Method and system for measuring thickness of multilayer coated lens

Abstract

The invention relates to the technical field of intelligent sensing systems, and provides a method and system for measuring the thickness of a multilayer coated lens, and the method comprises the steps: collecting reflection signals, dark field signals and blank reflection signals of a plurality of incident angles of the multilayer coated lens through an ellipsometer; decomposing the reflected signal of any incident angle to obtain a plurality of frequency band signals, and constructing a feature vector of each frequency band signal to obtain a plurality of frequency band signals of intermediate frequencies; obtaining a plurality of mixed layers of the multi-layer coated lens and the effective refractive index of the mixed layers; obtaining an optimal analog signal of the reflected signal of each incident angle; acquiring a real interference reflection signal of each incident angle, and obtaining the thickness of each layer of coating film under each incident angle through ellipsometric inversion; and the final thickness of each coating film is obtained by integrating the thickness of each coating film under multiple incidence angles. The invention aims to solve the problem that the thickness measurement accuracy of the ellipsometer is influenced by interface fuzziness caused by a mixed layer among multiple layers of coating films.

Description

Technical Field

[0001] This invention relates to the field of intelligent sensing system technology, specifically to a method and system for measuring the thickness of multi-layer coated lenses. Background Technology

[0002] The ellipsometer method utilizes polarized light to illuminate a film layer and then uses the amplitude ratio of the reflected light to determine the polarization. ) and phase difference ( The change in amplitude of the reflected light is then combined with the optical model to invert the thickness, thus determining the amplitude ratio of the reflected light. ) and phase difference ( The accuracy of thickness measurement is determined by the optical path difference; the thickness calculation of the ellipsometer depends on the accurate derivation of the optical path difference, and structural noise will directly lead to the deviation of the optical path difference calculation. Due to the influence of the deposition process, during film deposition (such as sputtering, evaporation, chemical vapor deposition CVD), the atoms / molecules of the later deposited film have a certain kinetic energy (especially the kinetic energy of ion bombardment is higher during sputtering), which will embed into the surface of the bottom film layer. The atoms of the bottom film layer will also diffuse to the upper film layer due to thermal motion, forming a "mixed region of mutual penetration"; at the same time, subsequent processing such as annealing / curing will significantly accelerate the diffusion rate of interface atoms, making the originally narrow transition layer wider and forming a distinct mixed layer, all of which will result in interface blurring (i.e., "transition mixed layer").

[0003] A blurry interface can lead to The signal exhibits "pseudo-features" such as peak broadening and sawtooth fluctuations. These structural noises can mask the true position and phase information of the interference peaks. If they are not removed, accurate polarization parameters cannot be extracted, resulting in a peak and valley position judgment error of 0.5-1 nm. Ultimately, this translates into a nanometer-level thickness inversion error, which cannot meet the requirements of high-precision measurement. Summary of the Invention

[0004] This invention provides a method and system for measuring the thickness of multi-layer coated lenses, to solve the problem that the interface blurring caused by the mixing layers between multi-layer coatings affects the accuracy of thickness measurement by ellipsometers. The specific technical solution adopted is as follows: This invention proposes a method for measuring the thickness of multi-layer coated lenses, the method comprising the following steps: The reflection signals, dark field signals, and blank reflection signals at several incident angles were collected from the multilayer coated lens using an ellipsometer. The reflected signal at any incident angle is decomposed to obtain several frequency band signals. Based on the amplitude variation of the frequency band signals with wavelength, feature vectors of each frequency band signal are constructed. Based on the feature vectors, several intermediate frequency band signals are obtained. The deviation between the peak wavelength of the intermediate frequency band signal and the standard characteristic wavelength of each coating layer is analyzed. Combined with the standard refractive index of each coating layer, several mixed layers of the multilayer coated lens and their equivalent refractive indices are obtained. Based on the standard refractive index of each coating layer and the equivalent refractive index of several mixed layers under the same incident angle of the reflected signal, combined with the incident angle of each coating layer, the transmission matrix of each coating layer and each mixed layer under each reflected signal is obtained, and then the overall transmission matrix is ​​generated; the thickness of each mixed layer is traversed, and several reflection coefficients are obtained according to the overall transmission matrix and a simulated signal is generated; the similarity relationship between the simulated signal and the frequency band signal of the middle frequency is analyzed, and the optimal simulated signal of the reflected signal at each incident angle is obtained. Based on the optimal simulated signal of the reflection signal at each incident angle, the true interference reflection signal at each incident angle is obtained, and the thickness of each coating layer at each incident angle is obtained by ellipsometry; by combining the thicknesses of each coating layer at multiple incident angles, the final thickness of each coating layer is obtained.

[0005] Optionally, the specific method for decomposing the reflected signal at any incident angle to obtain several frequency band signals includes: For any incident angle, the reflected signal is decomposed by wavelet packet to obtain several signal components, resulting in a total of 8 signal components, which serve as the 8 frequency band signals of the reflected signal.

[0006] Optionally, the feature vectors of each frequency band signal are obtained using the following method: For any frequency band signal, obtain several maxima points of that frequency band signal. For any one of these maxima points, start from that maxima point and traverse to both sides. The first point with a slope of 0 encountered on each side is taken as the two boundary points of that maxima point. The maxima point and its two boundary points constitute a characteristic peak. Obtain the characteristic peaks corresponding to each maxima point using the above method. It is worth noting that if a maxima point is close to the start or end point of the frequency band signal, resulting in the absence of a boundary point on one side, then the start or end point is used as the boundary point to form the characteristic peak. For any characteristic peak... The horizontal coordinate distance between the two boundary points of the peak is taken as the width of the characteristic peak. The number of characteristic peaks in the signal band, the mean width of all characteristic peaks, the maximum amplitude, the minimum amplitude, the mean slope and the variance of the slope of all points in the signal band are obtained. These six dimensions are used as the basic elements of the feature vector. The data of the same dimension of all signals in the same frequency band of the reflected signal are linearly normalized, and the results are used as the processing elements of the corresponding dimensions. The processing elements of all dimensions of the same signal band constitute the feature vector of the signal band.

[0007] Optionally, the classification based on the feature vector includes the following specific methods: For any two frequency band signals of the reflected signal at any incident angle, obtain the cosine similarity of the feature vectors of the two frequency band signals. Subtract the cosine similarity from 1 to obtain the difference as the feature distance between the two frequency band signals. Based on the feature distance, cluster all frequency band signals of the reflected signal. Use K-means clustering, with the feature distance as the distance metric and K value of 3, to obtain three clusters.

[0008] Optionally, the specific method for obtaining the frequency band signal of several intermediate frequencies includes: The mean frequency of all frequency band components in any cluster is obtained as the average frequency of that cluster. Clusters with the median average frequency are designated as intermediate clusters. All frequency band signals in the intermediate clusters are designated as frequency band signals with several intermediate frequencies. The reflected signals at each incident angle are analyzed by wavelet decomposition, eigenvector construction, and clustering to obtain the frequency band signals with several intermediate frequencies of the reflected signals at each incident angle.

[0009] Optionally, the specific method for defining the plurality of mixed layers and their equivalent refractive index of the multilayer coated lens is as follows: Obtain the standard characteristic wavelengths of each coating layer of a multilayer coated lens, as well as its standard refractive index, and form an interval between adjacent standard characteristic wavelengths to obtain several standard characteristic wavelength intervals. For any intermediate frequency band signal in the reflected signal at any incident angle, obtain the wavelengths corresponding to several maximum points in the intermediate frequency band signal, and use them as the wavelengths corresponding to several peaks. For any peak wavelength, the standard characteristic wavelength range in which the peak wavelength is located is obtained. The coating layer with the minimum standard characteristic wavelength in the two coating layers corresponding to the standard characteristic wavelength range is taken as the upper coating layer for the peak wavelength, and the coating layer with the maximum standard characteristic wavelength is taken as the lower coating layer for the peak wavelength. The difference between the peak wavelength and the standard characteristic wavelength of the upper coating layer is obtained. The ratio of the difference to the length of the standard characteristic wavelength range is taken as the lower offset factor for the peak wavelength. The length of the standard characteristic wavelength range is the difference between the standard characteristic wavelength of the lower coating layer and the standard characteristic wavelength of the upper coating layer. The ratio of the difference between the standard characteristic wavelength of the lower layer coating and the wavelength corresponding to the peak value to the length of the standard characteristic wavelength range is used as the upper layer offset factor for the wavelength corresponding to the peak value. The upper layer offset factor is used as the offset weight of the upper layer coating, and the lower layer offset factor is used as the offset weight of the lower layer coating. The standard refractive indices of the upper and lower layers coatings are weighted and summed using the offset weights of the upper and lower layers coatings. The result is used as the calculated refractive index for the wavelength corresponding to the peak value, and the wavelength corresponding to the peak value is used as a hybrid layer. Several peak wavelengths corresponding to the frequency band signals of each intermediate frequency of the reflected signal are obtained, and their refractive indices are calculated. Several peak wavelengths corresponding to the same upper and lower coating layers are taken as the same mixed layer. The average value of the calculated refractive indices of all peak wavelengths under the same mixed layer is taken as the equivalent refractive index of the mixed layer. Thus, several mixed layers of the reflected signal and their equivalent refractive indices are obtained.

[0010] Optionally, the specific method for obtaining the transmission matrix of each coating layer and each hybrid layer under each reflected signal includes: For a reflected signal at any incident angle, after obtaining the refractive index of each coating film and the equivalent refractive index of the mixed layer, the incident angle within each coating film is calculated wavelength by wavelength according to the law of refraction; simultaneously, the complex refractive index is calculated using the formula: Calculate complex refractive index ,in The extinction coefficient is used, therefore the refractive index of each coating is taken as the complex refractive index; the propagation phase value of each layer is calculated according to the propagation phase calculation formula and the polarization impedance formula. And polarization impedance values, and then construct the transmission matrix for different coating layers based on the transmission matrix.

[0011] Optionally, the specific method for obtaining the optimal analog signal of the reflected signal at each incident angle includes: Generating an analog signal based on the reflection coefficient: ,in Represents the p-polarized reflection coefficient. Let represent the s-polarization reflection coefficient; obtain the mean cosine similarity between the measured and simulated intermediate frequency band signals of the reflected signal, where the simulated intermediate frequency band signal corresponds to the thickness of a mixing layer, denoted as . Then, the thickness corresponds to the mean of a cosine similarity, which serves as the preferred value of the thickness. The thickness of each mixing layer is preset, and several thicknesses are obtained by iterating through them. Other thicknesses are iterated through to obtain the corresponding degree of optimization. The analog signal corresponding to the maximum value of the degree of optimization is taken as the best analog signal of the reflection signal.

[0012] Optionally, the method for obtaining the true interference reflection signal at each incident angle based on the optimal simulated signal of the reflection signal at each incident angle, and obtaining the thickness of each coating layer at each incident angle through ellipsometric inversion, includes the following specific methods: For any incident angle, the reflected signal is subtracted from its best analog signal, and then the blank reflected signal at that incident angle is subtracted. The remaining signal is taken as the true interference reflection signal at the incident angle corresponding to the reflected signal. The classic Levenberg-Marquardt algorithm for elliptic inversion was used to invert the thickness of each coating layer, and the thickness value of each coating layer was obtained.

[0013] The present invention also proposes a thickness measurement system for multilayer coated lenses, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above method.

[0014] The beneficial effects of this invention are as follows: This invention decomposes the received data and dynamically reverses the equivalent refractive index of the blurred interface hybrid layer based on the changes after decomposition. Furthermore, it simulates interference signals based on the equivalent refractive index, thereby removing the interference signals. Specifically, an ellipsometer is used to collect reflection signals from multiple incident angles of the multilayer coating, and environmental interference is eliminated using dark field signals and blank reflection signals. The reflection signals are decomposed to obtain frequency band signals. A feature vector reflecting the change pattern of the frequency band signals is constructed based on the peak wavelength changes in the frequency band signals, and the intermediate frequency band signals are extracted to initially reflect the interference signals generated by the hybrid layer. Based on the waveform of the intermediate frequency band signals and its relationship with the absorption peak wavelength distribution of the coating material, the refractive index of the hybrid layer is analyzed and the equivalent refractive index is obtained. Then, a transmission matrix of the coating film and the hybrid layer is constructed, and an overall transmission matrix is ​​built to generate the optimal simulated signal of the intermediate frequency band signals, which is used to remove interference from the reflection signals to obtain the true interference reflection signals, thereby achieving the final inversion of the thickness of each coating layer. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic flowchart of a method for measuring the thickness of a multilayer coated lens according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 The diagram illustrates a flowchart of a method for measuring the thickness of a multilayer coated lens according to an embodiment of the present invention. The method includes the following steps: Step S001: Collect reflection signals, dark field signals and blank reflection signals at several incident angles of the multilayer coated lens using an ellipsometer.

[0019] The purpose of this embodiment is to eliminate the interference of the mixed layers between the multilayer coatings on the interference characteristics during the thickness measurement of multilayer coated lenses using an ellipsometer. This requires first acquiring multiple reflection signals from multiple incident angles, and simultaneously recording the dark field signal and the blank reflection signal of the corresponding incident angle without a multilayer coated lens.

[0020] Specifically, an industry-standard ellipsometer is used for data acquisition. The ellipsometer mainly consists of a light source, polarizer, compensator, analyzer, and detector. The surface of the multi-layer coated lens is cleaned, and the test sample is fixed on the stage. The stage is adjusted so that the incident light perpendicularly illuminates the defect-free measurement point of the sample. The instrument's real-time monitoring interface ensures that the reflected light accurately enters the detector. The measurement program is started, and the instrument automatically acquires the polarization state changes of the reflected light point by point according to the preset incident angle and wavelength range, outputting the wavelength-incident angle corresponding to each set of measurements. (Unit: °) and (Unit: °) Parameters; Similarly, different incident angles (30°, 45°, 50°, 60° and 70° in this embodiment) are selected for reflection signal acquisition; the light source is turned off, and each dark field signal is acquired to eliminate detector noise. The result obtained by subtracting the received reflected light from the dark field signal is used as the reflection signal for each incident angle; the sample is removed, and the reflection signal of the blank substrate (of the same material as the sample to be tested) is acquired as the reference for subsequent background subtraction, thus obtaining the blank reflection signal for each incident angle; in this embodiment, the measurement band is set to 400-1200nm with a wavelength interval of 1nm to ensure that all interference peak characteristics of the multilayer film are captured.

[0021] Step S002: Decompose the reflected signal at any incident angle to obtain several frequency band signals, and construct the feature vector of each frequency band signal based on the amplitude change with wavelength in the frequency band signal. Classify the signal according to the feature vector to obtain several intermediate frequency band signals. Analyze the deviation between the peak wavelength corresponding to the intermediate frequency band signal and the standard characteristic wavelength of each coating layer. Combined with the standard refractive index of each coating layer, obtain several mixed layers of the multilayer coated lens and their equivalent refractive indices.

[0022] It should be noted that the root cause of interface blurring is the "physical mixing layer." Multilayer film interface blurring is not random noise, but rather a "transitional mixing layer" (5-20 nm thick) formed by atomic diffusion and chemical reactions. Its optical properties are a weighted superposition of adjacent layers, exhibiting clear physical laws. Referring to the idea of ​​"effective medium theory" (such as the Bruggeman model), all blurred interfaces are integrated into a "single equivalent mixing layer." This eliminates the need to separate layers or know the individual parameters of each interface; the equivalent parameters are inferred solely from the overall signal of the multilayer film. The original data is composed of a "superposition relationship": the measured original signal = the real signal of the multilayer film + the interference signal of the mixing layer (linear superposition, without complex coupling). The interference signal simulated based on the equivalent model can be precisely removed by directly separating the "original data - interference signal," eliminating peak broadening, tailing, and other interference caused by the mixing layer. This only affects the interference component and does not affect the position, amplitude, or phase characteristics of the interference peaks of the real signal. Therefore, by decomposing the received data and dynamically inferring the equivalent refractive index of the blurred mixing layer based on the changes after decomposition, the interference signal can be further simulated based on the equivalent refractive index, thus removing the interference signal.

[0023] Preferably, in one embodiment of the present invention, the reflected signal at any incident angle is decomposed to obtain several frequency band signals, and based on the amplitude variation with wavelength in the frequency band signals, feature vectors of each frequency band signal are constructed. Classification is then performed based on the feature vectors to obtain several intermediate frequency band signals. The specific method includes: For a reflected signal at any incident angle, wavelet packet decomposition is used to obtain several signal components. In this embodiment, the wavelet basis is selected as db4, and the decomposition level is set to 3, resulting in a total of 8 signal components, which serve as the 8 frequency band signals of the reflected signal. For any frequency band signal, several maxima points are obtained. For any maxima point, the signal is traversed to both sides from the maxima point. The first point on each side where the slope value changes sign is taken as the two boundary points of the maxima point. The maxima point and its two boundary points form a characteristic peak. The characteristic peaks corresponding to each maxima point are obtained according to the above method. It should be noted that if the maxima point is close to the beginning of the frequency band signal... If a starting or ending point results in the absence of a boundary point on one side, then the starting or ending point is used as the boundary point to form a characteristic peak. The horizontal coordinate distance between the two boundary points of any characteristic peak is taken as the width of the characteristic peak. The number of characteristic peaks in the frequency band signal, the mean width of all characteristic peaks, the maximum amplitude, the minimum amplitude, the mean slope of all points in the frequency band signal, and the variance of the slope are obtained. These six dimensions are used as the basic elements of the feature vector. The data of the same dimension of all frequency band signals of the reflected signal are linearly normalized, and the results are used as the processing elements of the corresponding dimensions. The processing elements of all dimensions of the same frequency band signal constitute the feature vector of the frequency band signal.

[0024] Furthermore, for any two frequency band signals of the reflected signal, the cosine similarity of the feature vectors of the two frequency band signals is obtained. The difference obtained by subtracting the cosine similarity from 1 is used as the feature distance between the two frequency band signals. Based on the feature distance, all frequency band signals of the reflected signal are clustered. In this embodiment, K-means clustering is used, and the distance metric is the feature distance. In this embodiment, the K value is described as 3, resulting in three clusters. The mean of the frequencies corresponding to all frequency band components in any cluster is obtained as the average frequency of the cluster. All clusters with the average frequency in the middle value (i.e., neither the maximum nor the minimum value) are considered as intermediate clusters. All frequency band signals in the intermediate clusters are considered as several intermediate frequency frequency band signals. Following the above method, the reflected signals at each incident angle are processed through wavelet decomposition, feature vector construction, and cluster analysis to obtain several intermediate frequency frequency band signals of the reflected signals at each incident angle.

[0025] It should be noted that since the high-frequency components originate from refractive index gradients (sawtooth-like small fluctuations), random electronic noise, environmental vibration interference, etc., and are characterized by irregular fluctuations and small amplitudes (≤0.2°), they are completely unrelated to the "gradual broadening" characteristic of interface ambiguity. On the other hand, the low-frequency components correspond to "ideal multilayer film interference signals without interface ambiguity," characterized by sharp peaks and peak widths ≤0.3nm, reflecting the true optical properties of the multilayer film itself, rather than interference from interface ambiguity. Therefore, the frequency band signal with the average frequency as the middle frequency among the three frequency band signals is selected as the object of subsequent analysis.

[0026] It should be further explained that the closer the wavelength corresponding to the peak in the mid-frequency band signal is to the standard characteristic wavelength corresponding to each coating layer, i.e., the wavelength of the absorption peak corresponding to the coating material, the greater the diffusion ratio and optical contribution of the corresponding film layer in the mixed layer reflected by the peak of the band signal, and the greater the corresponding offset weight. The determination of the mixed layer first requires determining the upper and lower coating layers of the mixed layer based on the distribution relationship between the peak wavelength and the standard characteristic wavelength, and then obtaining the equivalent refractive index based on the standard refractive index of the upper and lower coating layers and the offset weight.

[0027] Preferably, in one embodiment of the present invention, the deviation between the peak wavelength corresponding to the mid-frequency band signal and the standard characteristic wavelength of each coating layer is analyzed, and combined with the standard refractive index of each coating layer, several mixed layers of the multilayer coated lens and their equivalent refractive indices are obtained. The specific method includes: First, obtain the standard characteristic wavelengths (the standard characteristic wavelengths are the absorption peak wavelengths of the corresponding coating materials) and their standard refractive indices for each layer of the multilayer coated lens. Establish a range between adjacent standard characteristic wavelengths to obtain several standard characteristic wavelength ranges. For any intermediate frequency band signal in the reflected signal at any incident angle, obtain the wavelengths corresponding to several maxima in the intermediate frequency band signal, as several peak wavelengths. For any peak wavelength, obtain the standard characteristic wavelength range in which the peak wavelength is located. The coating layer corresponding to the minimum standard characteristic wavelength among the two coating layers in the standard characteristic wavelength range is taken as the upper coating layer for the peak wavelength, and the coating layer corresponding to the maximum standard characteristic wavelength is taken as the lower coating layer for the peak wavelength. Obtain the peak wavelength minus the value of the upper coating layer. The difference obtained from the standard characteristic wavelength is used as the ratio of the difference to the length of the standard characteristic wavelength interval, which is then used as the lower layer offset factor for the wavelength corresponding to the peak. The length of the standard characteristic wavelength interval is the difference between the standard characteristic wavelength of the lower layer coating and the standard characteristic wavelength of the upper layer coating. Simultaneously, the ratio of the difference between the standard characteristic wavelength of the lower layer coating and the wavelength corresponding to the peak is used as the upper layer offset factor for the wavelength corresponding to the peak. The upper layer offset factor is used as the offset weight of the upper layer coating, and the lower layer offset factor is used as the offset weight of the lower layer coating. The standard refractive indices of the upper and lower layers coatings are weighted and summed using the offset weights of the upper and lower layers coatings. The result is used as the calculated refractive index for the wavelength corresponding to the peak, and the wavelength corresponding to the peak is treated as a hybrid layer.

[0028] Furthermore, following the method described above, several peak wavelengths corresponding to the frequency band signals of each intermediate frequency of the reflected signal are obtained, and their refractive indices are calculated. Several peak wavelengths corresponding to the same upper and lower coating layers are taken as the same mixed layer, and the average of the calculated refractive indices of all peak wavelengths under the same mixed layer is taken as the equivalent refractive index of the mixed layer. Thus, several mixed layers of the reflected signal and their equivalent refractive indices are obtained. Several mixed layers of the reflected signal at each incident angle and their equivalent refractive indices are obtained following the method described above.

[0029] Thus, several mixed layers and their equivalent refractive indices for the reflected signals at each incident angle are obtained.

[0030] Step S003: Based on the standard refractive index of each coating layer and the equivalent refractive index of several mixed layers under the same incident angle of the reflected signal, and combined with the incident angle of each coating layer, the transmission matrix of each coating layer and each mixed layer under each reflected signal is obtained, and then the overall transmission matrix is ​​generated; the thickness of each mixed layer is traversed, and several reflection coefficients are obtained according to the overall transmission matrix and a simulated signal is generated; the similarity relationship between the simulated signal and the frequency band signal of the middle frequency is analyzed, and the optimal simulated signal of the reflected signal at each incident angle is obtained.

[0031] Preferably, in one embodiment of the present invention, based on the standard refractive index of each coating layer and the equivalent refractive index of several mixed layers under the same incident angle of the reflected signal, combined with the incident angle of each coating layer, the transmission matrix of each coating layer and each mixed layer under each reflected signal is obtained, and then the overall transmission matrix is ​​generated. The specific method includes: For a reflected signal at any incident angle, after obtaining the refractive index of each coating film and the equivalent refractive index of the mixed layer, according to the law of refraction ( ,in This indicates that the refractive index of air is 1. Indicates the air-side incident angle. Indicates the first The refractive index of the coating layer, Indicates the first The incident angle within each coating layer is calculated wavelength by wavelength; simultaneously, the complex refractive index is calculated using the following formula: Calculate complex refractive index ,in The extinction coefficient is represented by 0 (for transparent films), therefore the refractive index of each coating layer is taken as the complex refractive index (including the equivalent refractive index of the mixed layer); according to the propagation phase calculation formula and the polarization impedance (s-polarization / p-polarization) formula, the propagation phase value of each layer is calculated respectively. And polarization impedance values, and then construct the transmission matrix for different coating layers based on the transmission matrix.

[0032] Furthermore, based on the transmission matrices of all coating layers, the overall transmission matrix is ​​obtained by multiplying them sequentially from the bottom layer to the surface layer using matrix multiplication. It should be noted that the overall transmission matrices for s-polarization and p-polarization need to be calculated separately. The acquisition of complex refractive index and transmission matrix are well-known techniques in their respective fields and will not be elaborated upon in this embodiment. The overall transmission matrix of the reflected signal at each incident angle is obtained according to the above method.

[0033] Preferably, in one embodiment of the present invention, the thickness of each hybrid layer is traversed, and several reflection coefficients are obtained based on the overall transmission matrix to generate a simulated signal. The similarity relationship between the simulated signal and the frequency band signal at the intermediate frequency is analyzed to obtain the optimal simulated signal of the reflection signal at each incident angle. The specific method includes: For a reflected signal at any incident angle, the p-polarized reflection coefficient and s-polarized reflection coefficient are obtained based on the overall transmission matrix, which are existing technologies and will not be described in detail in this embodiment; an analog signal is generated based on the reflection coefficients: ,in Represents the p-polarized reflection coefficient. Let represent the s-polarization reflection coefficient; for the intermediate frequency band signal of the obtained simulated signal, obtain the mean cosine similarity between the measured intermediate frequency band signal and the simulated intermediate frequency band signal of the reflected signal. The simulated intermediate frequency band signal corresponds to the thickness of a mixing layer, denoted as . The thickness corresponds to the mean of a cosine similarity, which is used as the preferred degree of the thickness. In this embodiment, the thickness of each hybrid layer is preset, with an initial thickness of 5nm, a maximum thickness of 20nm (covering the range of common industrial hybrid layers), and a step size of 0.5nm, to obtain several thicknesses. Other thicknesses are obtained by traversing them according to the above method, and the corresponding preferred degree is obtained. The analog signal corresponding to the maximum preferred degree is used as the best analog signal of the reflection signal.

[0034] Thus, the optimal analog signal of the reflected signal at each incident angle is obtained.

[0035] Step S004: Based on the best simulated signal of the reflection signal at each incident angle, obtain the real interference reflection signal at each incident angle, and obtain the thickness of each coating layer at each incident angle through ellipsometry; combine the thicknesses of each coating layer at multiple incident angles to obtain the final thickness of each coating layer.

[0036] Specifically, for any incident angle, the reflected signal, since the optimal simulation signal is obtained by fusing and simulating the mid-frequency band signal, is actually the interference signal generated by the mixing layer. The remaining signal is obtained by subtracting the optimal simulation signal from the reflected signal, and then subtracting the blank reflected signal at that incident angle. The remaining signal portion is taken as the true interference reflection signal corresponding to that incident angle, i.e., the true... and .

[0037] Furthermore, the classic Levenberg-Marquardt (LM) algorithm for ellipsometry inversion is used to invert the thickness of each coating layer, obtaining the thickness value of each coating layer. Simultaneously, since different incident angles are measured, each incident angle corresponds to the thickness of each coating layer. For any coating layer, the range of its thickness values ​​at different incident angles is obtained, and a preset judgment threshold is set (in this embodiment, the judgment threshold is described as 0.1 nm). If the range is less than or equal to the judgment threshold, the average thickness value of the coating layer at different incident angles is taken as the thickness of that coating layer. If the range is greater than the judgment threshold, the average of the absolute values ​​of the differences between the thickness value of the coating layer at any incident angle and the thickness values ​​at other incident angles is obtained as the thickness deviation of the coating layer at that incident angle. The thickness value with the smallest thickness deviation at all incident angles is taken as the thickness of that coating layer. The thickness of each coating layer is obtained using the above method, realizing the thickness measurement of multi-layer coated lenses.

[0038] This concludes the embodiment.

[0039] Another embodiment of the present invention provides a thickness measurement system for multilayer coated lenses. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the above-described method steps S001 to S004.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of measuring the thickness of a multi-layer coated lens, the method comprising: The method comprises the following steps: Collecting reflection signals, dark field signals and blank reflection signals of the multilayer coated lens at a plurality of incident angles by an ellipsometer; Decomposing the reflection signal at an arbitrary incident angle to obtain a plurality of band signals, and constructing a feature vector of each band signal based on the change of the amplitude with the wavelength in the band signal, and classifying according to the feature vector to obtain band signals at intermediate frequencies; analyzing the deviation between the peak wavelength in the band signal at the intermediate frequency and the standard characteristic wavelength of each layer of coating, and combining the standard refractive index of each layer of coating to obtain a plurality of mixed layers and the equivalent refractive index of the multilayer coated lens; Based on the standard refractive index of each layer of coating and the equivalent refractive index of the plurality of mixed layers under the same incident angle, and combining the incident angles of each layer of coating, the transmission matrix of each layer of coating and each mixed layer under each reflection signal is obtained, and then the overall transmission matrix is generated; traversing the thickness of each mixed layer, and obtaining a plurality of reflection coefficients and generating simulated signals according to the overall transmission matrix, and analyzing the similarity between the simulated signals and the band signals at the intermediate frequencies to obtain the best simulated signal of the reflection signal at each incident angle; Based on the best simulated signal of the reflection signal at each incident angle, the real interference reflection signal at each incident angle is obtained, and the thickness of each layer of coating at each incident angle is obtained by ellipsometric inversion; and the thickness of each layer of coating is obtained by combining the thickness of each layer of coating under multiple incident angles.

2. The method of claim 1, wherein, The specific method for obtaining a plurality of band signals by decomposing the reflection signal at an arbitrary incident angle comprises: For the reflection signal at any incident angle, a plurality of signal components are obtained by wavelet packet decomposition, a total of 8 signal components, as 8 band signals of the reflection signal.

3. The method of claim 1, wherein the thickness of the multi-layer coated lens is measured by a method comprising: The specific method for obtaining the feature vector of each band signal comprises: For any band signal, a plurality of maximum points of the band signal are obtained, and for any maximum point, the first point with a slope of 0 traversed to both sides of the maximum point is taken as the two boundary points of the maximum point, and the maximum point and the two boundary points form a feature peak; the feature peaks corresponding to the maximum points are obtained according to the above method, and it is particularly pointed out that if the maximum point is close to the starting point or the ending point of the band signal, the starting point or the ending point is taken as the boundary point to form the feature peak; the horizontal distance between the two boundary points of any feature peak is taken as the width of the feature peak; the number of feature peaks in the band signal, the mean value of the widths of all feature peaks, the maximum value and the minimum value of the amplitude in the band signal, the mean value and the variance of the slope of all points in the band signal are obtained, and the six dimensions are taken as the basic elements of the feature vector; the same dimension data of all band signals of the reflection signal is linearly normalized to obtain the processing elements of the corresponding dimensions, and all dimensions of the same band signal are taken as the feature vector of the band signal.

4. The method of claim 3, wherein the thickness of the multi-layer coated lens is measured by, The specific method for classifying according to the feature vector comprises: For any two band signals of the reflection signal of any incident angle, cosine similarity of the eigenvectors of the two band signals is obtained, and a difference value obtained by subtracting the cosine similarity from 1 is taken as a feature distance of the two band signals, all band signals of the reflection signal are clustered based on the feature distance, K-means clustering is used, distance measurement uses the feature distance, K value is 3, and three clusters are obtained.

5. The method of claim 4, wherein the thickness of the multi-layer coated lens is measured by, The specific method for obtaining the band signals of the intermediate frequencies includes the following steps: An average value of the frequencies corresponding to all band components in any cluster is taken as an average frequency of the cluster, a cluster with the average frequency as an intermediate value in all clusters is taken as an intermediate cluster, and all band signals in the intermediate cluster are taken as the band signals of the intermediate frequencies.

6. The method of claim 1, wherein, The specific method for obtaining the mixed layers and the equivalent refractive indexes of the multi-layer coated lenses includes the following steps: Standard characteristic wavelengths of each layer of the multi-layer coated lens and standard refractive indexes thereof are obtained, an interval is formed between adjacent standard characteristic wavelengths, and a plurality of standard characteristic wavelength intervals are obtained. For any band signal of an intermediate frequency of the reflection signal of any incident angle, wavelengths corresponding to a plurality of maximum points in the band signal of the intermediate frequency are taken as a plurality of peak corresponding wavelengths. For any peak corresponding wavelength, a standard characteristic wavelength interval in which the peak corresponding wavelength is located is obtained, a layer coated with a standard characteristic wavelength corresponding to a minimum value of the standard characteristic wavelength interval is taken as an upper layer coated with the peak corresponding wavelength, a layer coated with a standard characteristic wavelength corresponding to a maximum value of the standard characteristic wavelength interval is taken as a lower layer coated with the peak corresponding wavelength, a difference value obtained by subtracting the standard characteristic wavelength of the upper layer from the peak corresponding wavelength is obtained, a ratio of the difference value to a length of the standard characteristic wavelength interval is taken as a lower offset factor of the peak corresponding wavelength, and the length of the standard characteristic wavelength interval is a difference value obtained by subtracting the standard characteristic wavelength of the upper layer from the standard characteristic wavelength of the lower layer. A ratio of a difference value obtained by subtracting the peak corresponding wavelength from the standard characteristic wavelength of the lower layer to the length of the standard characteristic wavelength interval is taken as an upper offset factor of the peak corresponding wavelength, the upper offset factor is taken as an offset weight of the upper layer, the lower offset factor is taken as an offset weight of the lower layer, the standard refractive indexes of the upper and lower layers are weighted and summed based on the offset weights of the upper and lower layers, a result obtained is taken as a calculated refractive index of the peak corresponding wavelength, and the peak corresponding wavelength is taken as a mixed layer. A plurality of peak corresponding wavelengths and calculated refractive indexes thereof are obtained for the band signals of the intermediate frequencies of the reflection signal, a plurality of peak corresponding wavelengths corresponding to the upper and lower layers are taken as the same mixed layer, an average value of the calculated refractive indexes of all peak corresponding wavelengths of the same mixed layer is taken as an equivalent refractive index of the mixed layer, and the mixed layers and the equivalent refractive indexes of the reflection signal are obtained.

7. The method of claim 1, wherein the thickness of the multi-layer coated lens is measured by a method comprising: The method comprises the following steps of: For the reflection signal of any incident angle, after the refractive index of each coating film and the equivalent refractive index of the mixed layer are obtained, the incident angle in each coating is calculated according to the refraction law for each wavelength; at the same time, the complex refractive index is calculated according to the complex refractive index calculation formula: Calculate the complex refractive index , wherein The extinction coefficient is represented, so the refractive index of each coating film is taken as the complex refractive index; According to the propagation phase calculation formula and the polarization impedance formula, the propagation phase values of each layer are calculated and the polarization impedance values, and then the transmission matrices of different coating layers are constructed according to the transmission matrix.

8. The method of claim 1, wherein the thickness of the multi-layer coated lens is measured by, The method comprises the following steps of: According to the reflection coefficient, a simulation signal is generated: wherein represents the p-polarized reflection coefficient, represents the s-polarized reflection coefficient; the average of the cosine similarity of the measured band signal of each intermediate frequency of the reflection signal and the simulated band signal of the intermediate frequency corresponding to the thickness of the mixed layer, is represented as then the thickness corresponds to an average cosine similarity, which is the preferred degree of the thickness; The thickness of each mixed layer is preset, and a plurality of thicknesses are obtained by traversal; the optimal degree corresponding to each thickness is obtained by traversal of other thicknesses, and the simulation signal corresponding to the maximum optimal degree is taken as the best simulation signal of the reflection signal.

9. The method of claim 1, wherein, The method comprises the following steps of: For the reflection signal of any incident angle, the best simulation signal of the reflection signal is subtracted, and then the blank reflection signal of the incident angle is subtracted, and the remaining signal part is taken as the real interference reflection signal of the incident angle corresponding to the reflection signal. The thickness inversion of each coating layer is performed by using a classical algorithm Levenberg-Marquardt algorithm of ellipsometric inversion, and the thickness value of each coating layer is obtained.

10. A system for measuring the thickness of a multi-layer coated lens, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein, The processor executes the computer program to realize the steps of the thickness measurement method of the multilayer coated lens according to any one of claims 1-9.

Images