Wafer morphology scanning signal processing method and computer program product

Abstract

The invention discloses a wafer morphology scanning signal processing method and a computer program product, and relates to the technical field of semiconductor manufacturing. The wafer morphology scanning signal processing method comprises the following steps: acquiring an original scanning signal of a target wafer surface morphology, and performing multi-resolution decomposition on the original scanning signal through wavelet transform to obtain a noise frequency band signal and a non-noise frequency band signal, namely decomposing the original scanning signal into different frequency band signals through multi-resolution decomposition, according to the method, the noise and the real feature of the surface morphology of the target wafer are separated, then signal denoising is performed on the noise frequency band signal related to the noise, and finally, the noise-removed reconstructed scanning signal can be obtained through inverse wavelet transform based on the noise frequency band signal after signal denoising and the non-noise frequency band signal. The reconstructed scanning signal can accurately reflect the real characteristics of the surface topography of the target wafer.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of semiconductor manufacturing, and particularly relates to a wafer topography scanning signal processing method and a computer program product. BACKGROUND

[0002] In the field of semiconductor manufacturing, accurate measurement of wafer surface topography is crucial to ensure chip quality and performance. However, due to the presence of instrument noise (such as measurement instruments), the measurement signal of wafer surface topography contains a large amount of interference information, which can mask the true characteristics of wafer surface topography, thereby reducing the accuracy of the measurement signal and increasing the difficulty of subsequent data analysis based on the measurement signal.

[0003] Currently, traditional noise identification and filtering techniques (such as filters) are generally used to process measurement signals to reduce interference information in the measurement signal. However, this traditional signal processing method cannot effectively distinguish noise from the true characteristics of wafer surface topography, resulting in a measurement signal after signal processing that cannot accurately reflect the wafer surface topography. SUMMARY

[0004] The embodiments of the present application provide a wafer topography scanning signal processing method and a computer program product, which can effectively separate noise from the true characteristics of wafer surface topography, thereby performing directional noise removal to improve the accuracy of the reconstructed scanning signal after signal processing.

[0005] In a first aspect, the present application provides a wafer topography scanning signal processing method, characterized in that it comprises:

[0006] obtaining an original scanning signal of a target wafer surface topography;

[0007] performing multi-resolution decomposition on the original scanning signal through wavelet transform to obtain a noise band signal and a non-noise band signal;

[0008] performing signal denoising on the noise band signal;

[0009] obtaining a reconstructed scanning signal through inverse wavelet transform according to the noise band signal after signal denoising and the non-noise band signal.

[0010] In a second aspect, the present application provides a wafer topography scanning signal processing device, comprising:

[0011] a signal acquisition module for obtaining an original scanning signal of a target wafer surface topography;

[0012] a signal decomposition module configured to perform multi-resolution decomposition on the original scanning signal by wavelet transform to obtain a noise band signal and a non-noise band signal;

[0013] a signal denoising module configured to perform signal denoising on the noise band signal;

[0014] a signal reconstruction module configured to obtain a reconstructed scanning signal by inverse wavelet transform based on the noise band signal after signal denoising and the non-noise band signal.

[0015] In a third aspect, an electronic device is provided. The device includes a memory and a program or instructions stored on the memory and executable on the processor, and the program or instructions, when executed by the processor, implement the wafer topography scanning signal processing method according to any one of the above aspects.

[0016] In a fourth aspect, a readable storage medium is provided. The readable storage medium stores a program or instructions, and the program or instructions, when executed by a processor, implement the wafer topography scanning signal processing method according to any one of the above aspects.

[0017] In a fifth aspect, a computer program product is provided. The instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the wafer topography scanning signal processing method according to any one of the above aspects.

[0018] The technical solutions provided in the embodiments of the present application at least have the following beneficial effects:

[0019] In the wafer topography scanning signal processing method provided in the embodiments of the present application, the original scanning signal of the target wafer surface topography is obtained, and the original scanning signal is decomposed into different frequency band signals by multi-resolution decomposition to obtain a noise band signal and a non-noise band signal, so that the separation between noise and the real features of the target wafer surface topography is realized, and then the signal denoising is performed on the noise band signal related to noise, and finally the reconstructed scanning signal without noise is obtained by inverse wavelet transform based on the noise band signal after signal denoising and the non-noise band signal, and the reconstructed scanning signal can accurately reflect the real features of the target wafer surface topography. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced. Those skilled in the art can obtain other drawings according to these drawings without creating creative labor.

[0021] Figure 1 is a flowchart of a wafer topography scanning signal processing method provided by an embodiment of the present application;

[0022] Figure 2 is a schematic diagram of a multi-resolution decomposition principle provided by an embodiment of the present application;

[0023] Figure 3 is a schematic diagram of a position distribution of D1 to D7 different waveband components provided by an embodiment of the present application;

[0024] Figure 4 is a schematic diagram of a size relationship between an energy proportion of each of D1 to D7 decomposition frequency band signals and a preset energy threshold provided by an embodiment of the present application;

[0025] Figure 5 is a schematic diagram of a principle of determining process warning information provided by an embodiment of the present application;

[0026] Figure 6 is a schematic diagram of a principle of a wafer topography scanning signal processing method provided by an embodiment of the present application;

[0027] Figure 7 is a schematic diagram of a simulated actual LER signal and a noise signal provided by an embodiment of the present application;

[0028] Figure 8 is a schematic diagram of a simulated actual LER signal and an LER signal after removing instrument noise provided by an embodiment of the present application;

[0029] Figure 9 is a structural schematic diagram of a wafer topography scanning signal processing device provided by an embodiment of the present application;

[0030] Figure 10 is a schematic diagram of a wafer topography scanning signal processing device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0031] The features and exemplary embodiments of various aspects of the present application will be described in detail below, in order to make the purposes, technical solutions and advantages of the present application clearer and more apparent, the present application will be further described in detail below in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, but not to limit the present application. The present application can be implemented without some of these specific details by those skilled in the art. The following description of the embodiments is only to provide a better understanding of the present application by showing examples of the present application.

[0032] It is to be noted that the relative terms such as first and second and the like in this context are used only to differentiate one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between such entities or operations. Moreover, the terms "comprising", "including", or any other variant thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... " does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0033] It should be noted that the acquisition, storage, use, processing, etc. of data in the technical solutions of the present application comply with the relevant provisions of national laws and regulations. In the embodiments of the present application, some industry existing solutions, components, models, etc. may be mentioned, which should be considered as exemplary, and the purpose is only to illustrate the feasibility in the implementation of the technical solutions of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0034] First, the noun terms related to one or more embodiments of the present application are explained.

[0035] Wavelet Transform (WT) is a signal analysis method that decomposes a signal into different scale (frequency) components.

[0036] Signal denoising refers to the process of extracting a target signal from a signal disturbed by noise.

[0037] Inverse Wavelet Transform (IWT) is the inverse process corresponding to Wavelet Transform, which is used to recombine the signals decomposed by Wavelet Transform.

[0038] Line Edge Roughness (LER) is a parameter used to quantify the quality of wafer surface topography, especially the irregularity of the line edge of the lithographic pattern formed in the semiconductor manufacturing process.

[0039] Semiconductor process mechanism refers to the physical, chemical or engineering process involved in the semiconductor manufacturing process.

[0040] Approximation component is the low-frequency part of a signal, reflecting the overall trend of the signal.

[0041] Detail component is the high-frequency part of a signal, reflecting the local mutation of the signal.

[0042] Soft threshold processing is a processing method for scaling signal coefficients.

[0043] Atomic Force Microscope (AFM) is a kind of high-resolution microscope used for researching and measuring the surface topography of a substance.

[0044] Secondly, in the field of semiconductor manufacturing, the accurate measurement of wafer surface topography is crucial to ensure the quality and performance of chips. However, due to the existence of instrument noise, the measurement signal of wafer surface topography contains a large amount of interference information, which will mask the true characteristics of wafer surface topography, thereby reducing the accuracy of the measurement signal and increasing the difficulty of subsequent data analysis based on the measurement signal.

[0045] Currently, the measurement signal is generally processed by traditional noise identification and filtering technology (such as filter) to reduce the interference information in the measurement signal. However, this traditional signal processing method is difficult to realize effective separation between noise and the true characteristics of wafer surface topography, thereby causing the measurement signal after signal processing to be unable to accurately reflect the wafer surface topography.

[0046] In view of the above technical problems, the present application provides a wafer topography scanning signal processing method and a computer program product. In the wafer topography scanning signal processing method provided in the embodiments of the present application, the original scanning signal of the target wafer surface topography is first obtained, and the original scanning signal is decomposed into noise band signal and non-noise band signal through wavelet transform. Secondly, the signal denoising is performed on the noise band signal related to the noise, and finally, based on the noise band signal and the non-noise band signal after signal denoising, the reconstructed scanning signal without noise can be obtained through inverse wavelet transform. The reconstructed scanning signal after signal processing reflects the true characteristics of the target wafer surface topography with high accuracy.

[0047] For example, the wafer topography scanning signal processing method provided in the embodiments of the present application can be applied to the production line of semiconductor manufacturing enterprises for quality evaluation of the surface topography of the produced wafer. In actual application, the original scanning signal of wafer surface topography is obtained, and the original scanning signal is decomposed into noise band signal and non-noise band signal through wavelet transform, and then the signal denoising can be performed on the noise band signal, and based on the noise band signal and the non-noise band signal after signal denoising, the reconstructed scanning signal is obtained through inverse wavelet transform, and the reconstructed scanning signal is stored in a data storage device. The operator can perform various mathematical analysis or quality evaluation based on the reconstructed scanning signal, such as defect detection of surface topography, line edge roughness, etc.

[0048] For example, in an ideal case, the line edge of a lithography pattern on a wafer should be smooth, but in actual processing, due to the non-uniformity of the light source, the chemical properties of the photoresist, the fluctuation of the exposure and development process, etc., the line edge irregularity will be caused. The line edge roughness (LER) can be used to quantify the irregularity of the line edge, so as to help optimize the processing technology of the semiconductor by evaluating the quality of the wafer processing technology. Therefore, by the wafer topography scanning signal processing method provided in the embodiments of the present application, the reconstructed scanning signal accurately reflecting the real characteristics of the wafer surface topography can be obtained, and the LER is determined based on the reconstructed scanning signal, so as to improve the quantization accuracy of the line edge irregularity.

[0049] It should be noted that the application scenarios described in the above embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. It is known to those skilled in the art that, with the emergence of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems. The wafer topography scanning signal processing method provided by the embodiments of the present application can be applied to various application scenarios that require multi-resolution decomposition of the scanning signal and accurate noise suppression of the noise band signal obtained after multi-resolution decomposition.

[0050] The wafer topography scanning signal processing method provided by the embodiments of the present application will be introduced below. In actual application, the execution subject of the wafer topography scanning signal processing method of the embodiments of the present application can be a terminal device, such as a desktop computer, a notebook computer, etc., or a remote device similar to a server. Of course, the embodiments of the present application can also adopt an execution subject in the form of a presentation software, such as a client, a software program, etc. installed in a terminal device, and the specific type of the execution subject corresponding to the technical solutions provided by the embodiments of the present application is not strictly limited herein, and can be flexibly selected according to actual application scenarios and actual needs.

[0051] The specific embodiments of the wafer topography scanning signal processing method, device, electronic equipment, storage medium and computer program product provided by the embodiments of the present application will be introduced below. First, a wafer topography scanning signal processing method is introduced.

[0052] Figure 1 The flowchart of the wafer topography scanning signal processing method provided by the embodiments of the present application is shown in FIG. 1. As shown in FIG. 1, the method includes steps S100 to S103. Figure 1

[0053] S100: Obtain the original scanning signal of the target wafer surface topography.

[0054] ​In one or more embodiments of the present application, in order to obtain the noise band signal and the non-noise band signal by multi-resolution decomposition in the subsequent steps, and then realize the targeted signal denoising of the noise band signal. In this step, the present application needs to obtain the original scanning signal of the target wafer surface topography, and the target wafer refers to the wafer to be processed.

[0055] Specifically, the original scanning signal of the target wafer surface topography is obtained.

[0056] It should be noted that in order to avoid random errors and improve the accuracy and reliability of the original scanning signal, repeated measurements are generally performed, and the number of repeated measurements is not limited in the present application, such as 5-10 times. In the present application, the specific way of obtaining the original scanning signal is not limited, which can be set according to the actual demand. For example, the surface topography of the target wafer is scanned by a scanning electron microscope (SEM) or an atomic force microscope (AFM). The AFM scans the surface topography of the target wafer by using a probe, and obtains the topography information of the target wafer surface by the interaction force between the probe and the target wafer surface. In one or more embodiments of the present application, the probe tip size of the AFM is calibrated, and the surface topography of the target wafer is scanned by the AFM according to the pre-set sampling interval and scanning length, and the original scanning signal is obtained. Of course, the specific size of the sampling interval and the scanning length is not limited in the present application, which can be set according to the actual demand, and in order to reconstruct the signal without distortion in the subsequent steps, the Nyquist sampling theorem needs to be met when scanning the surface topography of the target wafer, which stipulates that in order to reconstruct the signal without distortion, the sampling frequency is at least twice the highest frequency component in the signal.

[0057] The constraint conditions of the sampling interval and the scanning length are as follows:

[0058] f s ≥2f max (1)

[0059] Δx≥λ min / 2 (2)

[0060] Δx≥λ min / 3 (3)

[0061] In the above formula, f s is the sampling frequency, f max is the highest frequency of the signal; Δx is the sampling interval, and λ minThe minimum feature size refers to the size of the smallest structure that can be distinguished in the signal, such as the minimum line width, the minimum line spacing, and the like. Formula (1) constrains the sampling frequency to be at least twice the highest frequency, so in the spatial domain, the sampling interval is at least half of the minimum feature size, as shown in formula (2). In addition to noise and instrument limitations, in actual applications, a smaller sampling interval is usually selected to improve the accuracy and reliability of the measurement, that is, in one or more embodiments of the present application, formula (3) can be used as a constraint condition for the sampling interval. For example, if the minimum feature size of the wafer surface topography to be measured is 5 nm, the sampling interval should satisfy the constraint condition of not greater than 1.67 nm.

[0062] In the present application, the specific size of the scan length is not limited, and can be set according to actual needs, such as based on the required measurement accuracy and the limitations of the measuring instrument to determine the scan length. Since a smaller scan length will cause the measurement signal to be affected by noise and random errors, and a too long scan length will generate a large amount of data and require a long time, in one or more embodiments of the present application, the scan length is generally not less than 10 times the correlation length and generally not less than 2 microns. Taking the application scenario of determining the line edge roughness as an example, when the AFM is used to measure the target wafer surface topography, the scan length needs to be not less than 10 times the correlation length and generally not less than 2 microns, wherein the correlation length is the distance at which the autocorrelation function of the edge profile decays to 1 / e, reflecting the spatial distribution characteristics of the roughness. For example, in the photolithography process, the correlation length is usually 20-200 nanometers. If the correlation length is 50 nanometers, the minimum scan length needs to be not less than 0.5 microns, but in order to meet the universality, 2 microns is generally selected as the scan length.

[0063] S101: performing multi-resolution decomposition on the original scanning signal by wavelet transform to obtain a noise band signal and a non-noise band signal.

[0064] In one or more embodiments of the present application, in order to achieve targeted signal denoising in subsequent steps, in this step, the present application needs to perform multi-resolution decomposition on the original scanning signal obtained in step S100 by wavelet transform to realize signal separation of the true features reflecting the wafer surface topography and the noise in the original scanning signal, and obtain a noise band signal and a non-noise band signal.

[0065] Specifically, the original scanning signal obtained in step S100 is decomposed into a noise band signal and a non-noise band signal by wavelet transform, as follows:

[0066] Firstly, the original scanning signal is decomposed into decomposed band signals in different frequency and time scales by using Discrete Wavelet Transform (DWT), i.e. the original signal is recursively decomposed into approximation components and detail components by DWT.

[0067] As shown in the principle diagram of multi-resolution decomposition provided by the embodiment of the present application, the process of recursively decomposing the original scanning signal 200 into approximation components 201 and detail components 202 by using low-pass filters and high-pass filters is indicated in the principle diagram. Figure 2

[0068] The process of obtaining the approximation components and the detail components by multi-resolution decomposition can refer to the following formula:

[0069]

[0070]

[0071] f(x)=Af((x))+Df((x)) (6)

[0072]

[0073] In the above formula, Ψ j,k (x) is a wavelet function, φ j,k (x) is a scaling function, k represents a position / time, is a translation factor, x is a value of the signal at a spatial position; j represents a wavelength / frequency, is a scale factor; Af((x)) represents an approximation component, and Df((x)) represents a detail component DWT realizes the decomposition of the original scanning signal by recursively applying low-pass filters and high-pass filters, i.e. the low-frequency component (approximation component) of the signal is extracted by the low-pass filter, and the high-frequency component (detail component) of the signal is extracted by the high-pass filter; h0(k) is a system coefficient of the high-pass filter of the wavelet function, and h1(k) is a system coefficient of the low-pass filter of the scaling function. Formula (9) is a function formula of signal reconstruction, A j,k is an approximation component, and D j,k is a detail component.

[0074] It should be noted that the wavelet basis selected in the wavelet transform is not limited in the present application, and can be set according to actual needs. For example, a biorthogonal wavelet can be selected, and a 5th-order vanishing moment can effectively match the frequency band of noise, and simultaneously has a symmetric filter bank to avoid phase distortion generated in multi-resolution decomposition.

[0075] Secondly, the decomposed band signals in different frequency scales obtained by multi-resolution decomposition include noise band signals and non-noise band signals.

[0076] ​It should be noted that the decomposition level of the multi-resolution decomposition is not limited in the present application, and can be set according to actual needs. Since the wavelength analysis range covered by the semiconductor process mechanism in the semiconductor manufacturing field is 2-256 nm, in one or more embodiments of the present application, the present application can determine a plurality of decomposition frequency bands of the original scanning signal according to the correspondence between different fluctuation frequency bands and wavelength analysis ranges in the preset semiconductor process mechanism, and perform multi-resolution decomposition on the original scanning signal through wavelet transform based on the plurality of decomposition frequency bands to obtain a noise frequency band signal and a non-noise frequency band signal. The correspondence between different fluctuation frequency bands and wavelength analysis ranges of the semiconductor process mechanism can refer to Table 1:

[0077]

[0078]

[0079] Table 1

[0080] In the semiconductor manufacturing field, seven-level decomposition can cover a wavelength analysis range from 2 nm to 256 nm, and after seven-level decomposition, the lowest frequency approximation component (A7) covers the widest frequency range. Instrument noise is caused by the imperfection or noise of the measurement equipment (such as an AFM probe), i.e., in the present application, the sub-signals corresponding to the D1-D5 frequency bands are non-noise frequency band signals, which mainly contain real feature information of the wafer surface topography, and the sub-signals corresponding to the D6-D7 frequency bands are noise frequency band signals, and the instrument noise is mainly distributed in the noise frequency band signals.

[0081] As shown in Figure 3 Fig. 1 is a schematic diagram of the position distribution of D1 to D7 different waveband components provided by an embodiment of the present application, in which the horizontal coordinate is the position in space when the original scanning signal is subjected to wavelet transform, and the vertical coordinate is the size of the detail component (Wavelet Detail Coefficients). The upper right legend is D1-D7 from top to bottom.

[0082] Finally, in semiconductor manufacturing, the energy distribution of different decomposition frequency band signals of the wafer surface topography can reflect the working state of the corresponding equipment of the process mechanism. Therefore, after multi-resolution decomposition, the energy proportion of the decomposition frequency band signals of different frequency bands needs to be analyzed to ensure the effectiveness of the original scanning signal obtained in step S100, so that in the case of invalidity, the parameters of the equipment corresponding to a certain process mechanism are calibrated.

[0083] In one or more embodiments of the present application, first, the original detail energy of each decomposition frequency band signal can be determined according to the original detail component of the decomposition frequency band signal. Meanwhile, the decomposition frequency band corresponding to the termination of the multi-resolution decomposition process is taken as the target frequency band, and the original approximation energy of the original approximation component of the target frequency band corresponding decomposition frequency band signal is determined. Then, the total energy is determined according to the original detail energy and the original approximation energy of each decomposition frequency band signal.

[0084] Secondly, the energy proportion of each decomposition frequency band signal can be determined according to the original detail energy and the total energy of the decomposition frequency band signal.

[0085] The process of determining the original detail energy, the total energy and the energy proportion is as follows:

[0086]

[0087]

[0088] In the above formula, D j (x i ) is the detail component of the decomposition frequency band signal corresponding to the jth level, N1 is the signal length of the detail component of the decomposition frequency band signal corresponding to the jth level, D J is the decomposition frequency band corresponding to the termination of the multi-resolution decomposition process, E AJ is the original approximation energy of the original approximation component of the target frequency band corresponding decomposition frequency band signal, P j is the energy proportion of the decomposition frequency band signal corresponding to the jth level.

[0089] Finally, the effectiveness of the original scanning signal is determined according to the size relationship between the energy proportion of each decomposition frequency band signal and the preset energy threshold. In one or more embodiments of the present application, when the energy proportion of each decomposition frequency band signal is less than the preset energy threshold, the original scanning signal is effective, otherwise, when the energy proportion of each decomposition frequency band signal is not less than the preset energy threshold, the original scanning signal is invalid. Then, the present application can enter the step of signal denoising of the noise frequency band signal when the original scanning signal is effective. When the original scanning signal is invalid, it indicates that the semiconductor process mechanism is abnormal, and the process warning information for prompting the inspection of each process step in semiconductor manufacturing is determined, thereby assisting the optimization of semiconductor manufacturing process and improving the manufacturing quality. Therefore, in one or more embodiments of the present application, the present application needs to generate the process warning information for prompting that the semiconductor process mechanism is abnormal when the original scanning signal is invalid.

[0090] For example, Figure 4, which is a schematic diagram of the relationship between the energy proportions of the decomposed frequency band signals D1 to D7 and the preset energy thresholds provided in an embodiment of the present application. In the figure, the dotted straight line represents the preset energy threshold.

[0091] It should be noted that Table 1 clearly shows the correspondence between different fluctuation frequency bands in the semiconductor process mechanism and each process mechanism. Therefore, the present application can target a certain process manufacturing step based on the decomposed frequency band signal with abnormal energy ratio, thereby avoiding a comprehensive inspection of the entire process flow of semiconductor manufacturing and improving inspection efficiency.

[0092] In one or more embodiments of the present application, first, the present application may treat the decomposed frequency band signal whose energy proportion is greater than a preset energy threshold as an abnormal frequency band signal, and treat the decomposed frequency band corresponding to the abnormal frequency band signal as an abnormal frequency band.

[0093] Secondly, the present application can determine the abnormal process mechanism corresponding to the abnormal frequency band based on the correspondence between different fluctuation frequency bands and each process mechanism in the preset semiconductor process mechanism.

[0094] Finally, the present application can generate process warning information based on the determined abnormal process mechanism, and the process warning information is used to prompt a targeted inspection of the process steps involved in the abnormal process mechanism, such as equipment parameter calibration.

[0095] like Figure 5 , which is a schematic diagram of the principle of determining process warning information provided by an embodiment of the present application. In the figure, first, the original detail energy 501 of each decomposed frequency band signal 500 is determined, and the decomposed frequency band at the end of multi-resolution decomposition is used as the target frequency band 502, and the original approximate energy 503 of the target frequency band is determined; secondly, based on each original detail energy and the original approximate energy, the total energy 504 is determined, and then, for each decomposed frequency band signal, the energy proportion 505 of the decomposed frequency band signal can be determined based on the original detail energy of the decomposed frequency band signal and the total energy, and the validity 507 of the original scanning signal can be determined based on the energy proportion and a preset energy threshold 506; finally, based on the validity, an abnormal frequency band signal 508 is determined, and the abnormal frequency band corresponding to the abnormal frequency band signal is determined 509, and then the abnormal process mechanism associated with the abnormal frequency band is determined 510, and based on the abnormal process mechanism, process warning information 511 is determined.

[0096] S102: performing signal denoising on the noise frequency band signal.

[0097] In one or more embodiments of the present application, in order to obtain a reconstructed scanning signal free of instrument noise in subsequent steps, in this step, the present application needs to perform signal denoising on the noise band signal obtained in step S101.

[0098] Specifically, the noise band signal obtained in step S101 is subjected to signal denoising.

[0099] It should be noted that in the present application, the specific way of signal denoising of the noise band signal is not limited, and can be set according to actual needs. For example, hard threshold processing, that is, the original detail component of the noise band signal is set to zero, and the approximate component of the noise band signal is retained. However, this method of directly setting the original detail component to zero may cause discontinuity in the subsequent reconstruction of the scanning signal. Therefore, in order to effectively reduce the noise signal and as much as possible to envelope the effective signal, in one or more embodiments of the present application, the threshold parameter for distinguishing the effective signal and the noise signal can be determined according to the original detail component of the noise band signal, so that the original detail component of the noise band signal is subjected to soft threshold processing based on the threshold parameter. In this embodiment, the soft threshold processing can improve the smoothness of the signal through the shrinkage coefficient, thereby effectively reducing the discontinuity of the subsequent signal reconstruction.

[0100] Of course, in the present application, the determination method of the threshold parameter in the soft threshold processing is not limited, and can be set according to actual needs, such as manually setting the threshold parameter based on experience. In one or more embodiments of the present application, the number of sampling points can be determined according to the sampling interval and the scanning length corresponding to the original scanning signal, and then the noise standard deviation can be determined according to each detail coefficient in the original detail component of the noise band signal, and finally the threshold parameter can be determined based on the noise standard deviation and the number of sampling points.

[0101] The process of soft threshold processing and determining the threshold parameter can refer to the following formula:

[0102] n λ = sin(X) * max(|X| - λ, 0) (13)

[0103]

[0104] In the above formula, X represents a detail coefficient in the original detail component of the noise band signal, λ is the threshold parameter, and formula (13) is a soft threshold function; σ is the noise standard deviation of the noise band signal, median(|d i |) is the median of the absolute values of each detail coefficient in the original detail component of the noise band signal, and N2 is the signal length of the noise band signal. Since in each multi-resolution decomposition in the discrete wavelet transform, the signal length of the decomposition band is down-sampled (halved), the signal length of the noise band signal can be determined based on the signal length (number of sampling points) of the original scanning signal and the decomposition level of the noise band signal.

[0105] S103: obtaining a reconstructed scanning signal through inverse wavelet transform based on the noise band signal subjected to signal denoising and the non-noise band signal.

[0106] Specifically, the present application may obtain a reconstructed scanning signal through inverse wavelet transform based on the noise band signal after signal denoising in step S102 and the non-noise band signal obtained in step S101.

[0107] It should be noted that if hard threshold processing is used for signal denoising in step S102, then in this step, the present application can obtain a reconstructed scanning signal based on the original approximate component of the noise band signal, the original approximate component of the non-noise band signal, and the original detail component of the non-noise band signal through inverse wavelet transform. If soft threshold processing is used for signal denoising in step S102, then in this step, the present application can use the original detail component after soft threshold processing as the target detail component, and then obtain a reconstructed scanning signal based on the target detail component of the noise band signal, the original approximate component of the noise band signal, the original detail component of the non-noise band signal, and the original approximate component of the non-noise band signal through inverse wavelet transform.

[0108] In the above-mentioned method for processing wafer morphology scanning signals, the original scanning signal of the target wafer surface morphology can be obtained in step S100, and the original scanning signal is subjected to multi-resolution decomposition through wavelet transform in step S101 to obtain a noise band signal and a non-noise band signal. That is, the original scanning signal is decomposed into different frequency band signals through multi-resolution decomposition, thereby realizing the separation between noise and the real characteristics of the target wafer surface morphology, and then signal denoising is performed on the noise band signal related to the noise in step S102. Finally, in step S103, based on the noise band signal and the non-noise band signal after signal denoising, an inverse wavelet transform is performed to obtain a reconstructed scanning signal from which noise is removed. The reconstructed scanning signal can accurately reflect the real characteristics of the target wafer surface morphology.

[0109] like Figure 6 As shown, it is a schematic diagram of the principle of a method for processing a wafer topography scanning signal provided in an embodiment of the present application. In this figure, the original scanning signal 600 is subjected to multi-resolution decomposition to obtain a noise band signal 601 and a non-noise band signal 602. Then, after the noise band signal is denoised, the reconstructed scanning signal 603 is obtained by signal reconstruction based on the denoised noise band signal and the non-noise band signal.

[0110] The method for processing wafer topography scanning signals provided in this application can be used in the process of determining line edge roughness to improve the accuracy of line edge roughness. In one or more embodiments of this application, this application can determine the line edge roughness of the target wafer surface topography based on the reconstructed scanning signal, where the line edge roughness can be represented by the standard deviation or three times the standard deviation of the reconstructed scanning signal, which can be set according to actual needs.

[0111] As shown in Figure 7 , a schematic diagram of simulated actual LER signal and noise signal provided by the embodiment of the present application is shown. The horizontal coordinate in the diagram is the position in space when the original scanning signal is wavelet transformed, and the vertical coordinate is the signal size. The upper right legend from top to bottom is the simulated actual LER signal and noise signal. The RMS of the original scanning signal in the upper side of the diagram is the simulated actual LER signal, and the actual LER signal is 0.234 nm; the noise standard deviation is 0.248 nm.

[0112] As shown in Figure 8 , a schematic diagram of simulated actual LER signal and LER signal after removing instrument noise provided by the embodiment of the present application is shown. The horizontal coordinate in the diagram is the position in space when the original scanning signal is wavelet transformed, and the vertical coordinate is the signal size. The upper right legend from top to bottom is the simulated actual LER signal and LER signal after removing instrument noise. The RMS of the reconstructed scanning signal in the upper side of the diagram is the LER signal after removing instrument noise, and the LER signal after removing instrument noise is 0.217 nm, that is, the LER signal corresponding to the reconstructed scanning signal is reduced by 12.3%.

[0113] The processing method based on the wafer topography scanning signal. Correspondingly, the present application also provides a specific embodiment of a wafer topography scanning signal processing device.

[0114] As shown in Figure 9 , the wafer topography scanning signal processing device 900 provided by the embodiment of the present application includes a signal acquisition module 901, a signal decomposition module 902, a signal denoising module 903, and a signal reconstruction module 904.

[0115] The signal acquisition module 901 is configured to acquire an original scanning signal of a target wafer surface topography.

[0116] The signal decomposition module 902 is configured to perform multi-resolution decomposition on the original scanning signal through wavelet transform to obtain a noise band signal and a non-noise band signal.

[0117] The signal denoising module 903 is configured to perform signal denoising on the noise band signal.

[0118] The signal reconstruction module 904 is configured to obtain a reconstructed scanning signal through inverse wavelet transform according to the noise band signal after signal denoising and the non-noise band signal.

[0119] In some embodiments, the above device further includes a line edge roughness determination module, which is specifically configured to:

[0120] determine the line edge roughness of the target wafer surface topography according to the reconstructed scanning signal.

[0121] In some embodiments, the signal decomposition module 902 described above can be specifically configured to:

[0122] determine a plurality of decomposition frequency bands of the original scanning signal according to a preset correspondence between different fluctuation frequency bands and wavelength analysis ranges in a semiconductor process mechanism;

[0123] perform multi-resolution decomposition on the original scanning signal through wavelet transform according to the plurality of decomposition frequency bands, to obtain noise frequency band signals and non-noise frequency band signals.

[0124] In some embodiments, the signal decomposition module 902 described above can be further configured to:

[0125] for each of the noise frequency band signals and the non-noise frequency band signals, determine an original detail energy of the decomposition frequency band signal according to an original detail component of the decomposition frequency band signal;

[0126] determine a target frequency band as the decomposition frequency band corresponding to a termination of the multi-resolution decomposition process, and determine an original approximation energy according to an original approximation component of the decomposition frequency band signal corresponding to the target frequency band;

[0127] determine a total energy according to the original detail energy and the original approximation energy of each of the decomposition frequency band signals;

[0128] for each of the decomposition frequency band signals, determine an energy proportion of the decomposition frequency band signal according to the original detail energy and the total energy of the decomposition frequency band signal;

[0129] determine the effectiveness of the original scanning signal according to the energy proportions of the decomposition frequency band signals and a size relationship of a preset energy threshold value;

[0130] if the original scanning signal is effective, proceed to a step of performing signal denoising on the noise frequency band signals.

[0131] In some embodiments, the signal decomposition module 902 described above can be further configured to:

[0132] generate process warning information for prompting that an abnormality exists in the semiconductor process mechanism.

[0133] In some embodiments, the signal decomposition module 902 described above can be further configured to:

[0134] determine a decomposition frequency band signal whose energy proportion is greater than the preset energy threshold value as an abnormal frequency band signal;

[0135] determine the decomposition frequency band to which the abnormal frequency band signal belongs as an abnormal frequency band;

[0136] According to a preset corresponding relationship between different fluctuation frequency bands and process mechanisms in a semiconductor process mechanism, an abnormal process mechanism corresponding to the abnormal frequency band is determined;

[0137] Based on the abnormal process mechanism, the process warning information is generated.

[0138] In some embodiments, the signal denoising module 903 is specifically configured to:

[0139] According to the original detail component of the noise frequency band signal, a threshold parameter is determined, the threshold parameter being used to distinguish effective signals and noise signals in the noise frequency band signal;

[0140] Based on the threshold parameter, the original detail component of the noise frequency band signal is subjected to soft threshold processing.

[0141] In some embodiments, the signal denoising module 903 can also be configured to:

[0142] According to a sampling interval and a scanning length corresponding to the original scanning signal, a sampling point number is determined;

[0143] According to each detail coefficient in the original detail component of the noise frequency band signal, a noise standard deviation is determined;

[0144] According to the noise standard deviation and the sampling point number, a threshold parameter is determined.

[0145] In some embodiments, the signal reconstruction module 904 is specifically configured to:

[0146] The original detail component subjected to the soft threshold processing is taken as a target detail component;

[0147] According to the target detail component of the noise frequency band signal, the original approximation component of the noise frequency band signal, the original detail component of the non-noise frequency band signal, and the original approximation component of the non-noise frequency band signal, a reconstructed scanning signal is obtained through inverse wavelet transform.

[0148] A wafer topography scanning signal processing method is provided.

[0149] Figure 10 A hardware structure schematic diagram of a wafer topography scanning signal processing device is shown.

[0150] The wafer topography scanning signal processing device can include a processor 1001 and a memory 1002 having computer program instructions stored therein.

[0151] In particular, the processor 1001 can include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or can be configured to implement one or more integrated circuits that embody the embodiments of the present application.

[0152] The memory 1002 can include mass storage for data or instructions. As an example and not by way of limitation, the memory 1002 can include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disc (e.g., a compact disc (CD) or a digital versatile disc (DVD)), a tape drive, a USB drive, or a combination of two or more of these. Where appropriate, the memory 1002 can include removable or non-removable (or fixed) media, where appropriate. Where appropriate, the memory 1002 can be internal or external to the integrated gateway disaster recovery device. In particular embodiments, the memory 1002 is non-volatile, solid-state memory.

[0153] The processor 1001 implements the processing method of the wafer topography scanning signal in any of the above embodiments by reading and executing computer program instructions stored in the memory 1002.

[0154] In one example, the electronic device can further include a communication interface 1003 and a bus 1010. As shown, the processor 1001, the memory 1002, and the communication interface 1003 are connected by the bus 1010 and complete communication between each other. Figure 10

[0155] The communication interface 1003 is mainly used to realize the communication between the modules, devices, units and / or equipment in the embodiments of the present application.

[0156] The bus 1010 includes hardware, software, or both, that couples components of the electronic device to each other. As an example and not by way of limitation, the bus can include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand (IB) interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association local (VLB) bus, or another suitable bus or interconnect, or a combination of two or more of these. Where appropriate, the bus 710 can include one or more buses. Although the present application describes and illustrates a particular bus, the present application contemplates any suitable bus or interconnect. ​

[0157] In addition, in combination with the wafer topography scanning signal processing method in the above embodiments, the embodiments of the present application can provide a computer storage medium for implementation. The computer storage medium has computer program instructions stored thereon; the computer program instructions are executed by a processor to implement any of the wafer topography scanning signal processing methods in the above embodiments.

[0158] In addition, in combination with the wafer topography scanning signal processing method in the above embodiments, the embodiments of the present application can provide a computer program product for implementation, and the instructions in the computer program product are executed by a processor of an electronic device to enable the electronic device to perform a wafer topography scanning signal processing method provided by any of the above embodiments of the present application.

[0159] It should be noted that the present application is not limited to the specific configurations and processes described above and shown in the drawings. For the sake of brevity, detailed descriptions of well-known methods are omitted. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order of the steps, after understanding the spirit of the present application.

[0160] The functional blocks shown in the structural block diagrams described above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, etc. When implemented in software, the elements of the present application are program or code segments used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted through a data signal carried in a carrier wave over a transmission medium or communication link. The "machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segments can be downloaded via a computer network such as the Internet, an intranet, etc.

[0161] It should also be noted that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in an order different from the embodiments, or several steps can be performed simultaneously.

[0162] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0163] The above describes only specific implementation of the present application. For the convenience and brevity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described herein. It should be understood that the protection scope of the present application is not limited in this way. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed in the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A method for processing wafer topography scanning signals, characterized in that: include: Obtaining the original scanning signal of the target wafer surface topography; Performing multi-resolution decomposition on the original scanning signal through wavelet transform to obtain noise frequency band signals and non-noise frequency band signals; performing signal denoising on the noise frequency band signal; A reconstructed scanning signal is obtained by performing inverse wavelet transformation based on the noise band signal and the non-noise band signal after signal denoising.

2. The method according to claim 1, wherein After obtaining a reconstructed scanning signal by inverse wavelet transform based on the noise band signal and the non-noise band signal after signal denoising, the method further includes: The line edge roughness of the target wafer surface topography is determined according to the reconstructed scanning signal.

3. The method according to claim 1, wherein The original scanning signal is subjected to multi-resolution decomposition by wavelet transform to obtain a noise band signal and a non-noise band signal, including: Determining a plurality of decomposed frequency bands of the original scanning signal according to a correspondence between different fluctuation frequency bands and wavelength analysis ranges in a preset semiconductor process mechanism; According to the multiple decomposition frequency bands, the original scanning signal is subjected to multi-resolution decomposition through wavelet transformation to obtain a noise frequency band signal and a non-noise frequency band signal.

4. The method according to claim 3, wherein After performing multi-resolution decomposition on the original scan signal by wavelet transform to obtain a noise frequency band signal and a non-noise frequency band signal, the method further includes: For each decomposed frequency band signal in the noise frequency band signal and the non-noise frequency band signal, determining original detail energy of the decomposed frequency band signal according to an original detail component of the decomposed frequency band signal; The decomposition frequency band corresponding to the termination of the multi-resolution decomposition process is used as a target frequency band, and original approximate energy is determined according to original approximate components of the decomposition frequency band signal corresponding to the target frequency band; determining a total energy according to the original detail energy and the original approximate energy of each of the decomposed frequency band signals; For each of the decomposed frequency band signals, determining an energy proportion of the decomposed frequency band signal according to the original detail energy and the total energy of the decomposed frequency band signal; Determining the validity of the original scanning signal based on the energy proportion of each of the decomposed frequency band signals and a relationship between a preset energy threshold; When the original scanning signal is valid, the process proceeds to the step of performing signal denoising on the noise frequency band signal.

5. The method according to claim 4, characterized in that The method further comprises: In the case that the original scanning signal is invalid, process warning information is generated to prompt that an abnormality exists in the semiconductor process mechanism.

6. The method according to claim 5, wherein Generating process warning information for indicating that an abnormality exists in the semiconductor process mechanism, including: The decomposed frequency band signal whose energy proportion is greater than the preset energy threshold is regarded as an abnormal frequency band signal; taking the decomposed frequency band to which the abnormal frequency band signal belongs as the abnormal frequency band; Determining the abnormal process mechanism corresponding to the abnormal frequency band according to the corresponding relationship between different fluctuation frequency bands and each process mechanism in the preset semiconductor process mechanism; Based on the abnormal process mechanism, the process warning information is generated.

7. The method according to claim 1, wherein Performing signal denoising on the noise frequency band signal specifically includes: determining a threshold parameter according to an original detail component of the noise frequency band signal, wherein the threshold parameter is used to distinguish between a valid signal and a noise signal in the noise frequency band signal; Based on the threshold parameter, a soft threshold process is performed on the original detail component of the noise band signal.

8. The method according to claim 7, wherein Determining a threshold parameter according to the original detail component of the noise band signal specifically includes: Determining the number of sampling points according to the sampling interval and scanning length corresponding to the original scanning signal; determining a noise standard deviation according to detail coefficients in the original detail component of the noise frequency band signal; A threshold parameter is determined according to the noise standard deviation and the number of sampling points.

9. The method according to claim 7, wherein Obtaining a reconstructed scan signal by inverse wavelet transform according to the noise band signal and the non-noise band signal after signal denoising, specifically comprising: Using the original detail component after the soft threshold processing as the target detail component; A reconstructed scanning signal is obtained by inverse wavelet transform according to the target detail component of the noise band signal, the original approximate component of the noise band signal, the original detail component of the non-noise band signal and the original approximate component of the non-noise band signal.

10. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the method for processing wafer topography scanning signals according to any one of claims 1 to 7.

Images