Optical sensor for measuring film thickness

Abstract

The method for measuring film thickness involves irradiating the upper layer of a sample in a first region with a broadband illumination beam. The sample includes a substrate and a plurality of semiconductor structures formed between the substrate and the upper layer. A first reflectance spectrum of the sample is obtained in the first region. The first thickness of the upper layer in the first region is determined by applying an upper layer model to the first reflectance spectrum. The upper layer model is substantially unaffected by the plurality of semiconductor structures.

Description

Technical Field

[0001] Incorporation by Reference This disclosure claims the benefit of U.S. Provisional Patent Application No. 63 / 521,757, filed Jun. 19, 2023, and U.S. Non-Provisional Patent Application No. 18 / 501,672, filed Nov. 3, 2023, the entireties of which are incorporated herein by reference.

[0002] This disclosure relates to a measuring device and a method for measuring film thickness.

Background Art

[0003] In the manufacture of semiconductor devices (especially at the microscopic scale), various fabrication processes such as thin film formation deposition, etching mask generation, patterning, material etching and removal, and doping processes are carried out. These processes are repeatedly performed to form desired semiconductor device elements on a substrate. What is important for these processes is film thickness control and film uniformity control. Therefore, in the semiconductor industry, various film thickness characterization techniques have been developed and used.

Summary of the Invention

[0004] This disclosure relates to a method and an apparatus for measuring film thickness.

[0005] According to a first aspect of the present disclosure, a method for measuring film thickness is provided. The method includes irradiating an upper layer of a sample with a broadband illumination beam in a first region. The sample includes a substrate and a plurality of semiconductor structures formed between the substrate and the upper layer. A first reflectance spectrum of the sample in the first region is acquired. A first thickness of the upper layer in the first region is determined by applying an upper layer model to the first reflectance spectrum. The upper layer model is substantially not affected by the plurality of semiconductor structures.

[0006] In some embodiments, applying the upper layer model involves performing a Fourier transform (FT) of the first reflectance spectrum to obtain an FT spectrum. The first thickness is determined based on the first-order peak of the FT spectrum that has the highest intensity in the FT spectrum.

[0007] In some embodiments, the FT is performed on at least a first portion of the first reflectance spectrum that is substantially unaffected by the multiple semiconductor structures. In some embodiments, the FT is performed on both the first portion and a second portion of the first reflectance spectrum that is affected by the multiple semiconductor structures. In some embodiments, the FT spectrum further includes secondary peaks affected by the multiple semiconductor structures. Primary peaks are substantially unaffected by the multiple semiconductor structures. The first thickness is determined based on and without secondary peaks of the FT spectrum so that the overlayer model is substantially unaffected by the multiple semiconductor structures.

[0008] In some embodiments, a second portion of the first reflectance spectrum, which is affected by multiple semiconductor structures, is truncated before determining the first thickness of the upper layer. In some embodiments, a first portion below the critical wavelength is truncated.

[0009] In some embodiments, the position of the primary peak in the FT spectrum relative to the first thickness is calibrated experimentally or computationally. In some embodiments, the horizontal axis of the FT spectrum is scaled by n and k, which are the complex refractive indices of the upper layers, to obtain a scaled FT spectrum such that the position of the primary peak in the scaled FT spectrum is independent of n and k.

[0010] In some embodiments, obtaining a first reflectance spectrum of a sample involves collecting at least two reflected beams of substantially orthogonal polarization. The first reflectance spectrum is obtained as the ratio of the reflectances collected by the at least two reflected beams. In some embodiments, a critical wavelength is identified at which the ratio begins to differ from 1.

[0011] In some embodiments, the first reflectance spectrum of a sample is a reflection coefficient spectrum. A critical wavelength, which is the boundary between two regions of the reflection coefficient spectrum, is identified.

[0012] In some embodiments, the sample, broadband illumination beam, or a combination thereof is rearranged. The upper layer of the sample is illuminated in a second region with the broadband illumination beam. A second reflectance spectrum of the sample in the second region is obtained. The second thickness of the upper layer in the second region is determined by applying an upper layer model to the second reflectance spectrum. In some embodiments, multiple thicknesses of the upper layer in multiple regions are determined by an XY pattern or R-θ pattern over at least a portion of the sample. In some embodiments, a thickness of 5 to 100,000 of the upper layer is measured over part or all of the sample.

[0013] In some embodiments, the upper layer comprises a semiconductor material, and the broadband illumination beam has a wavelength range of 200 to 1000 nm.

[0014] In some embodiments, the upper layer includes silicon.

[0015] According to a second aspect of this disclosure, an apparatus is provided. The apparatus includes a handling stage configured to receive a sample. The sample includes a substrate, an upper layer, and a plurality of semiconductor structures formed between the substrate and the upper layer. A light source is configured to emit a broadband illumination beam. An optical system is configured to guide the broadband illumination beam to illuminate the upper layer of the sample, collect the reflected beam from the upper layer, and guide the reflected beam to a photodetector configured to acquire the reflectance spectrum of the sample. A controller is configured to determine the thickness of the upper layer by applying an upper layer model to the reflectance spectrum. The upper layer model is substantially unaffected by the plurality of semiconductor structures.

[0016] In some embodiments, the optical system includes a Schwarzschild objective and a knife-edge prism (KEP). The Schwarzschild objective includes a primary mirror and a secondary mirror. The KEP has a first surface and a second surface. The Schwarzschild objective and KEP are configured such that a broadband illumination beam is directed by the first surface of the KEP, through the aperture of the primary mirror to the secondary mirror, reflected by the primary mirror, and then reflected to the upper layer of the sample. The Schwarzschild objective and KEP are configured such that the reflected beam from the upper layer is reflected by the primary mirror to the secondary mirror, through the aperture of the primary mirror, reflected by the second surface of the KEP, and then directed to a photodetector.

[0017] In some embodiments, the optical system includes at least one selected from the group consisting of optical fibers, focusing lenses, system apertures, and polarizers. In some embodiments, the orientation of the polarizer is switchable between at least two substantially orthogonal orientations.

[0018] In some embodiments, the handling stage is an XY, XYZ, R-θ, or R-θ-Z stage. The light source includes at least one selected from the group consisting of a laser, laser diode, light-emitting diode, gas discharge light source, and laser-driven light source.

[0019] In some embodiments, the apparatus further includes a processing chamber configured for surface treatment of a sample.

[0020] In some embodiments, the photodetector is a broadband high-resolution spectrometer.

[0021] A processing system is provided according to a third aspect of the present disclosure. The processing system includes a wafer processing module and an optical apparatus. The optical apparatus includes a handling stage configured to receive a sample. The sample includes a substrate, an upper layer, and a plurality of semiconductor structures formed between the substrate and the upper layer. A light source is configured to emit a broadband illumination beam. The optical system is configured to guide the broadband illumination beam to illuminate the upper layer of the sample, collect the reflected beam from the upper layer, and guide the reflected beam to a photodetector configured to acquire the reflectance spectrum of the sample. A controller is configured to determine the thickness of the upper layer by applying an upper layer model to the reflectance spectrum. The upper layer model is substantially unaffected by the plurality of semiconductor structures.

[0022] In some embodiments, the wafer processing module is configured to perform surface treatment on a sample.

[0023] In some embodiments, the wafer processing module is configured to selectively remove a portion of the upper layer of the sample.

[0024] In some embodiments, the controller is configured to communicate a wafer map, which includes multiple upper layer thickness measurements, to a wafer processing module.

[0025] Please note that this summary section does not specify all embodiments and / or progressively novel aspects of the invention as described in this disclosure or claims. Instead, this summary provides only a preliminary consideration of different embodiments and corresponding novelties. For further details and / or anticipated aspects of the invention and embodiments, readers should refer to the detailed description sections and corresponding figures of this disclosure discussed further below. [Brief explanation of the drawing]

[0026] Aspects of the present disclosure will be better understood by reading the following detailed description in conjunction with the accompanying drawings. Note that, in accordance with standard practice in the industry, various features are not drawn to scale. In fact, the dimensions of the various features may be increased or decreased for clarity of discussion.

[0027] [Figure 1] A vertical cross-sectional view of a sample according to some embodiments of the present disclosure is shown. [Figure 2] A flowchart of a process for film thickness measurement according to some embodiments of the present disclosure is shown. [Figure 3A] A schematic diagram of an apparatus according to some embodiments of the present disclosure is shown. [Figure 3B] A perspective view of an apparatus according to some embodiments of the present disclosure is shown. [Figure 4A] A reflectance spectrum of a 1-μm-thick Si layer on a thin SiO2 layer on Si according to some embodiments of the present disclosure is shown. [Figure 4B] An FT spectrum of a 1-μm-thick Si layer on a thin SiO2 layer on Si according to some embodiments of the present disclosure is shown. [Figure 5] A plot of the critical wavelength versus the upper Si layer thickness above a typical structure on a Si substrate according to some embodiments of the present disclosure is shown. [Figure 6] A plot of the FT peak position versus the upper Si layer thickness on any structure on a Si substrate according to some embodiments of the present disclosure is shown. [Figure 7] A plot of the FT peak position versus the upper Si layer thickness on any structure on a Si substrate according to some embodiments of the present disclosure is shown. [Figure 8] A spectrum of the reflectance ratio of orthogonal polarization Rs to Rp of the upper Si layer on a typical semiconductor structure on a Si wafer according to some embodiments of the present disclosure is shown. [Figure 9] A spectrum of the reflectance ratio of the polarization signal, as well as the spectral reflectance of bare silicon and the spectral reflectance of the upper silicon layer on a typical semiconductor structure on a Si substrate, according to some embodiments of the present disclosure is shown. [Figure 10]The identification of the spectral envelope of a spectrum by some embodiments of this disclosure is shown. [Modes for carrying out the invention]

[0028] The following disclosure provides a variety of embodiments or examples for implementing various features of the subject matter provided. For the sake of brevity, specific examples of components and arrangements are described below. Naturally, these are merely examples and are not intended to be limiting. For example, in the following description, forming a first feature on or on a second feature may include embodiments in which the first and second features are formed in direct contact, and also embodiments in which additional features may be formed between the first and second features so that the first and second features are not in direct contact. In addition, the disclosure may repeat reference numbers and / or letters in various examples. This repetition is for the purpose of simplification and clarity and does not in itself determine the relationships between the various embodiments and / or configurations considered. Furthermore, spatially relative terms such as “top,” “bottom,” “underside,” “downward,” “lower,” “upper,” and “upper” may be used herein to facilitate the description of the relationship between one element or feature shown in the figures and another element or feature. Spatially relative terms are intended to include different orientations of the device in use or operation, in addition to the illustrated orientation. The apparatus may be in other orientations (rotated 90 degrees or otherwise), and spatially relative descriptors used herein may be interpreted accordingly.

[0029] The order in which the different steps described herein are considered is presented for the sake of clarity. In general, these steps can be performed in any suitable order. Furthermore, although each of the various features, techniques, structures, etc. described herein may be considered in separate sections of this disclosure, each concept is intended to be practiced independently or in combination with one another. Accordingly, the present invention can be embodied and considered in many different ways.

[0030] In drawings, the same reference numerals indicate the same or corresponding parts across multiple drawings. In addition, as used herein, words such as "a" and "an" generally mean "one or more" unless otherwise specified.

[0031] Furthermore, the terms “approximately,” “about,” “around,” and similar terms generally refer to a range that includes an identified value and any value in between, within a margin of 20%, 10%, or preferably 5%.

[0032] Thin-film optical measurement is one of the fundamentals of semiconductor manufacturing processes. Currently, the preferred techniques include spectral reflectance measurement (SR) and ellipsometry (SE), both based on the spectroscopic analysis of reflections from thin films. These techniques are widely used due to their best-in-class sensitivity, high throughput, reasonable cost, and non-destructive measurement capabilities. However, spectroscopic analysis typically requires information and assumptions about the layer stack and the structure formed on the wafer so that a multilayer model of the sample structure can be constructed. Such a multilayer model typically takes into account the upper layer of interest and at least one lower layer. A composite spectrum is then generated using the multilayer model, which can be compared to a measured spectrum to identify the target parameters, such as film thickness. However, if information about the layer thickness and the structure formed within the layer is incomplete, spectroscopic techniques may not provide accurate measurement results. Similar drawbacks exist when the optical properties of the layers and structures are unknown and / or difficult to obtain. Even when all the necessary information is available, actual measurement usually requires time and engineering effort to construct and optimize the multilayer model. Therefore, there is an unmet need for a technique that can measure thin film thickness with nanometer-level sensitivity using an upper layer model (or monolayer model) that is not based on or is not related to the underlying layers and structure.

[0033] The techniques described herein include, for example, apparatus and methods for measuring the thickness of an upper layer 121 formed on a structure 113 formed on a substrate 111 in sample 100 of Figure 1. In one example, this type of structure (e.g., 100) may be formed by obtaining two semiconductor wafers W1 and W2, on which at least one wafer (e.g., W1) has a complex semiconductor structure (e.g., 113) fabricated thereon, bonding them top-down or face-to-face, and then thinning one of the bonded wafers (e.g., W2) until only a thin Si layer (e.g., 121) remains on the structure 113. The objective is to measure the thickness of the upper Si layer (e.g., 121) remaining throughout sample 100 and provide a map of Si layer thicknesses to semiconductor processing tools for subsequent processing. In another example, sample 100 may be formed by depositing an upper layer 121 on top of the structure 113.

[0034] Both the upper layer 121 and the substrate 111 may contain silicon (Si), but may also contain other semiconductor materials commonly used in semiconductor manufacturing. Structure 113 may include any circuits, transistors, metal / dielectric patterns, junction interfaces, etc., as those skilled in the art will understand. For illustrative purposes, Si is used as an unrestricted example of the upper layer 121, and line patterns are used as unrestricted examples of structure 113. Thus, the upper layer 121 is also referred to as the upper Si layer 124.

[0035] One challenge to overcome is the lack of information regarding the properties of the structure 113 beneath the upper Si layer 124. Such properties may be either poorly defined or completely unknown. Even if such properties were known, another challenge would be to establish and optimize a robust multilayer model. For example, a known circuit could be formed directly beneath the upper Si layer 124. However, conventional models typically assume flat, uniform layers, making it difficult to construct multilayer models that include known circuits.

[0036] According to aspects of this disclosure, the method described herein utilizes the absorption properties of a top layer material, such as silicon (Si). An apparatus for measuring the spectral reflectance of the top Si layer 124 under broadband illumination is used, and the thickness of the top Si layer 124 is determined using the measured spectral dependence of absorption within the top Si layer 124 and reflection from the top Si layer 124. In one embodiment, a spectroreflectometer (SR) may be used to obtain such a reflectance spectrum. Hereinafter, the spectroreflectometer used in this method will be referred to as an absorption spectroreflectometer (aSR).

[0037] Continuing to refer to Figure 1, the spectral reflectance obtained when the upper Si layer 124 in Figure 1 is illuminated from above (as opposed to monochromatic illumination) in a broadband manner is generally influenced not only by the properties of the upper Si layer 124, whose thickness is specified, but also by the properties of the underlying structure (e.g., 113 and 111). When a thin film (e.g., the upper Si layer 124) is illuminated with broadband illumination, the reflection spectrum exhibits a characteristic periodicity due to constructive and destructive interference that differently affects the wavelength of the illumination beam as it traverses the thin film (e.g., 124), reflects from the bottom surface 122, and returns traversing the thin film towards the top surface 123. Semiconductor materials such as Si absorb a portion of the incident broadband illumination, and light transmittance generally increases with wavelength so that it is hardly absorbed in the infrared (IR) portion of the spectrum.

[0038] If, at a given wavelength, the absorption in the upper Si layer 124 in Figure 1 is large enough to cause complete attenuation of the illumination beam before it reaches the underlying structure (e.g., 113 and 111), then the reflectance of the upper Si layer 124 at that wavelength will substantially match the reflectance of the same Si layer without the underlying structure or any other film formed thereon. The same is expected to be true if a portion of the illumination beam that could have reached the underlying structure (e.g., 113 and 111) is reflected by the underlying structure (e.g., 113) and a portion of the illumination beam is completely attenuated as it crosses the upper Si layer 124 and returns towards the upper surface 123.

[0039] Therefore, due to the wavelength dependence of absorption in Si, it is expected that there will be a portion of the reflectance spectrum with sufficiently high absorption, and thus the reflectance spectrum of the upper Si layer 124 with the substrate feature will substantially coincide with the reflectance spectrum of the same Si layer without the substrate structure or any other film formed beneath it. Conversely, for wavelengths other than this portion of the reflectance spectrum, the spectrum will differ due to insufficient attenuation and will exhibit the periodic characteristics of the thin film reflectance spectrum. This periodicity may potentially be further altered by the substrate structure diffracting a portion of the illumination beam at even longer wavelengths, thus altering only the periodic characteristics of the thin film.

[0040] Figure 4A may show the reflectance spectrum 400A of a Si thin film lacking a substrate structure (e.g., an upper Si layer 124). The wavelength at which two portions 401 and 403 of the reflectance spectrum 400A of the upper Si layer 124 meet is called the "critical wavelength" (CWL) 411. In the example in Figure 4A, CWL 411 is approximately 430 nm. Since absorption in Si decreases with increasing wavelength, CWL 411 is the longest wavelength of the illumination beam at which the beam can be completely attenuated for a given thickness of the upper Si layer 124. As the thickness of the upper Si layer increases, it becomes possible to completely attenuate longer wavelengths, so CWL 411 exhibits a dependence on the thickness of the upper Si layer, as shown in plot 500 in Figure 5. This dependency can be advantageously used to measure the thickness of the upper Si layer 124, and in at least one embodiment, the CWL directly identified from the measurement spectrum can be used to measure the Si layer thickness, provided that the dependence of the CWL on the thickness is determined, for example, experimentally or by a calculation method if the experimental setup or computational model lacks the underlying structure (e.g., 113).

[0041] The advantage of this measurement method is that it does not require detailed information about the structure 113 beneath the upper Si layer 124. However, a disadvantage is that the method may require substantial absorption within the upper Si layer 124 in at least a portion of the illumination light spectrum to ensure the presence of CWL within the wavelength range of the broadband illumination beam.

[0042] If the aSR is equipped to modify the polarization orientation of the illumination beam so that two spectra can be obtained at substantially orthogonal polarizations, additional processes can be used to improve the accuracy of the measured CWL, such as analyzing the ratio of the reflectance spectra obtained at two substantially orthogonal polarizations, as shown in Figure 8. Point 811, where the ratio differs from 1, is the CWL in Figure 8. Point 911, where the ratio differs from 1, is the CWL in Figure 9, and plots 901, 903, and 905 show the reflectance ratio spectra for orthogonal polarization, the reflectance spectrum for bare silicon, and the reflectance spectrum for the upper silicon layer on a typical semiconductor structure on a Si substrate, respectively.

[0043] Figure 10 shows how CWL1011 can be identified from the reflectance spectrum 1000 (e.g., the reflectance spectrum 1001 or alternative reflectance ratio spectra of different polarizations) by calculating the upper envelope 1003 and lower envelope 1005 of the reflectance coefficient spectrum 1001 and identifying the CWL1011 where the upper envelope 1003 and lower envelope 1005 "contact". This method is particularly suitable for thicker upper Si layers where the reflectance spectrum exhibits high-density, low-pitch spectral periodicity.

[0044] However, in practice, further challenges arise because spectral measurement noise, the roughness of the upper and lower surfaces of the upper Si layer 124, or the weakly defined interface (e.g., 122) between the upper Si layer 124 and the (substrate) structure 113, i.e., the junction interface, adversely affect a portion of the illumination beam reflected from the lower surface 122 and therefore the measured reflection spectrum.

[0045] To overcome these challenges, in at least one embodiment, a Fourier transform (FT) can be applied to the acquired reflectance coefficient spectrum (e.g., Figure 4A) or reflectance ratio spectrum (e.g., Figure 8) to obtain the FT spectrum 400B. In Figure 4B, frequency analysis of the FT spectrum 400B is performed. The inventors have found that the primary peak 421 of the FT spectrum 400B is dominated by the periodicity of the reflectance spectrum 400A at the wavelength of portion 405 of the reflectance spectrum 400A, which is usually immediately to the right of the CWL 411. Portion 405 of the reflectance spectrum 400A is dominated by constructive and destructive interference of poorly attenuated illumination beams within the upper Si layer 124. For example, in Figure 8, the measured ratio of the spectrum obtained in orthogonal polarization and by the substrate structure exhibits a clear periodicity in the region from CWL (430 nm) to about 550 nm. The periodicity begins to change substantially at wavelengths above 550 nm due to diffraction from the substrate structure (e.g., 113).

[0046] In a non-restrictive example, a reflectance spectrum (e.g., 400A) can be converted to a frequency (or wavenumber) domain that includes the refractive index of Si. Here, frequency =

number

[0047] Figure 6 shows that the Fourier transform exhibits a peak frequency (e.g., wavenumber) with respect to the thickness of the upper Si layer 124, as in the case of thin film thickness measurement without the underlying structure. Therefore, by using the Fourier transform of the acquired reflectance spectrum, it is possible to avoid explicit identification of CWL411, as the FT spectrum 400B is little to no influence from the slowly changing non-vibrating portion (e.g., 401) of the reflectance spectrum acquired at wavelengths lower than CWL411. On the other hand, to reduce noise in the FT spectrum 400B, it may be advantageous to block a portion (e.g., 407) of the acquired reflectance spectrum (e.g., 400A) where diffraction from the underlying structure (e.g., 113 and 111) begins to affect the spectrum (e.g., wavelengths above approximately 550 nm in Figure 8).

[0048] However, the inventors have discovered that, for a wide range of cases, blocking of such spectral ranges is unnecessary, and the thickness of the upper Si layer 124 can be measured with sufficient accuracy by calculating the Fourier transform of the entire reflectance coefficient spectrum (e.g., 400A) or the reflectance ratio spectrum (e.g., 800) at different polarizations. This method has advantages in computational efficiency, thus enabling higher sampling rates and, consequently, higher measurement throughput. Furthermore, in specific cases where the attenuation in the upper Si layer 124 is insufficient for CWL411 to be present within the wavelength range of the broadband illumination beam, the periodicity of a portion of the acquired reflectance spectrum unaffected by the underlying structure (e.g., 113) can still allow for the determination of the thickness of the upper Si layer 124 with sufficient accuracy. However, since the Fourier transform calculation utilizes what can enable the maximum wavelength range, it is preferable that CWL411 be present in the acquired reflectance spectrum (e.g., 400A and 800). The periodic portion of the reflectance spectrum 400A (e.g., 405) preferably contains at least two to three, and preferably five or more, sufficient vibrations to ensure accurate Fourier transform peak (FT peak) calculations and thus accurate determination of the upper Si layer thickness.

[0049] The attenuation and reflection of the illumination beam from the upper Si layer 124 depend on the complex refractive index of Si, i.e., n and k, and therefore different calibrations (e.g., Figures 4 and 6) are required for different n and / or k refractive indices of the material used for the upper layer 121. In fact, n and k can vary even for the same material, e.g., Si, depending on the method and conditions for forming the upper Si layer 124 (e.g., the rest of a thinly bonded Si wafer). Calibrations performed for specific values ​​of n and k are extremely robust and reliable in that spectral noise does not significantly affect the FT peak and measured thickness. The disadvantage is that calibrations must be made for many possible values ​​of n and k.

[0050] To overcome the need for different calibrations for different complex refractive indices n and k, an "effective wavelength" can be used in the embodiments described earlier. The "effective wavelength" incorporates n and k, and the physical thickness of the upper Si layer 124 can be determined by excluding the values ​​of n and k from the calculation at the end of the calculation. For example, a scaled FT spectrum can be obtained by scaling or normalizing the horizontal axis of the FT spectrum by n and k. As a result, the position of the first-order peak in the scaled FT spectrum is independent of n and k. Such scaling can also make the spectral response from the upper Si layer 124 more periodic, and therefore the first-order peak more prominent. The advantage of this method is that it does not require multiple calibrations, but the method has the disadvantage that it is more sensitive to spectral noise, which may require the removal of parts of the acquired spectrum so that only parts with clear periodicity (e.g., 405) remain. Such a method requires many calculations because it is necessary to determine the spectral cutoff wavelength.

[0051] In practice, the FT peak or "equivalent thickness" calibration for thickness in Figure 6 can be determined using a simple computational model of the reflection of a broadband illumination beam from an absorption layer lacking a substrate structure. Linear fitting for the FT peak against thickness is usually sufficient, but higher-order fittings may also be employed. In particular, in Figure 6, plots 601, 602, 603, 604, 605, 606, 607, 608, 609, and 610 show the FT spectra of the upper Si layer 124 having thicknesses of 0.1 μm, 0.2 μm, 0.3 μm, 0.4 μm, 0.5 μm, 1 μm, 2 μm, 3 μm, 4 μm, and 5 μm, respectively. As shown in Figure 7, the position of the primary peak relates to the thickness of the upper Si layer 124.

[0052] Figure 2 shows a flowchart of process 200 for measuring film thickness according to several embodiments of the present disclosure. In step 210, the upper layer of the sample is illuminated with a broadband illumination beam. The sample includes a substrate and a plurality of semiconductor structures formed between the substrate and the upper layer. In step S220, the reflectance spectrum of the sample is obtained. In step 230, the thickness of the upper layer is determined by applying an upper layer model to the reflectance spectrum. The upper layer model is substantially unaffected by the plurality of semiconductor structures.

[0053] In some embodiments, the upper layer model is a single-layer model that is influenced by the upper layer 121 and substantially uninfluenced by the structure 113 and substrate 111. For example, the upper layer model may depend on information and knowledge of the upper layer 121 without depending on information and knowledge of the structure 113 and substrate 111. This differs from the prior art in which multilayer models are typically constructed to take into account the upper layer 121 and structure 113 (and possibly the substrate 111 as well). That is, the upper layer model may not include parameters or variables relating to structure 113 and substrate 111. Note that the upper layer 121 is a single layer in many embodiments. Therefore, the upper layer model may be a single-layer model. However, in other embodiments, the upper layer model may be applicable to a thickness-dominant layer located below the upper layer 121 or a thickness-dominant layer that is part of the upper layer 121. The thickness-dominant layer may have a much larger "effective thickness" than that of the other layers. For example, the upper layer 121 may be a thin surface roughness layer. The surface roughness layer may be a thin (e.g., about 3 nm) top layer that can be modeled or treated as an active material containing, for example, 50% air and 50% Si. While the surface roughness layer is usually negligible, in some cases, including it in the model may improve the degree of fit (GOF).

[0054] In a non-restrictive example, applying the upper layer model involves performing a Fourier transform (FT) of the reflectance spectrum 400A to obtain the FT spectrum 400B, and determining the thickness based on the primary peak 421 of the FT spectrum 400B, which has the greatest intensity in the FT spectrum 400B. For example, the position of the primary peak 421 in the FT spectrum 400B can be calibrated experimentally and / or computationally as a function of the thickness of the upper layer 121 (see, for example, Figures 6 and 7). Applying the “effective wavelength” method as discussed above, the position of the primary peak 421 in the scaled version of the FT spectrum 400B may be independent of the complex refractive indices n and k of the upper layer 121.

[0055] It should be noted that the FT spectrum 400B may further include at least one secondary peak 423 influenced by structure 113. Nevertheless, the primary peak 421 is substantially unaffected by structure 113, and its thickness is determined based on the primary peak 421 of the FT spectrum 400B, not on at least one secondary peak 423. As a result, the upper layer model is substantially unaffected by structure 113. In other words, the presence of at least one secondary peak 423 does not affect the relationship between the position of the primary peak 421 and the thickness of the upper layer 121, and is therefore irrelevant to the upper layer model.

[0056] In some embodiments, it can be assumed that the underlying structures (e.g., 113 and 111) of each selected measurement point / region on sample 100 are substantially the same, and therefore their influence on the obtained thickness is substantially the same as a constant thickness offset. In some cases, this constant thickness offset is so small that the primary peak position appears unaffected by the underlying structure. That is, the thickness identified based on the primary peak 421 may be a relative thickness rather than an absolute thickness measurement. Nevertheless, the absolute thickness can be obtained by subtracting a constant thickness offset from the relative thickness. Thus, it may be chosen to plot the absolute and / or relative thicknesses, for example, in Figure 7.

[0057] In one embodiment, the FT is performed on the entire reflectance spectrum, including both a first portion 409 and a second portion 407 of the reflectance spectrum 400A. The first portion 409 is substantially unaffected by the structure 113. The first portion 409 includes a portion 401 below the CWL and a portion 405 dominated by constructive and destructive interference of poorly attenuated illumination beams in the upper layer 121. The second portion 407 is affected by the structure 113. The second portion 407 of the reflectance spectrum 400A may appear in the FT spectrum 400B as at least one secondary peak 423 independent of the upper layer model.

[0058] In another embodiment, the FT is performed only on the first portion 409 which is substantially unaffected by the structure 113. Thus, the second portion 407 can be trimmed before determining the thickness of the upper layer 121. In yet another embodiment, the portion 401 below the CWL 411 can be trimmed from the first portion 409 so that it contains only the portion 405 for the FT.

[0059] It should be understood that the FT does not need to be performed strictly according to the divisions in Figure 4A. Instead, the FT can be performed flexibly over wavelength ranges such as 410-590 nm, 373-688 nm, 296-745 nm, 217-863 nm, etc. Preferably, the FT is performed over at least wavelength ranges that include CWL411 and at least two (e.g., 2, 3, 4, 5, 6, 7, etc.) vibrations to the right of CWL411 in the reflectance spectrum 400A. For example, the FT may be performed over at least a portion 405 of the reflectance spectrum 400A, preferably at least a first portion 409, preferably the whole. Furthermore, it should be noted that CWL411 may vary depending on the material and / or thickness of the upper layer 121. Thus, the various ranges considered herein for performing the FT are merely illustrative and not limiting.

[0060] In some embodiments, sample 100 can be repositioned relative to a broadband illumination beam so that the upper layer 121 of sample 100 can be illuminated in different regions by the broadband illumination beam. Thus, a different reflectance spectrum of sample 100 can be obtained, and by applying the upper layer model to the different reflectance spectrum, a different thickness of the upper layer 121 in different regions can be obtained. Furthermore, multiple thicknesses of the upper layer 121 can be specified in multiple regions in an XY pattern or R-θ pattern over part or all of sample 100.

[0061] In the example in Figure 4A, the reflectance spectrum 400A shows the reflectance coefficient against wavelength. In the example in Figure 8, the reflectance spectrum 800 shows the ratio of reflectances collected by (at least) two reflected beams of substantially orthogonal polarization. It should be understood that FT can be performed on the reflectance spectrum 800 as well as on the reflectance spectrum 400A. The explanation is provided above and is omitted here for brevity.

[0062] Figures 3A and 3B show schematic and perspective views, respectively, of the apparatus 300 according to several embodiments of the present disclosure. For example, the apparatus 300 may be used to perform a process 200 as discussed above. As shown, the apparatus 300 includes a handling stage 321 configured to receive a sample 100. A light source 301 is configured to emit a broadband illumination beam onto the sample 100.

[0063] The optical system is configured to guide a broadband illumination beam to illuminate the upper layer 121 of the sample 100, collect the reflected beam from the upper layer 121, and guide the reflected beam to the photodetector 339. In particular, the optical system may include an illumination optical fiber 302 so that the light source 301 and the illumination optical fiber 302 can function as a fiber illuminator. The optical system may also include a knife-edge prism (KEP) 303 having a first surface 303a and a second surface 303b. The optical system may further include a Schwarzschild objective mirror 310 including a primary mirror 311 and a secondary mirror 315.

[0064] In some embodiments, the primary mirror 311 may have larger dimensions than the secondary mirror 315. The primary mirror 311 has an aperture 313 (e.g., a central aperture) such that a broadband illumination beam can be directed through the aperture 313 of the primary mirror 301 by the first surface 303A of the KEP 303 to the secondary mirror 315. The secondary mirror 315 then reflects the broadband illumination beam to the primary mirror 311, which then reflects the broadband illumination beam to the upper layer 121 of the sample 100. The reflected beam from the upper layer 121 is then received by the primary mirror 311 and reflected to the secondary mirror 315. The reflected beam is then reflected by the secondary mirror 315, passes through the aperture 313 of the primary mirror 311 to the second surface 303b of the KEP 303, which can then direct the reflected beam to the photodetector 339.

[0065] In non-limiting examples, the apparatus 300 may be configured as a spectrophotometer, such as an absorption spectrophotometer (aSR). A broadband light source covering the ultraviolet (UV), visible (VIS), and near-infrared (NIR) portions of the light spectrum may be used. For example, a wavelength range of 200 to 1000 nm may be used. For specific applications, it may be necessary to further extend the wavelength range to the vacuum ultraviolet (VUV) and infrared (IR) portions of the light spectrum to adapt to thin or thick films or films with lower or higher absorption than usual. In other words, the light source 301 may be configured to emit a broadband illumination beam in a wavelength range defined by a lower and upper limit. The lower limit may be 0.1 nm, 1 nm, 10 nm, 100 nm, 200 nm, 300 nm, 400 nm, 500 nm, 600 nm, 700 nm, 800 nm, or any value in between. The upper limits may be 25,000nm, 15,000nm, 3,000nm, 1,000nm, 900nm, 800nm, 700nm, 600nm, 500nm, 400nm, 300nm, 200nm, 100nm, or any value in between.

[0066] In some embodiments, the light source 301 includes a laser-driven xenon (Xe) light source (LDL), a laser, a laser diode, a light-emitting diode, a gas discharge light source, a laser-driven light source, or any combination thereof. The light source 301 may include any suitable combination of lasers, LEDs, gas discharge light sources, etc. Preferably, the light source 301 includes an LDL. The light source 301 may be continuous (CW) or pulsed. The broadband illumination beam is guided to the upper Si layer 124 by an illumination optical system which may include optical fibers (e.g., 302) to enable remote placement of the light source 301, system aperture 305, and focusing lens (e.g., 310). The purpose of the focusing optical system is to focus the illumination beam onto the upper Si layer 124 at incident angles from 0° (perpendicular incidence) to 85°, for example, 0°, 5°, 10°, 20°, 30°, 40°, 45°, 50°, 60°, 70°, 80°, 85° and any value in between. The focusing optical system may include a Schwarzschild objective mirror (e.g., 310) that uses reflective optical systems (e.g., 311 and 315) to focus the illumination beam. Using a suitable mask, it may be possible to select a range of incident angles for illuminating or modulating the intensity level of a portion of the illumination beam with different incident angles. Upon reflection and / or refraction from the upper Si layer 124, a reflected beam is formed, which is captured by the focusing optical system and transmitted to the photodetector 339. Using KEP303, the illumination beam can be directed towards the focusing optical system (e.g., 310) and the reflected beam can be directed from the focusing optical system (e.g., 310). The use of a suitable mask allows for specular reflection from the upper Si layer 124 and collection of scattered light within a given range of reflection angles. In some embodiments, KEP303 may be implemented in the form of two mirrors, as shown in Figure 3B.

[0067] In some embodiments, the fiber illuminators of the illumination optics (e.g., 302 and 301) are imaged onto a broadband high-resolution spectrometer (e.g., 339) via focusing and converging optics to measure spectral reflectance. The detection optics may include a system aperture (e.g., 331) and a converging lens (e.g., 335) for focusing the reflected beam onto the input slit of the spectrometer (e.g., 339) or an optical fiber (e.g., 337). The detection optics may also include a beam splitter 333.

[0068] The focusing optical system may further include at least one polarizer that forms a polarized reflected beam, as needed. A movable or rotatable polarizer may be used to change the polarization orientation. The inventors have found that the use of polarization improves measurement performance when the upper Si surface (e.g., 123) is not perfectly smooth, for example due to roughness, grinding marks remaining from the grinding process of the bonded upper Si wafer, etc. The inventors have also found that the use of polarization improves measurement when the underlying structure (e.g., 113 and 111) is highly anisotropic or when measuring wafers having substantial roughness on the Si upper surface 123. The focusing beam may be linearly polarized, circularly polarized, etc. In some embodiments, the polarizer may be, for example, inserted into the illumination beam or as part of an illumination optical system integrated with the light source 301. An additional advantage of using a polarizer is that it allows for the identification of a critical wavelength (CWL) as the lower limit of the periodic portion of the reflectance spectrum, as the point at which the reflectance ratio obtained with orthogonal polarization deviates from 1. This allows for the non-computationally intensive removal of the non-periodic portion of the reflection spectrum below the CWL, enabling the identification of the Fourier transform peak (FT) with improved accuracy.

[0069] The illumination and reflected beams may have planar or circular wavefronts. In other embodiments, the beams can be modulated over a range of incident and reflection angles using phase masks positioned within the illumination beam path, the reflected beam path, or both. For example, a circular beam with a helical wavefront can be formed using a 2Pi phase mask positioned within the illumination beam. Such a beam may contain optical singularities that can be advantageously used to achieve an optimal refractive power distribution within the upper Si layer 124 and at the interface (e.g., 122) between the upper Si layer 124 and the substrate structure (e.g., 113). In addition, as those skilled in the art will understand, the output of the light source 301 can be monitored by employing a reference channel on a spectrometer (e.g., 339), so that variations in optical output can be excluded from calculations from the measured reflection spectrum.

[0070] In one embodiment, the apparatus 300 is mounted on a platform which can be integrated with a wafer processing module 341. The platform may include a wafer loading device and a wafer handling stage (e.g., 321) that can move and position a semiconductor wafer relative to an optical device so that the optical device can perform multiple consecutive measurements over a large portion of the wafer. The wafer handling stage may have rotational and translational degrees of freedom that enable measurements over at least 95% of the wafer surface with substantially uniform wafer coverage. Depending on the degrees of freedom of the wafer handling stage (e.g., 321) and the degrees of freedom of motion of the optical device (if so equipped), various sampling patterns such as measurements on an XY grid, an R-θ pole grid, etc., may be utilized. The wafer processing module 341 may include, for example, a processing chamber configured to perform a surface treatment on a sample 100 in a vacuum. Surface treatments include, but are not limited to, film deposition, etching, lithography patterning, doping, cleaning, heating, chemical mechanical polishing, etc.

[0071] The plasma processing module 341 may include a plasma processing chamber that is a capacitively coupled plasma processing apparatus, an inductively coupled plasma processing apparatus, a microwave plasma processing apparatus, a radial line slot antenna (RLSA™) microwave plasma processing apparatus, an electron cyclotron resonance (ECR) plasma processing apparatus, or other types of processing systems or combinations of systems. Therefore, it will be understood by those skilled in the art that the techniques described herein can be used in any of the diverse plasma processing systems. The plasma processing chamber can be used for a variety of operations, including but not limited to etching, deposition, cleaning, plasma polymerization, plasma chemical vapor deposition (PECVD), atomic layer deposition (ALD), atomic layer etching (ALE), etc. The structure of the plasma chamber is well known to those skilled in the art. It will be recognized that different and / or additional plasma process systems can be implemented while still utilizing the techniques described herein.

[0072] In some embodiments, the apparatus 300 and platform are used to perform multiple measurements across the wafer surface, with the thickness of the upper Si layer measured at each location. The set of measured thickness values ​​is then used to interpolate over unmeasured locations to generate a wafer map. The wafer map is then transferred to a processing tool, which can be used to perform wafer processing operations that require knowledge of the upper Si layer thickness distribution. In one example, the processing tool may selectively remove portions of the upper Si layer 124 to improve the uniformity of the upper Si layer thickness across the wafer. The number of measurements can vary, for example, from 5 to over 100,000 measurements per wafer (e.g., 5, 50, 100, 500, 1,000, 5,000, 10,000, 100,000 and any value in between). A high measurement density allows for the detection of "imprints" of the underlying structure and die edges within the wafer map, thereby enabling the alignment of measurement points to points within individual dies.

[0073] In one embodiment, the apparatus 300 may further include a controller 343. The components of the apparatus 300 can be connected to and controlled by the controller 343, which can optionally be connected to a corresponding memory storage unit and a user interface (not shown). Various processing operations (e.g., film thickness measurement operations and plasma operations) can be performed via the user interface, and various processing recipes and operations can be stored in the storage unit. Thus, a given substrate can be processed in the aforementioned plasma chamber by various microfabrication techniques.

[0074] The controller 343 is coupled to various components of the apparatus 300 to receive inputs from various components, including, but not limited to, the light source 301, illumination optical fiber 302, KEP 303, system apertures 305 and 331, Schwarzschild objective lens 310, handling stage 321, wafer processing module 341, beam splitter 333, focusing lens 335, collecting optical fiber 337 and / or photodetector 339, and to provide outputs to them. For example, the controller 343 may be configured to receive data from the photodetector 339 and determine the thickness of the upper layer 121 based on the reflectance spectrum by applying an upper layer model to the reflectance spectrum, for example by performing a Fourier transform. Of course, such a function can also be achieved manually.

[0075] The controller 343 may also be configured to adjust knobs and control settings for these components. For example, the handling stage 321 may be an XY, XYZ, R-θ, or R-θ-Z stage, and the controller 343 may be configured to move and / or rotate the sample 100 using adjustment knobs and control settings for the handling stage 321. Naturally, such adjustments may also be made manually.

[0076] The controller 343 can be implemented in various ways. In one example, the controller 343 is a computer. In another example, the controller 343 includes one or more programmable integrated circuits programmed to provide the functions described herein. For example, one or more processors (e.g., microprocessors, microcontrollers, central processing units, etc.), programmable logic devices (e.g., composite programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), etc.) and / or other programmable integrated circuits can be programmed with software or other programming instructions to implement the functions of a defined plasma process recipe. The software or other programming instructions may be stored in one or more non-temporary computer-readable media (e.g., memory storage devices, flash memory, DRAM memory, reprogrammable storage devices, hard drives, floppy disks, DVDs, CD-ROMs, etc.), and it should be further noted that when the software or other programming instructions are executed by the programmable integrated circuit, they cause the programmable integrated circuit to perform the processes, functions and / or capabilities described herein. Other variations may also be implemented.

[0077] The above description is provided in relation to the use of the described measuring apparatus and method for determining the remaining thickness of a bonded semiconductor wafer. It will be understood that the same apparatus and method are applicable to other thin film thickness measurement situations, provided the film exhibits sufficient absorption of the illumination beam.

[0078] The above description provides specific details, such as the specific shape of the processing system and the various components and processes used therein. However, it should be understood that the technology described herein may be implemented in other embodiments that deviate from these specific details, and such details are for illustrative purposes only, not limitation. Embodiments disclosed herein have been described with reference to the accompanying drawings. Similarly, specific figures, materials, and configurations have been given for illustrative purposes to ensure a good understanding. However, embodiments may be implemented without such specific details. Components having substantially the same functional structure are indicated by similar reference numerals, and redundant descriptions may be omitted.

[0079] To aid in understanding the various embodiments, various techniques have been described as multiple distinct operations. The order of description should not be interpreted as meaning that these operations are necessarily order-dependent. In fact, these operations do not need to be performed in the order presented. The described operations may be performed in a different order than in the described embodiments. In additional embodiments, various additional operations may be performed, and / or described operations may be omitted.

[0080] As used herein, “substrate” or “wafer” generally refers to an object processed according to the present invention. A substrate may include any material portion or structure of an element, in particular a semiconductor or other electronic element, and may be a base substrate structure such as a semiconductor wafer, or a layer on or superimposed on a base substrate structure such as a reticle or thin film. Therefore, a substrate is not limited to any particular base structure, underlying layer or superimposed layer, whether patterned or unpatterned, but is considered to include any combination of such layer or base structure and layers and / or base structures. The description may refer to certain types of substrates, but this is for illustrative purposes only.

[0081] The substrate may be any suitable substrate, such as a silicon (Si) substrate, a germanium (Ge) substrate, a silicon-germanium (SiGe) substrate, and / or a silicon-on-insulator (SOI) substrate. The substrate may contain a semiconductor material, such as a group IV semiconductor, a group III-V compound semiconductor, or a group II-VI oxide semiconductor. The group IV semiconductor may include Si, Ge, or SiGe. The substrate may be a bulk wafer or an epitaxial layer.

[0082] Those skilled in the art will understand that many variations of the operation of the technology described above can be made while achieving the same objectives as the present invention. Such variations are intended to be included within the scope of this disclosure. Therefore, the above description of embodiments of the present invention is not intended to be limiting. Rather, any limitations on embodiments of the present invention are presented in the following claims.

Claims

1. A method for measuring film thickness, The method involves irradiating the upper layer of a sample in a first region with a broadband illumination beam, wherein the sample includes a substrate and a plurality of semiconductor structures formed between the substrate and the upper layer. Obtaining a first reflectance spectrum of the sample in the first region, This includes determining the first thickness of the upper layer in the first region by applying the upper layer model to the first reflectance spectrum, The method wherein the upper layer model is substantially unaffected by the multiple semiconductor structures.

2. Applying the aforementioned upper layer model means The process involves performing a Fourier transform (FT) on the first reflectance spectrum to obtain the FT spectrum, The method according to claim 1, comprising determining the first thickness based on the primary peak of the FT spectrum having the maximum intensity in the FT spectrum.

3. The method according to claim 2, wherein the FT is performed on at least a first portion of the first reflectance spectrum that is substantially unaffected by the plurality of semiconductor structures.

4. The method according to claim 3, wherein the FT is performed on both the first portion and the second portion of the first reflectance spectrum that is affected by the plurality of semiconductor structures.

5. The FT spectrum further includes secondary peaks affected by the plurality of semiconductor structures, The aforementioned primary peak is substantially unaffected by the multiple semiconductor structures, and The method according to claim 4, wherein the first thickness is determined based on the primary peak of the FT spectrum, and not based on the secondary peak, such that the upper layer model is substantially unaffected by the plurality of semiconductor structures.

6. The method according to claim 3, further comprising truncating a second portion of the first reflectance spectrum that is affected by the plurality of semiconductor structures before determining the first thickness of the upper layer.

7. The method according to claim 6, further comprising truncating the first portion below the critical wavelength.

8. The method according to claim 2, further comprising experimentally or computationally calibrating the position of the primary peak of the FT spectrum with respect to the first thickness.

9. The method according to claim 8, further comprising scaling the horizontal axis of the FT spectrum by the complex refractive indices n and k of the upper layer to obtain the scaled FT spectrum such that the position of the primary peak of the scaled FT spectrum is independent of n and k.

10. Obtaining the first reflectance spectrum of the sample is Collecting at least two reflected beams of substantially orthogonal polarization, The method according to claim 1, comprising obtaining the first reflectance spectrum as the ratio of reflectances collected by the at least two reflected beams.

11. The method according to claim 10, further comprising specifying the critical wavelength at which the ratio begins to differ from 1.

12. The first reflectance spectrum of the sample is a reflection coefficient spectrum, The aforementioned method, The method according to claim 1, further comprising identifying a critical wavelength which is the boundary between two regions of the reflection coefficient spectrum.

13. Rearranging the aforementioned sample, the aforementioned broadband illumination beam, or a combination thereof, The upper layer of the sample is irradiated in a second region with the broadband illumination beam, Obtaining a second reflectance spectrum of the sample in the second region, The method according to claim 1, further comprising applying the upper layer model to the second reflectance spectrum to determine the second thickness of the upper layer in the second region.

14. The method according to claim 13, further comprising specifying a plurality of thicknesses of the upper layer in a plurality of regions using an X-Y pattern or an R-θ pattern over at least a portion of the sample.

15. The upper layer includes a semiconductor material, and The method according to claim 1, wherein the broadband illumination beam has a wavelength range of 200 to 1000 nm.

16. A handling stage configured to receive a sample, wherein the sample includes a substrate, an upper layer, and a plurality of semiconductor structures formed between the substrate and the upper layer. A light source configured to emit a broadband illumination beam, An optical system configured to guide the broadband illumination beam to illuminate the upper layer of the sample, collect the reflected beam from the upper layer, and guide the reflected beam to a photodetector configured to acquire the reflectance spectrum of the sample, An apparatus comprising: a controller configured to determine the thickness of the upper layer by applying the upper layer model to the reflectance spectrum, The aforementioned upper layer model is a device that is substantially unaffected by the multiple semiconductor structures.

17. The optical system includes a Schwarzschild objective lens and a knife-edge prism (KEP), The Schwarzschild objective lens includes a primary mirror and a secondary mirror, The KEP has a first surface and a second surface, The Schwarzschild objective and the KEP are configured such that the broadband illumination beam is directed by the first surface of the KEP through the aperture of the primary mirror to the secondary mirror, reflected by the primary mirror, and then reflected back to the upper layer of the sample, and The apparatus according to claim 16, wherein the Schwarzschild objective mirror and the KEP are configured such that the reflected beam from the upper layer is reflected by the primary mirror to the secondary mirror, passes through the aperture of the primary mirror and is reflected to the second surface of the KEP, and is then directed toward the photodetector.

18. The apparatus according to claim 16, wherein the optical system includes at least one selected from the group consisting of an optical fiber, a focusing lens, a system aperture, and a polarizer.

19. The handling stage is an X-Y, X-Y-Z, R-θ, or R-θ-Z stage, and The apparatus according to claim 16, wherein the light source includes at least one selected from the group consisting of a laser, a laser diode, a light-emitting diode, a gas discharge light source, and a laser driving light source.

20. The apparatus according to claim 16, further comprising a processing chamber configured for performing surface treatment on the sample.

Images