Interface-based defect inspection using second-harmonic generation

Abstract

To execute an inspection recipe, the inspection system may receive a reference second harmonic generation (SHG) image of the reference structure, where the reference structure has a multi-layer structure with one or more inversion-symmetric materials therein, based on scanning the reference structure with an illumination beam and collecting SHG light in response to the illumination beam. The system may further receive a test SHG image of the test structure, where the test structure and the reference structure have a common design, based on scanning the test structure with the illumination beam and collecting SHG light in response to the illumination beam. The system may further identify defects in the test structure by comparing the test and reference SHG images.

Description

[Technical Field]

[0001] [CROSS-REFERENCE TO RELATED APPLICATIONS] This application claims the benefit under 35 U.S.C. §119(e) of U.S. Provisional Patent Application No. 63 / 436,164, filed December 30, 2022, and U.S. Provisional Patent Application No. 63 / 449,491, filed March 2, 2023, the entire contents of both applications being incorporated herein by reference.

[0002] FIELD OF THE DISCLOSURE This disclosure relates generally to defect inspection, and more particularly to defect inspection using surface-selective second harmonic generation. [Background technology]

[0003] As semiconductor manufacturing advances from finFET (fin field effect transistor) architecture to gate-all-around (GAA) architecture nodes to achieve better performance, lower power, smaller area, and lower cost (PPAC), the associated process control requirements are becoming more stringent. For example, such devices incorporate multiple levels of channel regions and relatively small inter-channel spaces, which can simultaneously increase the number and types of potential defects that can occur during manufacturing and the difficulty of detecting those defects. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] U.S. Patent Application Publication No. 2022 / 0364850 [Patent Document 2] U.S. Patent Application Publication No. 2014 / 0300890 Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, there is a need to develop a system and method that addresses the above-listed drawbacks. [Means for solving the problem]

[0006] An inspection system is disclosed in accordance with one or more exemplary embodiments of the present disclosure. In one exemplary embodiment, the system includes a controller communicatively coupled to a detector in an optical subsystem. In another exemplary embodiment, the controller receives one or more reference second harmonic generation (SHG) images of one or more reference structures based on signals provided by the detector through scanning one or more reference structures with an illumination beam and collecting SHG light in response to the illumination beam, wherein each of the one or more reference structures includes a multi-layer structure having one or more inversion-symmetric materials within the multi-layer structure, and the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and one or more additional features in the multi-layer structure. In one exemplary embodiment, the controller receives one or more test SHG images of the one or more test structures based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting SHG light in response to the illumination beam, where the one or more test structures and the one or more reference structures have a common design. In another exemplary embodiment, the controller compares the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures.

[0007] An inspection system is disclosed in accordance with one or more exemplary embodiments of the present disclosure. In some exemplary embodiments, the system includes one or more illumination sources generating an illumination beam. Also, in some exemplary embodiments, the system includes a scanning subsystem including one or more elements configured to selectively direct the illumination beam at an off-axis angle of incidence toward the sample and to cause scanning movement of the illumination beam along a scan direction relative to the sample. Also, in some exemplary embodiments, the system includes a filter configured to reject a spectrum of the illumination beam and to allow a spectrum associated with second harmonic generation (SHG) of the illumination beam by at least a portion of the multilayer structure. Also, in some exemplary embodiments, the system includes a detector that captures light associated with the second harmonic of the first illumination beam. Also, in some exemplary embodiments, the system includes a controller communicatively coupled to the detector. In one illustrative embodiment, the controller receives one or more reference second harmonic generation (SHG) images of the one or more reference structures based on signals provided by the detector through scanning the one or more reference structures with the illumination beam and collecting SHG light in response to the illumination beam; wherein the one or more reference structures each have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, and the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and additional features in the multi-layer structure. In one exemplary embodiment, the controller receives one or more test SHG images of the one or more test structures based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting SHG light in response to the illumination beam, where the one or more test structures and the one or more reference structures have a common design. In another exemplary embodiment, the controller compares the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures.

[0008] A method is disclosed in accordance with one or more exemplary embodiments of the present disclosure, comprising generating one or more reference second harmonic generation (SHG) images of one or more reference structures based on signals provided by a detector through scanning one or more reference structures with an illumination beam and collecting SHG light in response to the illumination beam, wherein the one or more reference structures each have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, and the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and additional features in the multi-layer structure. An example embodiment method also includes generating one or more test SHG images of the one or more test structures based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting SHG light in response to the illumination beam, where the one or more test structures and the one or more reference structures have a common design, and comparing the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures.

[0009] Both the foregoing general description and the following detailed description are exemplary and explanatory only and do not necessarily limit the invention as claimed. The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the general description, serve to explain the principles of the invention.

[0010] Those skilled in the art will be able to better appreciate the numerous advantages of the present disclosure by reviewing the accompanying drawings. [Brief explanation of the drawings]

[0011] [Figure 1A] FIG. 1 is a block diagram of a second harmonic generation (SHG) inspection system according to one or more embodiments of the present disclosure. [Figure 1B] FIG. 1 is a simplified schematic diagram of an SHG inspection system according to one or more embodiments of the present disclosure. [Figure 2] FIG. 1 is a flow diagram depicting steps performed in a defect inspection method according to one or more embodiments of the present disclosure. [Figure 3] FIG. 1 is a schematic side view of a sample suitable for interfacial SHG defect inspection in accordance with one or more embodiments of the present disclosure. [Figure 4A] FIG. 1 is a schematic side view of an inversion-symmetric material formed of silicon, a first film formed as an interface layer, a second film formed as a high-k material, and a third film formed as a metal gate layer, according to one or more embodiments of the present disclosure. [Figure 4B] FIG. 4B is a schematic side view of the materials of FIG. 4A further including voids therein, according to one or more embodiments of the present disclosure. [Figure 5] FIG. 2 is a plot of optical extinction coefficient as a function of wavelength for silicon, in accordance with one or more embodiments of the present disclosure. [Figure 6] 1A-1C are schematic side views of gate regions of multi-channel field effect transistors (FETs) at intermediate process steps, according to one or more embodiments of the present disclosure. [Figure 7] FIG. 1 is a simplified simulation diagram of the intensity trend of interfacial SHG in a FET as a function of wavelength, according to one or more embodiments of the present disclosure. [Figure 8] FIG. 1 is a schematic diagram illustrating the generation of multiple difference images for multiple wavelengths in accordance with one or more embodiments of the present disclosure. [Figure 9A] 1A is a schematic side view of a gate region of a multi-channel FET at a first exemplary process step in accordance with one or more embodiments of the present disclosure. FIG. [Figure 9B] FIG. 10 is a schematic side view of a gate region of a multi-channel FET at a second exemplary process step, in accordance with one or more embodiments of the present disclosure. [Figure 9C]FIG. 10 is a schematic side view of a gate region of a multi-channel FET at a third exemplary process step, in accordance with one or more embodiments of the present disclosure. [Figure 9D] FIG. 10 is a schematic side view of a gate region of a multi-channel FET at a fourth exemplary process step, in accordance with one or more embodiments of the present disclosure. [Figure 10] FIG. 1 is a flow diagram depicting steps performed in an inspection recipe generation method according to one or more embodiments of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0012] Reference will now be made in detail to the disclosed subject matter, which is illustrated in the accompanying drawings. The present disclosure has been particularly shown and described with reference to certain embodiments and particular features thereof. The embodiments described herein should be considered illustrative rather than limiting. Various changes and modifications in form and detail may be made without departing from the spirit and scope of the present disclosure.

[0013] Embodiments of the present disclosure are directed to surface-selective defect inspection systems and methods based on second harmonic generation (SHG) technology. In particular, embodiments of the present disclosure are directed to interfacial SHG measurements for surface-selective inspection of structures having a thin film stack within which there is a substrate, the substrate having inversion symmetry that hinders SHG, such as, but not limited to, bulk forms of silicon. For purposes of this disclosure, the term inversion-symmetric material is used to describe a material that has inversion symmetry such that there is zero or weak SHG within the bulk of the material. In other words, the second-order nonlinear susceptibility (χ (2) ) is zero or sufficiently small, so that SHG within the bulk of the material is negligible for a particular application (e.g., the intensity of SHG within the bulk of the material induced by an illumination beam of a specified intensity is negligible).

[0014] As an example, the crystal structure of silicon is a diamond cubic lattice with inversion symmetry (e.g., centrosymmetric), so the electric field and polarization vectors in the Si bulk are invariant relative to the inversion system, thus preventing SHG. However, SHG can occur due to the higher-order nonlinear response of the inversion-symmetric material caused by electric dipoles on the Si surface, and the electric quadrupole response from the Si bulk when an external electric field is applied. In particular, when the inversion symmetry is broken along the direction perpendicular to the silicon surface, the second-order surface susceptibility at the interface becomes nonzero.

[0015] Further considerations herein have shown that the intensity of SHG generation at interfaces of such inversion-symmetric materials is highly sensitive to the presence of additional dipoles near the interface. As used herein, the terms intensity and amplitude, as they relate to SHG generation, are used interchangeably to refer to the intensity or amount of SHG light. Certain embodiments of the present disclosure are directed to systems and methods that utilize interface SHG to characterize one or more thin films near an inversion-symmetric substrate. Such measurements can, in turn, provide indirect measurements of the thin films based on their proximity to the inversion-symmetric substrate. Generally, such techniques may be suitable for identifying the types of defects that affect interface SHG at the nearest interface of an inversion-symmetric substrate. In some cases, such defects may be within 50 angstroms of the interface (e.g., 30 angstroms, 20 angstroms, or closer). It should be noted, however, that these are merely examples and not limitations of the present disclosure.

[0016] Further contemplated herein, such SHG measurements may be well suited for surface-selective defect inspection of various gate-all-around (GAA) field-effect transistor (FET) devices, including, but not limited to, GAA nanosheet FET devices, fork-sheet FET devices, complementary nanosheet FET devices, metal-on-silicon FET (MOSFET) devices, ferroelectric FET devices, ferroelectric memory devices (FeRAM), 2D-FET devices, 3D-FET devices, or 3D dynamic random access memory (DRAM) devices. Furthermore, such SHG measurements may be well suited for defects in 2D materials, including, but not limited to, transition metal dichalcogenides (e.g., MoS2, MoSe2, MoTe2, WS2, or WSe2) and III-IV chalcogenides (e.g., InSe or GaSe). These defects, especially when embedded, would be extremely difficult to detect using existing inspection techniques.

[0017] For example, within the gate region of many FET devices there is typically a thin film stack that includes one or more channel regions separated by a silicon substrate (e.g., a centrosymmetric material), as well as interfacial layers (ILs) made of various thin films, such as silicon dioxide (SiO2) and high-k dielectric materials (e.g., HfO2, ZrO2, HfSi x O y , HfO x N y etc.).

[0018] Such structures are prone to a variety of defects that are difficult to detect using conventional techniques, including, but not limited to, etch residues, channel bends, surface roughness after sacrificial SiGe separation from superlattice structures, and voids after metal gate (MG) fill processes. Many of these defects are buried defects within high aspect ratio (HAR) stacks, with feature sizes less than 10 nm. Detecting such defects has severely strained the capabilities of all current defect inspection strategies available in semiconductor fabs in terms of sensitivity, specificity, throughput, and capture speed. However, these defects are typically located near the channel and cannot be tolerated by process controls because they can have a significant impact on device performance and yield.

[0019] As one example, existing optical inspection techniques (e.g., optical bright-field inspection, etc.) may be suitable for detecting certain surface defects, but have limitations in their application to buried defects, and more generally, to defects having sizes substantially smaller than the illumination wavelength. As another example, scanning techniques, such as, but not limited to, electron beam inspection, through-focus scanning optical microscopy (TSOM), or atomic force microscopy (AFM), may generally provide high resolution for surface defects, but suffer from limited buried defect resolution, sample damage (especially when using the high landing voltages required to resolve buried defects), and / or low measurement throughput.

[0020] In certain embodiments, interface SHG light from interfaces of inversion-symmetric substrates (e.g., silicon) can be used to characterize neighboring materials or regions more generally. For example, interface SHG data for a sample of interest can be compared with interface SHG data from a reference sample to detect defects on the sample of interest. In addition, selective placement of a centrosymmetric substrate (e.g., silicon) in a region of interest (e.g., channel region) on the FET device provides a mechanism for interface SHG as well as highly selective metrology localized to the region of interest. Defect detection based on such interface SHG signals can, in turn, provide sensitive and selective performance, especially in regions that can most directly affect device performance.

[0021] Certain embodiments of the present disclosure are directed to various SHG inspection techniques or measurement modes. In certain embodiments, an illumination beam is directed toward a sample (e.g., containing FET devices) and SHG light is captured using a detector and a filter (e.g., a bandpass filter) for SHG signal isolation as the illumination beam is scanned across the sample. The illumination beam can generally be directed toward the sample at any angle, including orthogonal and off-axis angles of incidence. In certain embodiments, the SHG signal and, therefore, the signal-to-noise ratio (SNR) of the measurement can be enhanced by exposing the sample to additional stimuli. For example, the sample can be simultaneously illuminated with a separate light source and / or exposed to an electric field to excite dipoles and promote interface SHG. In certain embodiments, SHG signals associated with different wavelengths of incident illumination are measured during the inspection process. It is contemplated herein that the penetration depth of incident illumination into a sample can vary depending on its wavelength, allowing defects at different depths in the sample to be probed using different wavelengths of incident illumination.

[0022] The disclosed systems and methods based on interface-based defect inspection are nondestructive, high-throughput, and sensitive to additional optical excitation and external electric field disturbances at the channel surface / interface, regardless of the presence or absence of defects. The advantage of SHG behavior, as will be shown later, is that it is highly sensitive to the channel or silicon surface (or more generally, any inversion-symmetric material). As an example, the SHG behavior of a Si surface in an EPI process strongly depends on the surface binding of the Si surface to disilane or hydrogen in the EPI growth process. This results in greater sensitivity, enabling measurement of SHG intensity differences at hydrogen pressures between 0.8 and 1.5 Torr. More generally, surface binding with other species will produce different SHG behaviors, which can be used as a basis for defect inspection.

[0023] 1A-10, a selective metering system and method using SHG will now be described in more detail in accordance with one or more embodiments of the present disclosure.

[0024] Figure 1A is a block diagram of an SHG inspection system 100 according to one or more embodiments of the present disclosure. Figure 1B is a simplified schematic diagram of an SHG inspection system 100 according to one or more embodiments of the present disclosure.

[0025] In certain embodiments, an SHG inspection system 100 includes an illumination source 102 configured to generate an illumination beam 104, an illumination subsystem 106 having one or more optical elements that direct the illumination beam 104 toward a sample 108, a scanning subsystem 110 configured to scan the illumination beam 104 relative to the sample 108 (or vice versa), and a collection subsystem 112 having one or more optical elements that directs SHG light 114 generated by the sample 108 in response to the illumination beam 104 toward a detector 116. For example, the spectrum of the SHG light 114 may have twice the frequency or half the wavelength of the spectrum of the illumination beam 104.

[0026] Illumination source 102 may be any light source known in the art suitable for generating illumination beam 104 suitable for inducing SHG light 114 in sample 108. In certain embodiments, illumination source 102 is a laser light source, and thus illumination beam 104 is a coherent laser beam.

[0027] Furthermore, the spectral content of illumination beam 104 can be selected in any manner. For example, the representative wavelength (or center wavelength) of illumination beam 104 can be in any spectral range, including, but not limited to, the ultraviolet (UV), visible, infrared (IR), or near-IR range. In certain embodiments, illumination source 102 can include a laser light source. In certain embodiments, illumination source 102 is a tunable light source (e.g., a tunable laser light source or a tunable non-laser light source). This allows the wavelength, center wavelength, or more generally, the spectrum of illumination beam 104 to be tunable. For example, illumination source 102 can include, but is not limited to, a Ti:sapphire laser light source or a Yb-KGW laser light source.

[0028] The illumination beam 104 may generally have any temporal profile. In certain embodiments, the illumination beam 104 is formed as a train of pulses. The pulses may have any pulse duration. By way of non-limiting example, the illumination beam 104 may have pulse durations on the order of picoseconds, femtoseconds, or attoseconds, commonly referred to as ultrashort pulses. Such ultrashort pulses can advantageously provide high peak powers, which are suitable for efficiently inducing SHG in the sample 108. The pulses may also have any repetition rate, including, but not limited to, repetition rates in the kHz-MHz range.

[0029] The illumination subsystem 106 may include any combination of optical components suitable for directing the illumination beam 104 to the sample 108 and / or controlling the characteristics of the illumination beam 104. For example, the illumination subsystem 106 may include one or more lenses 118 that control the spot size of the illumination beam 104 on the sample 108. Alternatively, the illumination subsystem 106 may include one or more illumination control elements 120 that control parameters of the illumination beam 104, such as, but not limited to, the intensity, wavelength (or more generally, spectrum), polarization, spot size on the sample 108, or angle of incidence on the sample 108. For example, the illumination control elements 120 may include, but are not limited to, one or more polarizers, one or more spectral filters, one or more spatial filters, or one or more apodizers. Such illumination control elements 120 may be located in any suitable location, such as, but not limited to, a pupil plane or a field plane. Additionally, illumination subsystem 106 may direct illumination beam 104 onto the sample at any angle of incidence, including orthogonal incidence (eg, that depicted in FIG. 1B) and off-axis incidence.

[0030] The scanning subsystem 110 may include any component or combination of components suitable for providing relative motion between the illumination beam 104 and the sample 108. In certain embodiments, the scanning subsystem 110 includes a translation stage that translates (e.g., controls the position of) the sample 108 relative to the illumination subsystem 106 (e.g., a focusing lens, etc.). In certain embodiments, the scanning subsystem 110 includes one or more optical elements (e.g., beam scanning optics, etc.), such as, but not limited to, a galvanometer mirror coupled with an f-theta lens, that scan the illumination beam 104 across the sample 108.

[0031] In certain embodiments, the illumination subsystem 106 causes the illumination beam 104 to exhibit an asymmetric spot size on the sample 108. For example, the illumination beam 104 may be elongated along a direction perpendicular to the scan direction (e.g., the direction of relative motion between the illumination beam 104 and the sample 108) defined by the scanning subsystem 110. As an example, the illumination beam 104 may be focused into a line extending in a direction perpendicular to the scan direction. Such a configuration is depicted in FIG. 1B.

[0032] The collection subsystem 112 may include any combination of optical components suitable for directing the SHG light 114 from the sample 108 to the detector 116. In certain embodiments, the collection subsystem 112 includes one or more lenses 122 that collect light from the sample 108. In certain embodiments, the collection subsystem 112 includes one or more collection control elements 124 that control parameters of the collected light, such as, but not limited to, the intensity, wavelength (or more generally, spectrum), polarization, the point from which the light is collected on the sample 108, or the collection angle. For example, the collection control elements 124 may include, but are not limited to, one or more polarizers, one or more spectral filters, one or more spatial filters, or one or more apodizers.

[0033] In certain embodiments, the illumination control member 120 may include a first polarizer and the collection control member 124 may include a second polarizer. For example, the first polarizer may control the polarization of the illumination beam 104 incident on the sample 108, and the second polarizer may control the polarization of the SHG light 114 incident on the detector 116. Furthermore, the orientation of at least one of the first polarizer and the second polarizer may be adjusted to maximize the intensity of the SHG light 114 detected by the detector 116.

[0034] In certain embodiments, the collection subsystem 112 (e.g., collection control element 124) includes a filter 126 that selectively passes the SHG light 114 to the detector 116 or at least blocks reflected light associated with the spectrum of the illumination beam 104. For example, the filter 126 may include one or more spectral filters (e.g., dielectric filters, etc.), such as, but not limited to, a bandpass filter that selectively passes the SHG light 114, a bandreject filter that selectively rejects the spectrum of the illumination beam 104, or a lowpass filter (e.g., a lowpass wavelength filter) that blocks the spectrum of the illumination beam 104 and passes the spectrum of the SHG light 114. Alternatively, the filter 126 may include a dispersive element that spectrally disperses the light emitted by the sample 108, followed by a spatial filter that selectively passes the SHG light 114.

[0035] In certain embodiments, various components of the illumination subsystem 106 and the collection subsystem 112 are coupled or operated in tandem. For example, the SHG process is dependent on the polarization of the illumination beam 104. Therefore, the illumination subsystem 106 may include a polarizer (e.g., one of the illumination control elements 120) configured to provide the illumination beam 104 with a polarization that maximizes the intensity of the SHG light 114. The collection subsystem 112 may then include a polarizer (e.g., one of the collection control elements 124) configured to provide a filter 126 with the associated polarization.

[0036] The detector 116 may include any component or combination of components suitable for detecting the SHG light 114 and providing measurement data related to the SHG light 114. In certain embodiments, the detector 116 includes a single-pixel device, such as, but not limited to, a photodetector, an avalanche photodiode, or a photomultiplier tube. In certain embodiments, the detector 116 includes a multi-pixel device, such as, but not limited to, a charge-coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS) device. In certain embodiments, the detector 116 includes a spectrometer suitable for measuring the spectrum of light emitted from the sample 108 in response to the illumination beam 104. In general, the SHG inspection system 100 may include any number or type of detectors 116. This may make the SHG inspection system 100 more generally suitable for additional measurements beyond SHG measurements, such as, but not limited to, Raman spectroscopy and photoluminescence.

[0037] In certain embodiments, the SHG inspection system 100 includes one or more excitation sources 128 that enhance SHG generation associated with the illumination beam 104. The excitation sources 128 may include any source suitable for enhancing SHG generation associated with the illumination beam 104, such as, but not limited to, an additional illumination source or an electric field source.

[0038] In certain embodiments, the excitation source 128 includes an additional illumination source configured to generate the additional illumination beam 130. In this configuration, the SHG inspection system 100 can direct the additional illumination beam 130 at the same or different illumination angles toward the same portion of the sample 108 (e.g., the illumination beam 104 and the additional illumination beam 128 can overlap during scanning). For example, FIG. 1B depicts the additional illumination beam 130 incident on the sample 108 at an orthogonal angle of incidence. As a result, the additional illumination beam 130 can cause charge separation at one or more interfaces of the sample 108, thereby inducing a DC electric field and thereby promoting SHG and thereby increasing the intensity of the SHG light 114.

[0039] In certain embodiments, the excitation source 128 includes an electric field source (not explicitly shown in FIG. 1B ) configured to generate an electric field 132. For example, the electric field 132 can be oriented perpendicular to the surface of the sample 108 to modify the SHG behavior of the sample 108. As recognized, electric-field-induced second-harmonic generation (EFISH) is a third-order nonlinear process that relies on the interaction between an electric field and incident photons. In certain embodiments, a DC electric field 132 is applied to the symmetry-breaking surface, thereby enhancing the generation of interface SHG light 114 associated with the illumination beam 104. Specifically, the SHG light 114 is generated by an interface SHG process and can be enhanced by the promotion of third-order electrical susceptibility due to the external electric field 132. As contemplated herein, EFISH techniques have generally been used to analyze material properties but have not been used for surface-selective SHG metrology as disclosed herein.

[0040] In certain embodiments, SHG inspection system 100 further includes a controller 134, which includes one or more processors 136 configured to execute program instructions stored on a memory 138 (e.g., a storage medium). Controller 134 may be communicatively coupled to any component of SHG inspection system 100, including, but not limited to, detector 116. Execution of the program instructions by controller 134 may then cause one or more processors 136 to perform any step of the present disclosure. For example, controller 134 may execute an inspection recipe, receive metrology data related to SHG light 114 from sample 108 from detector 116, or receive reference and test SHG images derived from reference and test structures and identify defects in the test structures based on the reference and test SHG images.

[0041] The one or more processors 136 of the controller 134 may include any processor or processing element known in the art. For purposes of this disclosure, the terms "processor" or "processing element" are defined broadly to encompass any device having one or more processing or logic elements (e.g., one or more microprocessor devices, one or more application-specific integrated circuit (ASIC) devices, one or more field-programmable gate arrays (FPGAs), or one or more digital signal processors (DSPs)). In this sense, the one or more processors 136 may include any device configured to execute algorithms and / or instructions (e.g., program instructions stored in memory). In one embodiment, the one or more processors 136 may be embodied as a desktop computer, a mainframe computer system, a workstation, an image computer, a parallel processor, a networked computer, or any other computer system configured to execute programs configured to operate or work in conjunction with the SHG inspection system 100 as described elsewhere in this disclosure.

[0042] Furthermore, various subsystems of SHG inspection system 100 may include processors or logic elements suitable for performing at least a portion of the steps described herein. Accordingly, the above description should be construed as merely illustrative and not limiting on the embodiments of the present disclosure. Furthermore, the steps described throughout the present disclosure may be performed by a single controller 134 or, alternatively, by multiple controllers. Additionally, controller 134 may include one or multiple controllers housed within a common housing or multiple housings. In this manner, any controller or combination of controllers may be individually packaged as a module suitable for integration into SHG inspection system 100.

[0043] Memory 138 may include any storage medium known in the art suitable for storing program instructions executable by one or more associated processors 136. For example, memory 138 may include a non-transitory storage medium. Alternatively, memory 138 may include, but is not limited to, read-only memory (ROM), random access memory (RAM), magnetic or optical storage devices (e.g., disks), magnetic tape, solid-state drives, etc. It is further noted that memory 138 may be housed with one or more processors 136 within a common controller housing. In some embodiments, memory 138 may be remote from the physical location of one or more processors 136 and controller 134. For example, one or more processors 136 in controller 134 may access a remote memory (e.g., a server) accessible via a network (e.g., the Internet, an intranet, etc.).

[0044] 2-10, interfacial SHG for defect inspection will now be described in more detail in accordance with one or more embodiments of the present disclosure.

[0045] 2 is a flow diagram depicting steps performed in a defect inspection method 200 according to one or more embodiments of the present disclosure. It should be understood that the embodiments and enabling technologies previously described herein in the context of SHG inspection system 100 apply to method 200. However, it is further noted that method 200 is not limited by the architecture of SHG inspection system 100.

[0046] In certain embodiments, method 200 generates, at step 202, one or more reference SHG images of one or more reference structures based on signals provided by detector 116 from scanning the one or more reference structures by illuminating the one or more reference structures with illumination beam 104 and collecting SHG light 114 in response to illumination beam 104. Additionally, the reference structure may have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, where SHG light 114 is associated with one or more interfaces between the one or more inversion-symmetric materials and additional features in the multi-layer structure. In certain embodiments, method 200 generates one or more test SHG images of one or more test structures at step 204 based on signals provided by the detectors through scanning the one or more test structures by illuminating the one or more test structures with the illumination beam 104 and collecting SHG light 114 in response to the illumination beam 104, where the test structure and a reference structure have a common design. In certain embodiments, method 200 identifies defects in the test structures at step 206 by comparing the one or more test SHG images with the one or more reference SHG images. For example, defects may be identified at step 206 based on differences between the reference and test images.

[0047] In certain embodiments, the reference structure may be known to be free of defects of interest, and the test structure may have the same design (e.g., intended feature layout) as the reference structure. As a result, when the test structure is defect-free, the reference SHG image and the test SHG image will be equivalent within a specified tolerance. For example, the present contemplated application allows for differences between the reference and test SHG images based on allowable variations in the reference and test structures within design and / or manufacturing tolerances. Nevertheless, the present contemplated application allows for defects in the reference structure to be identified by measurable differences between the reference and test images.

[0048] Various techniques can be used to generate the reference image or images, all of which are within the spirit and scope of the present disclosure. In certain embodiments, the reference image is generated based on a single reference structure, e.g., one known to be defect-free. In certain embodiments, the reference image is generated based on a combination (e.g., average, etc.) of multiple images of one or more reference features on the same or separate specimen 108.

[0049] Additionally, the SHG light 114 used to generate the reference and test SHG images may be captured using any suitable technique or combination of techniques. In certain embodiments, the reference and test SHG images are generated based on illuminating the sample 108 with the illumination beam 104 and capturing the associated SHG light 114 without using any additional excitation source 128. In certain embodiments, the reference and test SHG images are generated based on illuminating the sample 108 with the illumination beam 104 and capturing the associated SHG light 114 while using one or more additional excitation sources 128. For example, the reference and test SHG images may be generated using the SHG light 114 in the presence of an additional illumination beam 130 and / or an electric field 132 (e.g., as depicted in FIG. 1B ). Furthermore, multiple sets of reference and test SHG images may be generated using different configurations of the illumination beam 104, the additional illumination beam 130, or the electric field 132.

[0050] 3-9D, defects suitable for inspection based on measurements of SHG light 114 will now be described in more detail, in accordance with one or more embodiments of the present disclosure. For example, the defects depicted in FIGS. 3-9D may be detected by SHG inspection system 100 and / or method 200. It should be understood, however, that the defects depicted in FIGS. 3-9D are merely illustrative and do not limit the types of defects that may be detected using SHG inspection system 100 and / or method 200.

[0051] FIG. 3 is a schematic side view of a sample 108 suitable for interfacial SHG defect inspection according to one or more embodiments of the present disclosure.

[0052] In certain embodiments, the sample 108 comprises at least one inversion-symmetric material 302 (eg, an inversion-symmetric substrate) and one or more films 304 (eg, thin films) of additional material.

[0053] The inversion-symmetric material 302 can include various materials known in the art that have inversion symmetry, such as, but not limited to, centrosymmetric materials (e.g., centrosymmetric crystals) and isotropic materials (e.g., glass). These materials will have zero or weak SHG in the bulk of the inversion-symmetric material 302. Furthermore, the inversion-symmetric material 302 can be of any phase, such as, but not limited to, quartz, glass, or ceramics. In certain embodiments, the inversion-symmetric material 302 is reflective at least with respect to the wavelength of the SHG light 114, such that the SHG light 114 returns from the sample 108 and is collected by the collection subsystem 112 on the same side of the sample 108 as the illumination subsystem 106. In certain embodiments, the inversion-symmetric material 302 includes a crystalline semiconductor substrate, such as, but not limited to, silicon.

[0054] The one or more films 304 may comprise any type of material known in the art. In certain embodiments, the one or more films 304 may also have inversion symmetry. In certain embodiments, the one or more films 304 may include SiO2 (e.g., 111) and / or high-k materials, such as, but not limited to, HfO2, ZrO2, and HfSi. x O y or HfO x N yExamples of suitable materials include those described above. In certain embodiments, the sample 108 can have a ferroelectric thin film stack including one or more HfO2 films 304 doped with a material such as, but not limited to, Zr, Al, Gd, La, Si, Sr, or Y. In certain embodiments, the sample 108 can have one or more HfO2 films 304 with an associated oxide, such as, but not limited to, SiO2 / HfO2 / AlO3 or HfO2 / ZrO2 / HfO2. In certain embodiments, the sample 108 can be a transition metal dichalcogenide-based material, such as, but not limited to, MoS2, MoSe2, MoTe2, WS2, or WSe2. In certain embodiments, the sample 108 can be a III-IV chalcogenide-based material, such as, but not limited to, InSe or GaSe. In certain embodiments, the inversion-symmetric material 302 can be at least partially surrounded by air or other gas.

[0055] As contemplated herein, surface-selective defect inspection can be performed on portions of the sample 108 containing the inversion-symmetric material 302 based on interfacial SHG occurring at the interface between the inversion-symmetric material 302 and a surrounding material (e.g., air, film 304, etc.). Consequently, SHG defect inspection may be suitable for thin film metrology of a wide range of devices and materials (e.g., features on the sample 108), including, but not limited to, GAA-FET devices, including non-limiting examples thereof, GAA nanosheet FET devices, fork-sheet FET devices, complementary nanosheet FET devices, MOSFET devices, ferroelectric FET devices, ferroelectric memory devices (FeRAM), 2D-FET devices, or 3D dynamic random access memory (DRAM) devices.

[0056] 4A and 4B are side views of thin film stacks formed of materials common to at least some FET devices. Specifically, FIG. 4A is a schematic side view of an inversion-symmetric material 302 formed of silicon, a first film 304a formed as an interfacial layer (IL) (e.g., SiO), a second film 304b formed as a high-k material (e.g., an annealed high-k material and / or a high-k material after creating an interface dipole engineering (IDE) layer), and a third film 304c formed as a metal gate (MG) layer, according to one or more embodiments of the present disclosure.

[0057] As depicted in FIG. 4A, the interface of the film 304 on the inversion-symmetric material 302 can have electric dipoles 402 that can promote interface SHG. The electric dipoles 402 in FIG. 4A can be primarily associated with the breaking of the inversion symmetry at the interface of the inversion-symmetric material 302. In certain applications, the number and / or concentration of the electric dipoles 402 are further controlled by an IDE layer (not explicitly shown). Such an IDE layer can be fabricated for various purposes, including but not limited to, to control the threshold voltage (V t ) is intended to modify the threshold voltage. Furthermore, the threshold voltage can be made particularly sensitive to the electric dipole 402 in the gate region as well as to defects in that region.

[0058] According to the present contemplation, interface SHG associated with interfaces of inversion-symmetric material 302 may also be sensitive to defects that, while not directly associated with the electric dipoles 402, still affect the local distribution of the electric dipoles 402 and, therefore, the local distribution of the SHG light 114. As an example, FIG. 4B is a schematic side view of the materials depicted in FIG. 4A , further including an internal void 404, according to one or more embodiments of the present disclosure. The presence of such defects may result in localized changes in the electric dipoles 402 (e.g., charge / trap distribution) at the neighboring interfaces of the inversion-symmetric material 302 (here, silicon), and thus localized modifications of the characteristics of the SHG light 114. Furthermore, the presence of such defects in different locations may result in different behavior. For example, if such defects were located exactly where the inversion-symmetric material 302 is not present, the resulting SHG behavior may be substantially different, potentially resulting in no measurable SHG light 114 at all. Therefore, the systems and methods disclosed herein based on interface SHG can perform selective defect inspection at locations surrounding such inversion-symmetric material 302. Furthermore, it should be understood that the depiction of void 404 in Figure 4B is merely illustrative, and that any type of defect that affects the local distribution of electric dipoles 402 near the interface of inversion-symmetric material 302, and thus the local distribution of SHG light 114 on sample 108, can be detected using the systems and methods disclosed herein.

[0059] Further contemplated herein is the selective presence of inversion symmetric material 302, such as, but not limited to, silicon, in specific areas of interest of semiconductor devices, such as, but not limited to, in the gate / channel regions of FET devices, GAA-FET devices, etc. As a result, the systems and methods disclosed herein enable selective defect inspection in areas that can critically affect device performance.

[0060] 5-7, the depth dependent interfacial SHG signal when using wavelength tuning will now be described in more detail in accordance with one or more embodiments of the present disclosure.

[0061] As contemplated herein, depth-dependent interface SHG information can be generated by controlling the wavelength of the illumination beam 104 (and thus the SHG light 114 generated at half the wavelength of the illumination beam 104) in conjunction with the extinction properties of the sample 108. Specifically, the extinction properties of the sample 108 can limit both the initial penetration depth of the illumination beam 104 into the sample 108 and the extent to which the SHG light 114 generated below the surface can propagate out of the plane and be available for detection.

[0062] 5 is a plot of the optical extinction coefficient (k) as a function of wavelength for silicon (which is an example of an inversion-symmetric material 302 common in semiconductor applications) in accordance with one or more embodiments of the present disclosure. As depicted in FIG. 5, silicon is largely transparent at wavelengths above 380 nm, with its extinction peak near 290 nm.

[0063] FIG. 6 is a schematic side view of the gate region of a multi-channel FET 602 at an intermediate process step, according to one or more embodiments of the present disclosure. For example, the FET 602 has three channels 604a-c, each at a different depth. In FIG. 6, the FET 602 includes various inversion-symmetric materials 302 (e.g., silicon) as bottom and middle substrates, surrounded by an inner spacer 606, source / drain EPI regions 608, gate spacers 610, and an interface layer 612 (e.g., a dielectric layer, such as, but not limited to, SiO2). FIG. 6 also depicts various penetration depths and associated SHG light 114 for different wavelengths of the illumination beam 104, as well as a defect 614 in the second channel 604b for illustrative purposes.

[0064] FIG. 7 is a simplified simulation of the intensity trend of interfacial SHG within FET 602 as a function of wavelength in accordance with one or more embodiments of the present disclosure.

[0065] 5-7, increasing the wavelength of illumination beam 104 can result in both increased penetration of illumination beam 104 and increased transmittance of SHG light 114 at half its wavelength. For silicon, the effect of increasing wavelength is particularly evident when the wavelength of SHG light 114 is greater than or equal to about 290 nm (e.g., that associated with the peak extinction coefficient in FIG. 6).

[0066] 6, SHG light 114 can be generated at multiple interfaces (particularly surface-selective interfaces between adjacent materials of interest, such as inversion-symmetric materials 302, as described herein) within a given penetration depth (defined not only by the penetration of illumination beam 104 but also by the transmitted intensity of SHG light 114 from that depth). As a result, the intensity of SHG light 114 can be based, in large part, on the cumulative emission of interfaces within that penetration depth, enabling depth-dependent inspection measurements.

[0067] In certain embodiments, test SHG images (and associated reference SHG images) are generated at multiple wavelengths to perform depth-resolved inspection metrology (e.g., defect depth identification). In particular, the multiple wavelengths of the illumination beam 104 can be selected so that the associated SHG light 114 falls within a spectral region where the extinction of the inversion-symmetric material 302 varies. In this manner, increasing the wavelength progressively increases the SHG light 114, thereby characterizing additional interfaces of interest. As an example, FIG. 7 illustrates the increase in SHG light 114 as a function of wavelength, and further illustrates the individual plateaus of SHG light 114 associated with various channels 604 within the FET 602. In this manner, the trend of SHG light 114 as a function of wavelength can provide specific information related to the structure of the FET 602. Furthermore, collecting cumulative information about the various interfaces allows for quantitative depth-dependent defect detection (eg, defect depth discrimination), or at least separate defect detection per channel 604.

[0068] In certain embodiments, defects can be identified based on reference and test SHG images at a particular designated wavelength, which can be selected to provide a known penetration depth. For example, in some applications, it may not be desirable to perform a wavelength scan, for reasons such as measurement efficiency. In such cases, reference and test SHG images based on measurements of SHG light 114 at known wavelengths associated with known penetration depths can provide cumulative or average information about all interfaces of interest within that penetration depth. As an example, a designated wavelength known to interrogate all three channels of FET 602 (e.g., a wavelength within the plateau for the lower channel as shown in FIG. 8) can provide information about all three channels, taking into account extinction as a function of depth, if necessary. Furthermore, additional assumptions or knowledge about the various channels and / or extinction trends as a function of depth can be used to generate specific information about any of the interfaces of interest.

[0069] In certain embodiments, depth-resolved defect inspection is performed based on one or more difference images. For example, a difference image may be generated by subtracting a first SHG image generated with the illumination beam 104 at a first wavelength from a second SHG image generated with the illumination beam 104 at a second wavelength. Further, reference and test difference images may be generated by subtracting the first and second reference SHG images from corresponding ones of the first and second test SHG images. In this manner, defects may be identified based on a comparison of the reference and test difference images (e.g., in step 206).

[0070] In certain embodiments, multiple difference images are generated based on different combinations of SHG images generated using different wavelength combinations. Figure 8 is a schematic diagram illustrating the generation of multiple difference images for wavelengths λ1-λ4 (e.g., those depicted in Figure 6) in accordance with one or more embodiments of the present disclosure. In particular, Figure 8 depicts a first difference image 802 resulting from the subtraction of SHG images generated using wavelengths λ1 and λ2, a second difference image 804 resulting from the subtraction of SHG images generated using wavelengths λ2 and λ3, a third difference image 806 resulting from the subtraction of SHG images generated using wavelengths λ3 and λ4, and a second difference image 808 resulting from the subtraction of SHG images generated using wavelengths λ1 and λ4, although additional difference images for other wavelength combinations are within the spirit and scope of the present disclosure. Additionally, a set of reference difference images (e.g., difference images 802-808 depicted in FIG. 8) can be generated based on the reference SHG image, and a set of test difference images (e.g., difference images 802-808 depicted in FIG. 8) can be generated based on the test SHG image. In this manner, defects can be identified (e.g., in step 206) by comparing the set of reference images with the set of test images. As an example, defects can be identified by subtracting reference and test difference images generated using the same wavelength combination.

[0071] Additionally, depth information (e.g., precise depth values ​​and / or associated channels 604) for those defects can be determined using wavelength combinations. For example, the difference image 802 associated with wavelengths λ1 and λ2 can be used to identify surface defects and / or defects associated with the first channel 604a because of the limited penetration depth at such wavelengths, as depicted in FIG. 6. Increasing the wavelength in turn increases the associated penetration depth, thereby providing discriminatory and quantitative depth information for deeper defects.

[0072] 9A-9D, various non-limiting examples of defects in a multi-channel FET 602 that may be detected using the interface SHG techniques disclosed herein are depicted. Specifically, FIGS. 9A-9D depict defect inspection of a multi-channel FET 602 at different process steps (e.g., different stages during the manufacturing process). It is contemplated herein that interface SHG enables highly sensitive defect inspection on or near the inversion-symmetric material 302 at various stages of manufacturing. In particular, the SHG light 114 associated with the inversion-symmetric material 302 (here, silicon) may be selective to gate regions (e.g., various channels 604 within the gate region) critical to the performance of the FET 602, thereby selectively identifying defects within those regions that may have a significant impact on the performance of the FET 602 (e.g., as fabricated).

[0073] 9A is a schematic side view of the gate region of a multi-channel FET 602 at a first exemplary process step, according to one or more embodiments of the present disclosure. Among other features, FIG. 9A depicts a dummy gate 902, inversion-symmetric material 302 associated with the channel 604, and a SiGe region 904. This structure may be associated with a process step that etches recesses 906 in the SiGe region 904 between the inversion-symmetric material 302. Non-limiting examples of defects that may be identified by interfacial SHG defects based on the proximity of the inversion-symmetric material 302 to the interface include, but are not limited to, unetched regions 908 (e.g., where the etching process was unsuccessful) and etch residue 910 (e.g., SiGe residue in the recesses 906).

[0074] 9B is a schematic side view of the gate region of multi-channel FET 602 at a second exemplary process step according to one or more embodiments of the present disclosure. Figure 9B is similar to Figure 9A, except that the defects depicted in Figure 9A are absent and source / drain EPI regions 608 are depicted. Non-limiting examples of defects that may be identified by interfacial SHG defects at this process step based on their proximity to the inversion-symmetric material 302 interface include, but are not limited to, void defects 912 in any of the source / drain EPI regions 608.

[0075] 9C is a schematic side view of the gate region of multi-channel FET 602 at a third exemplary process step according to one or more embodiments of the present disclosure. Figure 9C is similar to Figure 9B, except that the defects depicted in Figure 9B are absent and that dummy gate 914 and SiGe region 904 have been removed. Non-limiting examples of defects that may be identified by interfacial SHG defects at this process step based on the proximity of inversion-symmetric material 302 to the interface include, but are not limited to, surface roughness 916, channel bending 918, or etch residue 920 (e.g., partially or fully unetched regions) in any of inversion-symmetric material 302.

[0076] 9D is a schematic side view of the gate region of the multi-channel FET 602 at a fourth exemplary process step according to one or more embodiments of the present disclosure. Figure 9D is similar to Figure 9C, except that the defects depicted in Figure 9C are not present and the interface layer 612, high-k / I-D layer 922, and metal gate layer 924 have been created. Non-limiting examples of defects that may be identified by interface SHG defects at this process step based on their proximity to the inversion-symmetric material 302 interface include, but are not limited to, void defects 926 in any of the metal gate layers 924.

[0077] 5-9D , it should be understood that they are presented for illustrative purposes only and should not be construed as limiting. Rather, defect inspection using SHG light 114 from interfaces of an inversion-symmetric material 302, as disclosed herein, can be used to identify any type of defect that affects the generation of such SHG light 114, which may generally include various defect types, such as, but not limited to, surface roughness, surface profile distortions, voids, underfill defects, overfill defects, etch residue, or inclusion defects. Such defects may be present on the interfaces of the inversion-symmetric material 302 and / or in close proximity to the interfaces that affect the electric dipoles 402 that are responsible for the generation of the interface SHG light 114. In some cases, such defects may be located within 50 angstroms of the interface (e.g., 30 angstroms, 20 angstroms, or closer). It should be noted, however, that these are merely examples and are not intended to limit the present disclosure. Furthermore, the device types or regions that may be inspected for defects are not limited to the depicted FET 602, but rather may include, but are not limited to, ferroelectric layers in a ferroelectric FET device, 2D layers (e.g., transition metal dichalcogenides, III-IV chalcogenides, etc.) in a 2D-FET device, or layers in a 3D-DRAM device. Additionally, defect inspection of any non-FET device having one or more inversion-symmetric materials 302 is also contemplated herein.

[0078] 2, in certain embodiment method 200, at step 208, at least some of the defects are classified based on a comparison of the one or more test SHG images to the one or more reference SHG images. By way of non-limiting example, any of the defects depicted in Figures 9A-9D may be classified and differentiated at step 208. The defects may be classified using any suitable technique.

[0079] In certain embodiments, defects are classified based on design data of the structure being inspected. For example, defect locations in the test SHG images (and / or comparison images between the reference and test SHG images) can be associated with specific features based on design data, such as the intended layout for the relevant process steps at which the images were generated. This information can then be used to classify the defects based on known defect mechanisms in relevant regions at the relevant process steps. Furthermore, depth information provided based on the wavelengths or wavelength combinations used to generate the images, as described herein, can be used to further isolate and identify the three-dimensional locations of the defects and facilitate classification (e.g., based on known defect mechanisms in relevant regions at the relevant process steps).

[0080] In certain embodiments, defects are classified (e.g., via pattern recognition) based on patterns in the test SHG image (and / or comparison images between the reference and test SHG images). For example, FIG. 9C illustrates multiple defects 916-920 of different types (e.g., different classification classes) that may appear in a common region of a test SHG image (e.g., generated using wavelengths sufficient to probe the relevant depth). These defects may induce different variations in the SHG light 114 from nearby interfaces in the inversion-symmetric material 302. More specifically, different types of defects may affect the electric dipole 402 associated with the generation of SHG light 114 at such interfaces differently, due to differences in the surface structure of the interface, the composition of the associated materials (e.g., film 304, depicted generally in FIG. 3), material thickness, etc. As a result, various types of defects may generate distinct patterns suitable for classification. Furthermore, when multiple wavelengths are used as described herein, various defects can produce patterns in one or more difference images (e.g., difference images 802-808 depicted in FIG. 8), any of which can be used for defect classification.

[0081] In certain embodiments, defects are classified using supervised or unsupervised machine learning techniques. For example, a supervised machine learning algorithm can be trained on some combination of reference SHG images, test SHG images of samples with known defects, comparison images (e.g., images generated by subtracting the reference and test SHG images), difference images at different inspection wavelengths, or the like. The trained supervised machine learning algorithm can then classify defects identified on new test structures using corresponding input images, i.e., images of the type used for training but generated using the new test structures. Alternatively, unsupervised machine learning techniques can classify defects based directly on an analysis of similar inputs generated using some test structures.

[0082] In certain embodiments, defects are classified based on multiple sets of reference and test SHG images generated using different combinations or characteristics of illumination beam 104, additional illumination beam 130, or electric field 132 (e.g., those depicted in FIG. 1B ). For example, different types of defects may affect the generation of SHG light 114 differently under different inspection conditions, which can be used to classify the defects. Furthermore, the multiple sets of reference and test SHG images can be used in conjunction with any classification technique, including the use of design data, the use of defect depth information, pattern recognition techniques, machine learning techniques, or any combination thereof.

[0083] Reference is now made to FIG. 10, which is a flow diagram depicting steps performed in an inspection recipe generation method 1000 in accordance with one or more embodiments of the present disclosure. Applicant's note is that it should be understood that the embodiments and enabling technologies previously described herein in the context of SHG inspection system 100 apply to method 1000. However, it is further noted that method 1000 is not limited to the architecture of SHG inspection system 100. Furthermore, method 1000 can provide an inspection recipe for performing method 200 previously described herein.

[0084] The inspection recipe generally describes various parameters of the SHG inspection system 100 being measured. In certain embodiments, the inspection recipe incorporates various parameters of the illumination beam 104, such as, but not limited to, intensity, wavelength, polarization, spot size on the sample 108, or angle of incidence relative to the sample 108. In certain embodiments, the inspection recipe incorporates various parameters of the collected SHG light 114, such as, but not limited to, polarization. In certain embodiments, the inspection recipe incorporates various parameters of the detector 116, such as, but not limited to, gain settings. In certain embodiments, the inspection recipe incorporates various parameters of the excitation source 128 that enhance the SHG process. For example, the inspection recipe may incorporate various parameters of the additional illumination beam 130, such as, but not limited to, intensity, wavelength, polarization, spot size on the sample 108, or angle of incidence relative to the sample 108. As another example, the inspection recipe may incorporate various parameters of the electric field 132, such as, but not limited to, the field strength and the angle relative to the sample 108.

[0085] In certain embodiment method 1000, a sample 108 is selected for inspection at step 1002. The sample 108 may have any suitable structure, including some type of inversion-symmetric material 302 suitable for producing interface SHG light 114.

[0086] In certain embodiments, method 1000 includes selecting inspection parameters for inspecting specimen 108 for defects at step 1004. The inspection parameters may correspond to a preliminary inspection recipe and, therefore, may include operating parameters of SHG inspection system 100. For example, the inspection parameters may include characteristics of illumination beam 104, such as, but not limited to, wavelength, polarization, intensity, spot size, or angle of incidence. Alternatively, the inspection parameters may include characteristics of detector 116 or filter 126. Alternatively, the inspection parameters may include whether excitation source 128 is used and related parameters of additional illumination beam 130 and / or electric field 132 to be used.

[0087] The inspection parameters may also include techniques for identifying and / or classifying defects based on the reference and test SHG images generated using the selected parameters, thereby enabling the inspection parameters to include techniques for performing method 200 or any step therein.

[0088] In certain embodiments, method 1000 may identify one or more defects in the specimen 108 using the selected inspection parameters in step 1006. For example, method 200 or any step therein may be performed in step 1006.

[0089] In certain embodiments, method 1000 checks whether the defects were correctly identified in step 1008. For example, in step 1008, the identified defects may be verified using one or more alternative inspection systems (e.g., TEM, SEM, TSOM, optical bright field imaging, or some other alternative technique).

[0090] In certain embodiments, method 1000 may include, when the defects are properly identified (eg, verified), saving an inspection recipe based on the selected inspection parameters in step 1010 .

[0091] In certain embodiments, method 1000 selects new inspection parameters (e.g., a new preliminary inspection recipe) in step 1012. These new inspection parameters may include any new combination of parameters described in connection with step 1004. Additionally, step 1012 may incorporate additional steps, such as, but not limited to, applying additional algorithms or machine learning techniques to identify inspection parameters suitable for testing a particular sample 108.

[0092] In certain embodiments, step 1008 is repeated with the new parameters. If successful, the inspection recipe can be saved (e.g., step 1010). If unsuccessful, steps 1012, 1014, and 1008 can be repeated multiple times. For example, steps 1012, 1014, and 1008 can be repeated until inspection parameters are found that correctly identify defects on specimen 108 that can be verified by one or more additional inspection systems.

[0093] The subject matter described herein is sometimes depicted as various components embedded within or connected or coupled to other components. It should be understood that such illustrated architectures are merely exemplary, and that in fact, many other architectures that achieve the same functionality can be implemented. Conceptually, any arrangement of components that achieves the same functionality is effectively "integrated" to achieve the desired functionality. Thus, any two components herein that are combined to achieve a particular function can be considered to be "integrated" with each other to achieve the desired functionality, regardless of the architecture or intervening components. Similarly, any two components so integrated can be considered to be "connected" or "coupled" to each other to achieve the desired functionality, and any two components that can be so integrated can be considered to be "combinable" with each other to achieve the desired functionality. Examples of what can be coupled include, but are not limited to, physically interactable and / or physically interacting elements, and / or wirelessly interactable and / or wirelessly interacting elements, and / or logically interactable and / or logically interacting elements.

[0094] The present disclosure and many of its attendant advantages will be understood from the foregoing description, and it will also be apparent that various changes can be made in the form, construction and arrangement of the parts without departing from the disclosed subject matter or diminishing all of its essential advantages. The described form is illustrative only, and it is the intent of the following claims to encompass and embrace all such modifications. It is the appended claims which further define the invention.

Claims

1. 1. An inspection system comprising: a controller communicatively coupled to the detector in the optical subsystem, the controller having one or more processors configured to execute program instructions, the program instructions comprising: receiving one or more reference SHG images of the one or more reference structures based on signals provided by the detector through scanning the one or more reference structures with an illumination beam and collecting second harmonic generated (SHG) light in response to the illumination beam, wherein the one or more reference structures each have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, and wherein the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and one or more additional features in the multi-layer structure; receiving one or more test SHG images of the one or more test structures on the specimen based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting SHG light in response to the illumination beam, wherein the one or more test structures and the one or more reference structures have a common design; and comparing the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures; an inspection system that causes the one or more processors to execute an inspection recipe by

2. 10. The inspection system of claim 1, wherein the one or more processors further comprise: classifying at least one of the defects based on at least one of a known design of the structure under test, a discernible pattern in the one or more SHG images of the test associated with a known class of defects, and a machine learning algorithm trained on additional SHG images of known defects of a known class; an inspection system configured to execute the inspection recipe by

3. 10. The inspection system of claim 1, wherein the defect is: An inspection system for detecting at least one of surface roughness, surface profile distortion, voids, underfill defects, overfill defects, etch residue, and non-coalescence defects in the multi-layer structure.

4. 10. The inspection system of claim 1, wherein a spot size of the illumination beam on the specimen is elongated along a direction perpendicular to a scanning direction.

5. 10. The inspection system of claim 1, further comprising: an additional illumination source generating an additional illumination beam; one or more optical elements configured to direct the additional illumination beam toward the sample; and overlapping the illumination beam and the additional illumination beam during the scanning of the one or more structures under test, wherein the spectrum of the additional illumination beam is selected to enhance the SHG light.

6. 10. The inspection system of claim 1, further comprising: An inspection system comprising an electric field source that generates an electric field above the specimen, the electric field enhancing the SHG light.

7. 2. An inspection system according to claim 1, wherein the one or more reference structures include a single reference structure, and the one or more reference SHG images relate to the single reference structure.

8. 2. An inspection system as described in claim 1, wherein the one or more reference structures include two or more reference structures, and the one or more reference SHG images relate to the two or more reference structures.

9. 10. The inspection system of claim 1, wherein the spectrum of the illumination beam is tunable.

10. 10. The inspection system of claim 9, wherein the one or more processors further include: generating two or more sets of the reference and test SHG images based on the illumination beams at different wavelengths; and identifying a depth of at least one defect based on the two or more sets of the reference and test SHG images; An inspection system configured to execute program instructions.

11. 11. The inspection system of claim 10, wherein identifying the depth of at least one defect based on the two or more sets of the reference and test SHG images comprises: generating difference images between at least some of the test SHG images for the different wavelengths; and identifying the depth of at least one defect based on the difference image; Inspection system.

12. 10. The test system of claim 1, wherein the multi-layer structure is associated with at least one of a field effect transistor (FET) and a memory device.

13. 13. The inspection system of claim 12, wherein the multilayer structure relates to a gate region of at least one of a metal oxide semiconductor FET (MOSFET), a planar FET, a FinFET, a gate-all-around (GAA) FET, a GAA nanosheet FET, a fork-sheet FET, a complementary nanosheet FET, a ferroelectric FET, a 2D-FET, and a 3D-FET.

14. 13. The test system of claim 12, wherein the multi-layer structure is for at least one of a ferroelectric memory device and a 3D dynamic random access memory (DRAM) device.

15. 10. The inspection system of claim 1, wherein the one or more processors further include: generating two or more sets of the reference and test SHG images based on the illumination beams at different wavelengths; and identifying a depth of at least one defect based on the two or more sets of the reference and test SHG images; An inspection system configured to execute program instructions.

16. 16. The inspection system of claim 15, wherein the multi-layer structure is a field effect transistor (FET) having two or more channels in a gate region of the FET, and wherein identifying the depth of at least one defect based on the two or more sets of the reference and test SHG images comprises: An inspection system that identifies a channel in which the at least one defect is located.

17. 10. The inspection system of claim 1, wherein the one or more additional features are: The inspection system includes at least one of a high-k layer, an interface dipole engineering layer, a metal gate, and a SiGe region.

18. 2. The inspection system of claim 1, wherein the one or more inversion symmetric elements are: Silicon-containing inspection systems.

19. 1. An inspection system comprising: one or more illumination sources that generate illumination beams; an illumination subsystem having one or more lenses configured to direct the illumination beam onto the sample at an off-axis angle of incidence; a scanning subsystem including at least one of one or more translation stages and one or more optical elements, the scanning subsystem being further configured to cause scanning motion of the illumination beam relative to the sample along a scan direction; a filter configured to reject a spectrum of the illumination beam and to allow a spectrum associated with second harmonic generation (SHG) of the illumination beam by at least a portion of the sample; a detector for capturing light associated with a second harmonic of the illumination beam; a controller communicatively coupled to the detector; the controller having one or more processors configured to execute program instructions, the program instructions comprising: receiving one or more reference SHG images of the one or more reference structures based on signals provided by the detector through scanning the one or more reference structures with an illumination beam and collecting SHG light in response to the illumination beam, wherein the one or more reference structures each have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, and wherein the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and one or more additional features in the multi-layer structure; receiving one or more test SHG images of the one or more test structures on the sample based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting the SHG light in response to the illumination beam, wherein the one or more test structures and the one or more reference structures have a common design; and comparing the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures; an inspection system that causes the one or more processors to execute an inspection recipe by

20. 20. The inspection system of claim 19, further comprising: a first polarizer for controlling the polarization of the illumination beam incident on the sample; a second polarizer for controlling the polarization of the SHG light incident on the detector; An inspection system comprising:

21. 21. The inspection system of claim 20, wherein an orientation of at least one of the first polarizer and the second polarizer is adjusted to maximize the intensity of the SHG light detected by the detector.

22. 20. The inspection system of claim 19, wherein the detector comprises: An inspection system comprising at least one of a photomultiplier tube, a charge coupled device, and a photodiode.

23. 20. The inspection system of claim 19, wherein the scanning subsystem comprises the one or more translation stages, the one or more translation stages controlling the position of the specimen.

24. 20. The inspection system of claim 19, wherein the scanning subsystem comprises the one or more optical elements, the one or more optical elements controlling the position of the illumination beam.

25. 20. The inspection system of claim 19, wherein the one or more processors further comprise: classifying at least one of the defects based on at least one of a known design of the structure under test, a discernible pattern in the one or more SHG images of the test associated with a known class of defects, and a machine learning algorithm trained on additional SHG images of known defects of a known class; an inspection system configured to execute the inspection recipe by

26. 20. The inspection system of claim 19, wherein the defect comprises: An inspection system for detecting at least one of surface roughness, surface profile distortion, voids, underfill defects, overfill defects, etch residue, and non-coalescence defects in the multi-layer structure.

27. 20. The inspection system of claim 19, wherein a spot size of the illumination beam on the specimen is elongated along a direction perpendicular to the scanning direction.

28. 20. The inspection system of claim 19, further comprising: an additional illumination source generating an additional illumination beam; one or more optical elements configured to direct the additional illumination beam toward the sample; and overlapping the illumination beam and the additional illumination beam during the scan, the spectrum of the additional illumination beam being selected to enhance the SHG light.

29. 20. The inspection system of claim 19, further comprising: An inspection system comprising an electric field source that generates an electric field above the specimen, the electric field enhancing the SHG light.

30. 20. The inspection system of claim 19, wherein the one or more reference structures include a single reference structure, and the one or more reference SHG images relate to the single reference structure.

31. 20. An inspection system as described in claim 19, wherein the one or more reference structures include two or more reference structures, and the one or more reference SHG images relate to the two or more reference structures.

32. 20. The inspection system of claim 19, wherein the spectrum of the illumination beam is tunable.

33. 33. The inspection system of claim 32, wherein the one or more processors further include: generating two or more sets of the reference and test SHG images based on the illumination beams at different wavelengths; and identifying a depth of at least one defect based on the two or more sets of the reference and test SHG images of the SHG images; An inspection system configured to execute program instructions.

34. 34. The inspection system of claim 33, wherein identifying the depth of at least one defect based on the two or more sets of reference and test SHG images comprises: generating difference images between at least some of the test SHG images for the different wavelengths; and identifying the depth of at least one defect based on the difference image; Inspection system.

35. 20. The test system of claim 19, wherein the multi-layer structure is associated with at least one of a field effect transistor (FET) and a memory device.

36. 36. The inspection system of claim 35, wherein the multilayer structure is associated with a gate region of at least one of a metal oxide semiconductor FET (MOSFET), a planar FET, a FinFET, a gate-all-around (GAA) FET, a GAA nanosheet FET, a fork-sheet FET, a complementary nanosheet FET, a ferroelectric FET, a 2D-FET, and a 3D-FET.

37. 36. The test system of claim 35, wherein the multi-layer structure is for at least one of a ferroelectric memory device and a 3D dynamic random access memory (DRAM) device.

38. 20. The inspection system of claim 19, wherein the one or more processors further include: generating two or more sets of the reference and test SHG images based on the illumination beams at different wavelengths; and identifying a depth of at least one defect based on the two or more sets of the reference and test SHG images; An inspection system configured to execute program instructions.

39. 39. The inspection system of claim 38, wherein the multi-layer structure is a field effect transistor (FET) having two or more channels in a gate region of the FET, and wherein identifying the depth of at least one defect based on the two or more pairs of the reference and test SHG images comprises: An inspection system that identifies a channel in which the at least one defect is located.

40. 36. The inspection system of claim 35, wherein the one or more additional features include: The inspection system includes at least one of a high-k layer, an interface dipole engineering layer, a metal gate, and a SiGe region.

41. 41. The inspection system of claim 40, wherein the inversion symmetric material comprises: Silicon-containing inspection systems.

42. 1. A method comprising: generating one or more reference SHG images of the one or more reference structures based on signals provided by a detector from scanning the one or more reference structures with an illumination beam and collecting second harmonic generated (SHG) light in response to the illumination beam, wherein the one or more reference structures each have a multi-layer structure with one or more inversion-symmetric materials within the multi-layer structure, and wherein the SHG light is associated with one or more interfaces between the one or more inversion-symmetric materials and one or more additional features in the multi-layer structure; generating one or more test SHG images of the one or more test structures based on signals provided by the detector through scanning the one or more test structures with the illumination beam and collecting the SHG light in response to the illumination beam, wherein the one or more test structures and the one or more reference structures have a common design; and comparing the one or more test SHG images with the one or more reference SHG images to identify defects in the one or more test structures; method.

43. 43. The method of claim 42, further comprising: A method for classifying at least one of the defects based on at least one of a known design of the multilayer structure, identifiable patterns in the SHG images associated with known classes of defects, and a machine learning algorithm trained on additional SHG images of known classes of known defects.

44. 43. The method of claim 42, wherein the spectrum of the illumination beam is tunable.

45. 45. The method of claim 44, further comprising: generating two or more sets of the reference and test SHG images based on the illumination beam at different wavelengths; and identifying a depth of at least one defect based on the two or more sets of reference and test SHG images of the SHG images; method.

46. 46. ​​The method of claim 45, wherein identifying the depth of at least one defect based on the two or more sets of reference and test SHG images comprises: generating difference images between at least some of the test SHG images for the different wavelengths; and identifying the depth of at least one defect based on the difference image; method.

Images