alignment marks comprise a pair of second moiré interference gratings. At step 1310, a second misalignment of the first substrate relative to the second substrate is measured based on a second detected misalignment of the third alignment mark relative to the fourth alignment mark, measured using the third and fourth interference patterns associated with the second moiré interference grating pair. This measurement of the second detected misalignment includes measuring a third displacement of the third interference pattern in a first direction and a fourth displacement of the fourth interference pattern in a second direction.

[0091] As disclosed herein, in one embodiment, X-rays are directed to a first substrate and a second substrate configured for bonding together. The X-rays are directed to first and third alignment marks in the first substrate, and to second and fourth alignment marks in the second substrate. Fluorescent X-rays are detected as they are emitted from the first and second alignment marks to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment at the first and second alignment marks. X-rays are transmitted through the first and second substrates using X-ray Talbot-Laur interferometry to measure a second misalignment of the first and second substrates based on a second detected misalignment at the third and fourth alignment marks.

[0092] Example 1. A method for measuring misalignment between substrates, the method comprising: directing X-rays to a first substrate and a second substrate in a bonded configuration but not yet bonded together, the directing comprising directing the X-rays to a first alignment mark and a third alignment mark in the first substrate, and to a second alignment mark and a fourth alignment mark in the second substrate; detecting fluorescent X-rays emitted from the first alignment mark and from the second alignment mark to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first alignment mark relative to the second alignment mark; and detecting at least some of the X-rays transmitted through the first substrate and through the second substrate using X-ray Talbot-Laur interferometry to measure a second misalignment of the first substrate relative to the second substrate based on a second detected misalignment of the third alignment mark relative to the fourth alignment mark.

[0093] Example 2. The method as described in Example 1, wherein the first alignment mark, the second alignment mark, the third alignment mark, and the fourth alignment mark comprise metal.

[0094] Example 3. The method as described in either Example 1 or 2, wherein the metal comprises at least one of copper (Cu), tungsten (W), cobalt (Co), ruthenium (Ru), or molybdenum (Mo).

[0095] Example 4. The method as described in any one of Examples 1 to 3, wherein the first detected misalignment is detected by a first detector sensitive to these fluorescent X-rays to measure the intensity of these fluorescent X-rays in relation to the thickness of the metal, and wherein the second detected misalignment is detected by a second detector sensitive to these X-rays to measure the intensity of these X-rays transmitted relative to the TL interference pattern at the second detector.

[0096] Example 5. The method as described in any one of Examples 1 to 4, wherein the first alignment mark and the third alignment mark are located on the top surface of the first substrate, and the second alignment mark and the third alignment mark are located on the top surface of the second substrate.

[0097] Example 6. The method as described in any one of Examples 1 to 5, wherein the top surface of the first substrate faces the top surface of the second substrate, or wherein the top surface of the first substrate faces X-rays directed to the first substrate.

[0098] Example 7. A method for measuring misalignment between substrates, the method comprising: directing X-rays to a first substrate and a second substrate in a bonded configuration but not yet bonded together, the directing comprising directing the X-rays to a first alignment mark in the first substrate and to a second alignment mark in the second substrate; detecting fluorescent X-rays emitted from the first substrate and the second substrate in response to the X-rays irradiating the first alignment mark and the second alignment mark; and measuring a first misalignment of the first alignment mark relative to the second alignment mark based at least in part on wavelengths of the fluorescent X-rays corresponding to metals in the first alignment mark and the second alignment mark.

[0099] Example 8. The method of Example 7 further includes: distinguishing the wavelength based on the metal to measure the first misalignment, wherein the metal includes at least one of copper (Cu), tungsten (W), cobalt (Co), ruthenium (Ru), or molybdenum (Mo).

[0100] Example 9. The method as described in either Example 7 or 8, wherein the first misalignment is detected by a first detector that is sensitive to these fluorescent X-rays to measure the intensity of these fluorescent X-rays in relation to the metal thickness.

[0101] Example 10. The method as described in any one of Examples 7 to 9, wherein the first detector is a silicon drift detector (SDD).

[0102] Example 11. The method as described in any one of Examples 7 to 10, wherein the first alignment mark and the second alignment mark comprise a common material.

[0103] Example 12. A method for measuring misalignment between substrates, the method comprising: directing X-rays to a first substrate and a second substrate bonded together for subsequent bonding; transmitting the X-rays through a first alignment mark in the first substrate and through a second alignment mark in the second substrate using an X-ray Talbot-Laur interferometer, wherein the first alignment mark and the second alignment mark comprise a first pair of moiré interference gratings; and measuring a first misalignment of the first substrate relative to the second substrate based on detecting a first detected misalignment of the first alignment mark relative to the second alignment mark using a first interference pattern and a second interference pattern associated with the first moiré interference grating pair, the detection of the first detected misalignment comprising measuring a first displacement of the first interference pattern in a first direction and a second displacement of the second interference pattern in a second direction opposite to the first direction.

[0104] Example 13. The method as described in Example 12, wherein the first detected misalignment is linearly correlated with a first sum of the absolute values ​​of the first displacement and the second displacement, and wherein the first misalignment is at least 10 times smaller than the first sum.

[0105] Example 14. The method as described in one of Examples 12 or 13, wherein the first moiré interference grating is aligned with the beam splitter grating of the X-ray Talbot-Lau interferometer.

[0106] Example 15. The method of any one of Examples 12 to 14, further comprising: transmitting the X-rays through a third alignment mark in the first substrate and a fourth alignment mark in the second substrate using an X-ray Talbot-Laur interferometer, wherein the third alignment mark and the fourth alignment mark comprise a second moiré interference grating pair; and measuring a second misalignment of the first substrate relative to the second substrate based on measuring a second detected misalignment of the third alignment mark relative to the fourth alignment mark using a third interference pattern and a fourth interference pattern associated with the second moiré interference grating pair, the measurement of the second detected misalignment comprising measuring a third displacement of the third interference pattern in the first direction and a fourth displacement of the fourth interference pattern in the second direction.

[0107] Example 16. The method as described in any one of Examples 12 to 15, wherein the second misalignment is linearly related to a second sum of the absolute values ​​of the third and fourth displacements, and wherein the second misalignment is at least 50 times smaller than the second sum.

[0108] Example 17. The method as described in any one of Examples 12 to 16, wherein the first detection lower limit of the first moiré interference grating pair is equal to the Talbot fringe period of the analyzer grating of the X-ray Talbot-Laur interferometer divided by the first moiré amplification factor of the first moiré interference grating pair, and wherein the first detection upper limit of the first moiré interference grating pair is equal to ,in, It is the first spacing of the first alignment mark, and It is the second spacing of the second alignment mark.

[0109] Example 18. The method as described in any one of Examples 12 to 17, wherein the second detection lower limit of the second moiré interference grating pair is equal to the Talbot fringe period divided by the second moiré amplification factor of the second moiré interference grating pair, and wherein the second detection upper limit of the second moiré interference grating pair is equal to ,in, It is the third spacing of the third alignment mark, and It is the fourth spacing of the fourth alignment mark.

[0110] Example 19. The method as described in any one of Examples 12 to 18, wherein the first detection upper limit is greater than the second detection lower limit.

[0111] Example 20. The method as described in any one of Examples 12 to 19, wherein the X-rays are simultaneously transmitted through the first pair of moiré interference gratings and the second pair of moiré interference gratings.

[0112] Example 21. A method for measuring misalignment between semiconductor substrates, the method comprising: transmitting first X-rays through a first substrate and a second substrate bonded together for subsequent bonding with each other using an X-ray Talbot-Lau (TL) interferometer, including transmitting the first X-rays through a first alignment mark in the first substrate and a second alignment mark in the second substrate to generate second X-rays; transmitting the second X-rays through a TL phase grating and through a TL analyzer grating to generate third X-rays; and receiving the third X-rays at an X-ray detector to measure a first misalignment of the first substrate relative to the second substrate based on detecting a first misalignment of the first alignment mark relative to the second alignment mark using the X-ray detector.

[0113] Example 22. The method of Example 21, wherein receiving the third X-rays at the X-ray detector further comprises: performing a line scan by moving one of the TL phase grating or the TL analyzer grating on the first alignment mark and the second alignment mark; recording a line scan signal from the output of the X-ray detector during the line scan, while the X-ray detector receives the third X-rays; and detecting the first detected misalignment based on the line scan signal.

[0114] Example 23. The method as described in either Example 21 or 22, wherein detecting the first detected misalignment based on the line scan signal further comprises: using a first reference line scan signal repository indexed to a calibrated misalignment value to match the line scan signal to the first detected misalignment.

[0115] Example 24. The method as described in any one of Examples 21 to 23, wherein the X-ray detector includes a silicon drift detector (SDD).

[0116] Example 25. The method of any one of Examples 21 to 24, wherein receiving the third X-rays at the X-ray detector further comprises: generating image data of the first alignment mark and the second alignment mark using the X-ray detector, wherein the X-ray detector is a flat panel X-ray image detector; and detecting the first detected misalignment based on the image data, including using a second reference image data repository indexed to calibrated misalignment values ​​to match the image data to the first detected misalignment.

[0117] Example 26. The method as described in any one of Examples 21 to 25, wherein the first alignment mark includes a first moiré grating element and the second alignment mark includes a second moiré grating element, wherein the first moiré grating element and the second moiré grating element together form a moiré interference grating pair, and wherein the second X-rays include a moiré interference pattern.

[0118] Although this disclosure has been described with reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. Various modifications and combinations of the illustrative embodiments and other embodiments will be apparent to those skilled in the art upon reference to this specification. Therefore, the appended claims are intended to cover any such modifications or embodiments.

Claims

1. A method for measuring misalignment between substrates, the method comprising: The X-rays are directed to a first substrate and a second substrate that are bonded together but not yet bonded together. The guidance includes directing the X-rays to a first alignment mark and a third alignment mark in the first substrate, and to a second alignment mark and a fourth alignment mark in the second substrate. Fluorescent X-rays emitted from the first alignment mark and from the second alignment mark are detected to measure a first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first alignment mark relative to the second alignment mark; as well as At least some of the X-rays transmitted through the first substrate and the second substrate are detected using X-ray Talbot-Laur interferometry to measure a second misalignment of the first substrate relative to the second substrate based on a second detected misalignment of the third alignment mark relative to the fourth alignment mark.

2. The method as described in claim 1, wherein, The first alignment mark, the second alignment mark, the third alignment mark, and the fourth alignment mark are made of metal.

3. The method as described in claim 2, wherein, The metal includes at least one of copper (Cu), tungsten (W), cobalt (Co), ruthenium (Ru), or molybdenum (Mo).

4. The method of claim 2, wherein, The first detected misalignment is detected by a first detector that is sensitive to these fluorescent X-rays to measure the intensity of these fluorescent X-rays in relation to the thickness of the metal, and the second detected misalignment is detected by a second detector that is sensitive to these X-rays to measure the intensity of these X-rays transmitted relative to the TL interference pattern at the second detector.

5. The method of claim 1, wherein, The first alignment mark and the third alignment mark are located on the top surface of the first substrate, and the second alignment mark and the third alignment mark are located on the top surface of the second substrate.

6. The method of claim 5, wherein, The top surface of the first substrate faces the top surface of the second substrate, or the top surface of the first substrate faces the X-rays directed to the first substrate.

7. A method for measuring misalignment between substrates, the method comprising: The X-rays are directed to a first substrate and a second substrate that are in a bonded configuration but not yet bonded together. The guiding includes directing the X-rays to a first alignment mark in the first substrate and to a second alignment mark in the second substrate. Detect fluorescent X-rays emitted from the first substrate and the second substrate in response to X-ray irradiation of the first alignment mark and the second alignment mark; as well as The first misalignment of the first alignment mark relative to the second alignment mark is measured, at least in part, based on the wavelengths of these fluorescent X-rays that correspond to the metals in the first alignment mark and the second alignment mark.

8. The method of claim 7, further comprising: The wavelength is distinguished based on the metal used to measure the first misalignment, wherein the metal includes at least one of copper (Cu), tungsten (W), cobalt (Co), ruthenium (Ru), or molybdenum (Mo).

9. The method of claim 7, wherein, The first misalignment was detected by a first detector that is sensitive to these fluorescent X-rays to measure the intensity of these fluorescent X-rays in relation to the thickness of the metal.

10. The method of claim 9, wherein, The first detector is a silicon drift detector (SDD).

11. The method of claim 7, wherein, The first alignment mark and the second alignment mark share a common material.

12. A method for measuring misalignment between substrates, the method comprising: X-rays are directed to a first and a second substrate that are bonded together for subsequent bonding. These X-rays are transmitted using an X-ray Talbot-Laur interferometer through a first alignment mark in the first substrate and through a second alignment mark in the second substrate, wherein the first and second alignment marks comprise a first moiré interference grating pair; and The first misalignment of the first substrate relative to the second substrate is measured based on the detection of a first misalignment of the first alignment mark relative to the second alignment mark using a first interference pattern and a second interference pattern associated with the first moiré interference grating pair. The detection of the first misalignment includes measuring a first displacement of the first interference pattern in a first direction and a second displacement of the second interference pattern in a second direction opposite to the first direction.

13. The method of claim 12, wherein, The first detected misalignment is linearly correlated with a first sum of the absolute values ​​of the first displacement and the second displacement, wherein the first misalignment is at least 10 times smaller than the first sum.

14. The method of claim 12, wherein, The first moiré interference grating is aligned with the beam splitter grating of the X-ray Talbot-Lau interferometer.

15. The method of claim 12, further comprising: Using this X-ray Talbot-Lau interferometer, these X-rays are transmitted through a third alignment mark in the first substrate and a fourth alignment mark in the second substrate, wherein the third and fourth alignment marks comprise a second moiré interference grating pair; and The second misalignment of the first substrate relative to the second substrate is measured based on the second detected misalignment of the third alignment mark relative to the fourth alignment mark using the third and fourth interference patterns associated with the second moiré interference grating pair. The measurement of the second detected misalignment includes measuring the third displacement of the third interference pattern in the first direction and the fourth displacement of the fourth interference pattern in the second direction.

16. The method of claim 15, wherein, The second misalignment is linearly related to the second sum of the absolute values ​​of the third and fourth displacements, wherein the second misalignment is at least 50 times smaller than the second sum.

17. The method of claim 15, wherein, The first detection limit of the first moiré interference grating pair is equal to the Talbot fringe period of the analyzer grating of the X-ray Talbot-Laur interferometer divided by the first moiré amplification factor of the first moiré interference grating pair, and Wherein, the first detection upper limit of the first moiré interference grating pair is equal to ,in, It is the first spacing of the first alignment mark, and It is the second spacing of the second alignment mark.

18. The method of claim 17, wherein, The second detection lower limit of the second moiré interference grating pair is equal to the Talbot fringe period divided by the second moiré amplification factor of the second moiré interference grating pair, and The second detection upper limit of the second moiré interference grating pair is equal to ,in, It is the third spacing of the third alignment mark, and It is the fourth spacing of the fourth alignment mark.

19. The method of claim 18, wherein, The upper limit of the first detection is greater than the lower limit of the second detection.

20. The method of claim 15, wherein, These X-rays simultaneously penetrate through both the first pair of moiré interference gratings and the second pair of moiré interference gratings.

21. A method for measuring misalignment between semiconductor substrates, the method comprising: Using an X-ray Talbot-Lau (TL) interferometer, first X-rays are transmitted through a first substrate and a second substrate that are bonded together for subsequent bonding, including transmitting these first X-rays through a first alignment mark in the first substrate and a second alignment mark in the second substrate to generate second X-rays; These second X-rays are transmitted through a TL phase grating and through a TL analyzer grating to generate a third X-ray; as well as These third X-rays are received at an X-ray detector to measure the first misalignment of the first substrate relative to the second substrate based on a first detected misalignment of the first alignment mark relative to the second alignment mark detected using the X-ray detector.

22. The method of claim 21, wherein, Receiving these third X-rays at the X-ray detector further includes: Line scanning is performed by moving one of the TL phase grating or the TL analyzer grating on the first alignment mark and the second alignment mark; During the line scan, as the X-ray detector receives these third X-rays, the line scan signal from the output of the X-ray detector is recorded; and The first misalignment is detected based on the line scan signal.

23. The method of claim 22, wherein, Detecting the first misalignment based on the line scan signal further includes: Use a first reference line scan signal repository with an index of a calibrated misalignment value to match the line scan signal to the first detected misalignment.

24. The method of claim 22, wherein, The X-ray detector includes a silicon drift detector (SDD).

25. The method of claim 21, wherein, Receiving these third X-rays at the X-ray detector further includes: Image data of the first alignment mark and the second alignment mark are generated using an X-ray detector, wherein the X-ray detector is a flat-panel X-ray image detector; and Detecting the first misalignment based on the image data includes using a second reference image data repository indexed as calibrated misalignment values ​​to match the image data to the first misalignment.

26. The method of claim 21, wherein, The first alignment mark includes a first moiré grating element, and the second alignment mark includes a second moiré grating element, wherein the first moiré grating element and the second moiré grating element together form a moiré interference grating pair, and wherein the second X-rays include a moiré interference pattern.

Images