A high-precision non-destructive detection method and system for ion implantation defects

Abstract

The present invention discloses a high-precision non-destructive detection method and system for ion implantation defects, relating to the technical field of non-destructive detection. The method includes: selecting a measurement light source to collect reflected light intensity data, representing the intensity distribution of polarized light using the Stokes vector, calculating the ellipsometry parameters and determining the refraction angle data, calculating the complex refractive index through the Fresnel formula and determining the refractive index and extinction coefficient, and analyzing the overall complex dielectric constant of the material; dividing the crystal material region, obtaining the complex dielectric constant of the optimized region using the non-linear regression method, and constructing a three-layer composite medium model in combination with the region occupancy ratio. The method of the present invention can reveal the microscopic inhomogeneity inside the material through numerical analysis by calculating the Stokes parameters, improve the resolution of the damaged region, and can accurately distinguish the distributions of crystalline silicon, amorphous silicon and strained silicon after ion implantation by dividing the crystal material region, thereby improving the analysis ability of the microscopic structural changes of the material.

Description

Technical Field

[0001] The present invention relates to the technical field of nondestructive testing, and in particular to a high-precision nondestructive testing method and system for ion implantation defects. Background Art

[0002] Ion implantation technology, as a key process in the fields of semiconductor manufacturing and materials engineering, is widely used in doping regulation, surface modification, and nanostructure preparation. However, during the ion implantation process, lattice damage, strain, and non-uniform doping defects are often introduced. These defects directly affect the electrical, optical, and mechanical properties of the material. Therefore, high-precision nondestructive testing technology for ion implantation defects has always been an important research direction in semiconductor material analysis and quality control. Existing testing methods mainly include transmission electron microscopy, scanning electron microscopy, Raman spectroscopy analysis, photoluminescence, etc. Spectroscopic ellipsometry (SE) has gradually become an important tool for ion implantation defect detection in recent years due to its non-contact, high-precision, and broad-spectrum characteristics.

[0003] However, traditional ellipsometry methods mainly rely on single-angle or fixed-wavelength measurement modes, making it difficult to accurately analyze the changes in the complex dielectric constant in the depth direction of the material and unable to effectively distinguish the defect distribution in different depth layers. Secondly, existing spectroscopic ellipsometry methods are difficult to accurately analyze the defect distribution in different depths of the material. The calculated complex dielectric constant is usually the overall average value and cannot directly reflect the hierarchical nature of the internal structure of the material, resulting in limited accuracy of the calculation results. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a high-precision nondestructive testing method for ion implantation defects to solve the problems that traditional ellipsometry methods mainly rely on single-angle or fixed-wavelength measurement modes, making it difficult to accurately analyze the changes in the complex dielectric constant in the depth direction of the material and unable to effectively distinguish the defect distribution in different depth layers. Secondly, existing spectroscopic ellipsometry methods are difficult to accurately analyze the defect distribution in different depths of the material. The calculated complex dielectric constant is usually the overall average value and cannot directly reflect the hierarchical nature of the internal structure of the material, resulting in limited accuracy of the calculation results.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a high-precision nondestructive testing method for ion implantation defects, which includes:

[0008] Select a measurement light source to collect reflected light intensity data, represent the intensity distribution of polarized light using the Stokes vector, calculate the ellipsometry parameters and determine the refraction angle data, calculate the complex refractive index through the Fresnel formula and determine the refractive index and extinction coefficient, and analyze the overall complex dielectric constant of the material;

[0009] Conduct regional division of the crystal material, obtain the complex dielectric constant of the optimized region using the non-linear regression method, construct a three-layer composite dielectric model in combination with the regional occupancy ratio, optimize the calculation of the optimal solution of the regional occupancy ratio through the Lagrange optimization method, perform depth stratified Fourier transform to obtain the complex dielectric constant distribution in the depth domain, and perform integral calculation on the complex dielectric constant distribution data of different layers to confirm the normalized weight to calculate the material occupancy ratio of different layers;

[0010] Build a deep learning model based on DNN, optimize the model parameters using the Bayesian optimization method, perform prediction on the material defect distribution, and perform prediction matching degree prediction;

[0011] Encrypt the prediction data and perform secure storage.

[0012] As a preferred solution of the high-precision non-destructive detection method for ion implantation defects described in the present invention, wherein: the analysis of the overall complex dielectric constant of the material includes:

[0013] Based on spectroscopic ellipsometry for ion implantation defect detection, measure the reflected light intensities polarized in the parallel p and perpendicular s directions respectively, convert circularly polarized light into linearly polarized light by placing a quarter-wave plate QWP in the reflected light beam path, and sequentially adjust the polarizer to make it transmit right-handed circularly polarized and left-handed circularly polarized light, and measure the reflected light intensities after passing through right-handed and left-handed circularly polarized light;

[0014] Collect the reflected light intensity data after the polarizer by placing a wavelength modulation type wave plate in the light beam path;

[0015] Use the Stokes vector to represent the intensity distribution of polarized light and calculate the amplitude ratio and phase difference of the ellipsometry parameters;

[0016] Determine the reflection coefficient and determine the refraction angle data through Snell's law, and calculate the complex refractive index through the Fresnel formula and determine the refractive index and extinction coefficient;

[0017] Analyze and calculate the response value of the material to the electric field and the light absorption ability value of the material according to the refractive index and extinction coefficient, and calculate the overall complex dielectric constant.

[0018] As a preferred solution of the high-precision non-destructive detection method for ion implantation defects described in the present invention, wherein: the conduct of regional division of the crystal material, calculation of the optimal solution of the regional occupancy ratio, and confirmation of the normalized weight to calculate the material occupancy ratio of different layers includes:

[0019] During the ion implantation process, the crystal structure of silicon materials is divided into crystalline silicon regions, amorphous silicon regions, and strained silicon regions. The complex dielectric constants corresponding to crystalline silicon, amorphous silicon, and strained silicon are determined based on experimental data. Among them, the initial complex dielectric constant values of crystalline silicon, amorphous silicon, and strained silicon at different wavelengths are obtained from the database respectively. The nonlinear regression method is used to set the optimization goal as the minimum error between the three complex dielectric constants and the overall complex dielectric constant, and the Newton iteration method is used for optimization iteration until the iteration error no longer changes significantly, then the iteration is stopped and the three complex dielectric constants are determined;

[0020] The overall complex dielectric constant is calculated as the sum of the products of the three complex dielectric constants and the corresponding proportion values of crystalline silicon, amorphous silicon, and strained silicon respectively, and a three-layer composite medium model is constructed, where 、 and represent the proportion values of crystalline silicon, amorphous silicon, and strained silicon respectively;

[0021] By using the Lagrangian optimization method, the goal of maximizing the fitting degree of the complex dielectric constant model is set, and the optimal solutions of 、 and are optimized and calculated;

[0022] The overall complex dielectric constant is subjected to deep-layer stratified Fourier transform A-FFT to obtain the complex dielectric constant distribution in the depth domain;

[0023] Based on the complex dielectric constant after Fourier transform, the data corresponding to short wavelength, medium wavelength, and long wavelength are extracted respectively for inverse Fourier transform, and the complex dielectric constant distributions of the surface layer, medium-depth layer, and deep layer materials are obtained respectively. And according to the integral value calculation of the complex dielectric constant distribution data of different layers, the normalized weights of each layer are determined;

[0024] According to the normalized weights of different layers, calculate the products of the optimal proportion values of 、 and obtained by the Lagrangian optimization method respectively, and the material proportion values of different layers are obtained and normalized for verification.

[0025] As a preferred scheme of the high-precision non-destructive detection method for ion implantation defects described in the present invention, wherein: the deep learning model is constructed based on DNN, and the Bayesian optimization method is used to optimize the model parameters for predicting the material defect distribution, including,

[0026] Construct a deep learning model based on DNN, including an input layer, a hidden layer, and an output layer. Combine the complex permittivity data of different depth layers and the material occupancy ratio data corresponding to each depth layer to form a data matrix, and input it into the deep learning model through the input layer. Predict the material defect distribution through the hidden layer, including the predicted value of the material defect distribution area and the depth prediction value, and output it through the output layer.

[0027] Use the Bayesian optimization method to adjust the neural network parameters. Set the optimization goal to maximize the model prediction accuracy. Through Bayesian optimization iteration and update, stop the iteration when the update change is no longer obvious, and obtain the optimized model parameters.

[0028] As a preferred solution of the high-precision non-destructive detection method for ion implantation defects of the present invention, wherein: the prediction of the prediction matching degree refers to using the calibrated experimental measurement defect distribution data to calculate the matching degree of the predicted material defect distribution data of the deep learning model. Set an error threshold based on the experimental data. If the matching degree is greater than or equal to the error threshold, it is judged as accurate prediction.

[0029] As a preferred solution of the high-precision non-destructive detection method for ion implantation defects of the present invention, wherein: the selection of the measurement light source to collect the reflected light intensity data refers to selecting a supercontinuum laser source as the measurement light source, using a femtosecond pulsed light source, using a polarization modulator to generate incident light with a specific polarization state, allowing the incident light to pass through an adjustable polarizer to adjust the polarization direction, and collecting the reflected light intensity data through a high-sensitivity photodetector.

[0030] As a preferred solution of the high-precision non-destructive detection method for ion implantation defects of the present invention, wherein: the data encryption and secure storage of the predicted data refer to encrypting the predicted defect distribution data using the AES-256 encryption standard and performing encrypted storage through the KMS security key management.

[0031] In the second aspect, the present invention provides a system for a high-precision non-destructive detection method for ion implantation defects, including

[0032] An optical data acquisition module that selects and controls the measurement light source, adjusts the polarization state of the incident light, collects the reflected light intensity data, and measures the response of the material to different polarized lights.

[0033] An optical parameter calculation module that analyzes the polarization information of the light, calculates the response parameters of the material to the light, calculates the ellipsometry parameters, determines the refraction angle data, and further obtains the refractive index and extinction coefficient.

[0034] A three-layer composite medium module that divides the crystalline silicon, amorphous silicon, and strained silicon regions according to the material measurement data, obtains the complex permittivity of different regions through an optimization method, and constructs a composite medium model.

[0035] The material layer analysis module performs in-depth layer-by-layer Fourier transform, analyzes the complex dielectric constant distribution of different depth layers, calculates the normalized weights of different layers, and finally obtains the material proportion of different layers.

[0036] The deep learning defect prediction module uses DNN to predict the defect distribution, identifies the damaged areas in the material, and adjusts the model parameters through Bayesian optimization.

[0037] The matching degree verification module calculates the matching degree between the DNN prediction result and the experimental measurement data to judge the reliability of the prediction result.

[0038] The data security storage module securely stores the measurement data, calculation data, and prediction data.

[0039] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the high-precision non-destructive detection method for ion implantation defects as described in the first aspect of the present invention is implemented.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the high-precision non-destructive detection method for ion implantation defects as described in the first aspect of the present invention is implemented.

[0041] The beneficial effects of the present invention are as follows: by calculating the Stokes parameters, the microscopic inhomogeneity inside the material can be revealed through numerical analysis, improving the resolution of the damaged areas. Through the regional division of the crystal material, the distributions of crystalline silicon, amorphous silicon, and strained silicon can be accurately distinguished after ion implantation, improving the analysis ability of the microscopic structure changes of the material. By combining spectroscopic ellipsometry with layer-by-layer Fourier transform, material regional division, and optimization calculation, the solution can achieve precise layer-by-layer analysis of the microscopic structure and defects of the material based on high-precision polarization optical measurement. By calculating the Stokes vector, ellipsometric parameters, and deriving the refractive index and extinction coefficient through the Fresnel formula, the optical response of the material can be quantitatively described and directly correlated with the internal structure changes of the material after ion implantation, realizing non-destructive, high-precision, and depth-resolvable ion implantation defect detection. Description of the Drawings

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0043] Figure 1It is a schematic flow diagram of the high-precision non-destructive detection method for ion implantation defects in Example 1.

[0044] Figure 2 It is a schematic structural diagram of the high-precision non-destructive detection system for ion implantation defects in Example 1.

[0045] Figure 3 It is a schematic flow diagram of the crystal region division process of the high-precision non-destructive detection method for ion implantation defects in Example 1.

[0046] Figure 4 It is a schematic structural diagram of the DNN deep learning model of the high-precision non-destructive detection method for ion implantation defects in Example 1. Detailed implementation manners

[0047] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention with reference to the accompanying drawings of the specification.

[0048] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0049] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that can be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that excludes other embodiments.

[0050] Example 1, referring to Figures 1 to 4 , which is the first embodiment of the present invention. This embodiment provides a high-precision non-destructive detection method for ion implantation defects, including the following steps:

[0051] S1. Select a measurement light source to collect reflected light intensity data, use the Stokes vector to represent the intensity distribution of polarized light, calculate the ellipsometry parameters and determine the refraction angle data, calculate the complex refractive index through the Fresnel formula and determine the refractive index and extinction coefficient, and analyze the overall complex dielectric constant of the material.

[0052] Preferably, selecting a measurement light source to collect reflected light intensity data means using a supercontinuum laser source as the measurement light source, with a spectral range of 190 nm - 1200 nm, using a femtosecond pulsed light source (<100 fs), using a polarization modulator to generate incident light with a specific polarization state, allowing the incident light to pass through an adjustable polarizer to adjust the polarization direction, and collecting the reflected light intensity data through a high-sensitivity photodetector.

[0053] By using a supercontinuum laser as the measurement light source, compared with traditional single-wavelength laser sources, the optical responses of materials at multiple wavelengths can be obtained simultaneously, improving the spectral resolution of defect detection. Using a femtosecond pulsed light source can reduce time-resolved errors, reduce the influence of surface roughness and ambient light on the measurement, improve the spatio-temporal resolution of the signal, ensure more accurate measurement of ion implantation defects. By generating incident light with a specific polarization state through a polarization modulator and adjusting the polarization direction with a tunable polarizer, the response information of the material to different polarization directions can be comprehensively obtained. Using a high-sensitivity photodetector to collect the reflected light intensity data can ensure accurate measurement of the material reflection characteristics even in an environment with a low signal-to-noise ratio, improve the signal-to-noise ratio of the data, and enhance the measurement stability.

[0054] Furthermore, analyze the overall complex dielectric constant of the material, including:

[0055] Based on spectroscopic ellipsometry for ion implantation defect detection, measure the reflected light intensities polarized in the parallel p and perpendicular s directions respectively. By placing a quarter-wave plate QWP in the reflected beam path to convert circularly polarized light into linearly polarized light, and sequentially adjusting the polarizer to transmit right-handed circularly polarized and left-handed circularly polarized light, measure the reflected light intensities after passing through right-handed and left-handed circular polarization;

[0056] By placing a wavelength-modulated wave plate in the beam path and sequentially selecting polarized light in the 45° and 135° directions through a tunable polarizer, collect the reflected light intensities after the 45° and 135° polarizers;

[0057] Use the Stokes vector to represent the intensity distribution of polarized light, expressed as:

[0058] ;

[0059] ;

[0060] ;

[0061] ;

[0062] where represents the total light intensity, represents the degree of linear polarization from 0° to 90°, represents the degree of linear polarization from 45° to 135°, represents the degree of circular polarization, and respectively represent the reflected light intensities polarized in the parallel (p) and perpendicular (s) directions in the reflected light, and respectively represent the components with polarization directions of 45 degrees and 135 degrees, and respectively represent the light intensities of right-handed circular polarization and left-handed circular polarization;

[0063] And calculate the amplitude ratio and phase difference of the ellipsometry parameters, expressed as:

[0064] ;

[0065] ;

[0066] where represents the amplitude ratio, represents the phase difference;

[0067] Determine the reflection coefficient and determine the refraction angle data through Snell's law, and calculate the complex refractive index through Fresnel's formula and determine the refractive index and extinction coefficient, expressed as:

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] where and respectively represent the reflection coefficients in the parallel (p) and perpendicular (s) directions in the reflected light, i represents the imaginary unit, and can be expressed as , represents the incident angle, represents the refraction angle, determined by Snell's law, and can be expressed as , represents the refractive index of the incident medium, i.e., air, represents the complex refractive index of the material, n represents the effective refractive index, k represents the extinction coefficient, and can be determined by the formula where represents the intensity of light propagating to a depth x in the material, represents the intensity of the incident light, represents the wavelength of light;

[0073] Analyze and calculate the response value of the material to the electric field and the light absorption ability value of the material according to the refractive index and extinction coefficient, and calculate the overall complex dielectric constant, expressed as:

[0074] ;

[0075] ;

[0076] ;

[0077] wherein represents the response value of the material to the electric field, represents the light absorption ability value of the material, represents the overall complex dielectric constant.

[0078] Ion implantation defect detection by spectroscopic ellipsometry can improve the sensitivity and resolution of defect detection, can accurately analyze the optical properties of materials without damaging the samples. By measuring the reflected light intensities polarized in the parallel (p) and perpendicular (s) directions respectively, it can comprehensively capture the response of the material to light with different polarization directions, thereby enhancing the detection ability for material anisotropy and microstructural changes. During the adjustment process of the quarter-wave plate (QWP) and polarizer, by converting circularly polarized light into linearly polarized light and successively measuring the reflected light intensities after passing through right-handed and left-handed circular polarization, the optical rotatory power and chiral information of the material can be extracted;

[0079] By calculating the Stokes parameters, the reflected light characteristics of the material can be accurately analyzed. Especially when there are complex lattice defects, the microscopic inhomogeneity inside the material can be revealed through numerical analysis, improving the resolution of the damaged area. Calculating the amplitude ratio and phase difference of the ellipsometric parameters can directly reflect the interaction of the material with different polarized lights, can avoid the complexity of directly measuring the dielectric constant, and at the same time improve the measurement accuracy and stability, making the detection process more repeatable. Using Snell's law and Fresnel's formula to calculate the reflection coefficient and refractive index provides a non-destructive and high-precision method to measure the optical properties of materials. By accurately calculating the refraction angle, the understanding of the ion implantation depth and its influence can be improved, which can be further used for ion doping control in semiconductor manufacturing. Determining the extinction coefficient in the complex refractive index enables the quantification of the light absorption ability of the material, so as to be used to predict the optical loss of the material at different wavelengths;

[0080] Using the non-destructive spectroscopic ellipsometry method and combining with the calculation of Stokes vectors, ellipsometric parameters and complex dielectric constants, high-precision automated detection is achieved, reducing the complexity of sample preparation and improving the detection efficiency.

[0081] S2, perform crystal material region division, use the non-linear regression method to obtain the complex dielectric constant of the optimized region, construct a three-layer composite dielectric model in combination with the region occupancy ratio, optimize the calculation of the optimal solution of the region occupancy ratio by the Lagrangian optimization method, perform depth stratified Fourier transform to obtain the complex dielectric constant distribution in the depth domain, perform integral calculation on the complex dielectric constant distribution data of different layers, and confirm the normalized weight to calculate the material occupancy ratio of different layers;

[0082] Preferably, the crystal material region is divided, the optimal solution of the region occupancy ratio is calculated, and the normalized weights are used to calculate the material occupancy ratios of different layers, including:

[0083] During the ion implantation process, the crystal structure of silicon material is divided into crystalline silicon region, amorphous silicon region and strained silicon region. Based on experimental data, the complex dielectric constants corresponding to crystalline silicon, amorphous silicon and strained silicon are determined respectively. Among them, the initial complex dielectric constant values of crystalline silicon, amorphous silicon and strained silicon at different wavelengths are obtained from the database, and the non-linear regression method is used to set the optimization goal as the minimum error between the three complex dielectric constants and the overall complex dielectric constant, and the Newton iteration method is used for optimization iteration until the iteration error no longer changes significantly, then the iteration is stopped and the three complex dielectric constants are determined;

[0084] According to the sum of the products of the three complex dielectric constants and the corresponding ratios of crystalline silicon, amorphous silicon and strained silicon as the overall complex dielectric constant, a three-layer composite dielectric model is constructed, which is expressed as:

[0085] ;

[0086] Where represents the overall complex dielectric constant, , and represent the complex dielectric constants of crystalline silicon, amorphous silicon and strained silicon respectively, , and represent the ratios of crystalline silicon, amorphous silicon and strained silicon respectively;

[0087] By using the Lagrangian optimization method, the goal of maximizing the fitting degree of the complex dielectric constant model is set, and the optimal solutions of , and are optimized. The Lagrangian function is expressed as:

[0088] ;

[0089] Where represents the Lagrangian multiplier, which is used to ensure that the sum of all ratios is equal to 1;

[0090] Perform a deep-layer Fourier transform A-FFT on the overall complex dielectric constant to obtain the complex dielectric constant distribution in the depth domain, which is expressed as:

[0091] ;

[0092] ;

[0093] Where represents the complex permittivity after Fourier transform of the depth domain z, represents the wavelength of the overall complex permittivity, represents the depth-frequency component inside the material, reflecting the oscillatory behavior of the complex permittivity varying with depth z, and n represents the effective refractive index;

[0094] Based on data corresponding to short wavelength, medium wavelength, and long wavelength are extracted respectively for inverse Fourier transform, and the complex permittivity distributions of the surface layer, medium-depth layer, and deep-layer materials are obtained respectively. According to the integral values of the complex permittivity distribution data of different layers, the normalized weights of each layer are determined, expressed as:

[0095] ;

[0096] where represents the normalized weight of the i-th layer, and represent the upper and lower boundaries of the i-th depth layer respectively, represents the complex permittivity distribution data of the i-th layer, and A represents the sum of the integrals of the complex permittivities of all depth layers;

[0097] According to the normalized weights of different layers, calculate the product of the optimal occupancy ratios obtained by the Lagrangian optimization method for 、 and respectively, to obtain the material occupancy ratios of different layers and perform normalization verification.

[0098] By dividing the crystal material region, it is possible to accurately distinguish the distributions of crystalline silicon, amorphous silicon, and strained silicon after ion implantation, improve the analytical ability of the material microstructure changes. When constructing a three-layer composite dielectric model, by using the weighted sum of the complex permittivity ratios of crystalline silicon, amorphous silicon, and strained silicon, the change of the overall complex permittivity can be described more accurately. By using the Lagrangian optimization method, the non-physical solutions that may be brought by unconstrained optimization are avoided, improving the stability of the model and the credibility of the calculation results. At the same time, by maximizing the fitting degree of the complex permittivity model, it can be ensured that the calculation results of the material ratios are optimal within the global range. Through depth-stratified Fourier transform, the overall complex permittivity can be analyzed in the depth direction, realizing the spatial stratified characterization of material properties;

[0099] Multiply the normalized weights by the optimal occupancy ratios obtained by the Lagrangian optimization method to calculate the material occupancy ratios of different layers and perform normalization verification to ensure that the finally obtained material occupancy ratios numerically conform to the actual physical laws;

[0100] By combining spectroscopic ellipsometry with depth-stratified Fourier transform (A-FFT) for material region division and optimization calculation, the solution can, based on high-precision polarization optical measurement, achieve precise stratified analysis of the microstructure and defects of materials, and obtain the optimal material composition ratio based on mathematical optimization methods, enabling the defect detection to expand from traditional two-dimensional surface information to high-resolution three-dimensional spatial characterization. By calculating the refractive index and extinction coefficient through Stokes vectors, ellipsometric parameters, and Fresnel formula derivation, the optical response of the material can be quantitatively described and directly correlated with the internal structural changes of the material after ion implantation, realizing non-destructive, high-precision, and depth-resolvable ion implantation defect detection.

[0101] S3, construct a deep learning model based on DNN, use the Bayesian optimization method to optimize the model parameters, predict the material defect distribution, and predict the prediction matching degree;

[0102] Preferably, construct a deep learning model based on DNN, use the Bayesian optimization method to optimize the model parameters, and predict the material defect distribution, including,

[0103] Construct a deep learning model based on DNN, including an input layer, a hidden layer, and an output layer. Combine the complex dielectric constant data of different depth layers and the corresponding material ratio data of each depth layer to form a data matrix, and input the data matrix into the deep learning model through the input layer. Predict the material defect distribution through the hidden layer, including the predicted values of the material defect distribution region and the depth prediction value, and output through the output layer;

[0104] Use the Bayesian optimization method to adjust the neural network parameters, set the optimization goal to maximize the model prediction accuracy, stop the iteration when the update change is no longer obvious through Bayesian optimization iteration, and obtain the optimized model parameters.

[0105] Through the deep learning model based on DNN, the complex dielectric constant data and the material ratio data can be combined, enabling the material defect distribution prediction to have stronger global and data-driven capabilities. The neural network can capture the complex characteristics of defects in different depth layers through the non-linear mapping ability of the hidden layer, thereby realizing more detailed material damage analysis. Optimizing the neural network parameters through the Bayesian optimization method can avoid the inefficiency of traditional grid search or random search methods in the high-dimensional hyperparameter space.

[0106] Furthermore, predicting the prediction matching degree means using the calibrated experimental measurement defect distribution data to calculate the matching degree between the predicted material defect distribution data of the deep learning model, setting an error threshold based on the experimental data. If the matching degree is greater than or equal to the error threshold, it is judged that the prediction is accurate.

[0107] Through prediction verification of the matching degree, it is possible to ensure that the prediction results of the defect distribution of the deep learning model are highly consistent with the actual experimental measurement data, thereby improving the credibility and practical usability of the model. By calculating the prediction matching degree of the deep learning model, the deviation between the prediction data and the experimental measurement data can be quantified, and whether the prediction is qualified can be automatically judged by setting an error threshold. This makes the evaluation system of the entire model more standardized and automated, without relying on manual analysis of the prediction results, and improves the intelligence and efficiency of the material defect detection process.

[0108] S4. Encrypt the prediction data and store it securely;

[0109] Preferably, encrypting the prediction data and storing it securely means encrypting the predicted defect distribution data using the AES-256 encryption standard and storing it encrypted through KMS secure key management.

[0110] Through encrypted data storage, unauthorized access to the data during storage or transmission can be prevented, achieving the effect of encrypting sensitive data. And through secure key management, the data access protection effect of users can be further improved.

[0111] This embodiment also provides a system for a high-precision non-destructive detection method of ion implantation defects, including

[0112] An optical data acquisition module that selects and controls the measurement light source, adjusts the polarization state of the incident light, acquires the reflected light intensity data, and measures the response of the material to different polarized lights;

[0113] An optical parameter calculation module that analyzes the polarization information of light, calculates the response parameters of the material to light, calculates the ellipsometry parameters, determines the refraction angle data, and further obtains the refractive index and extinction coefficient;

[0114] A three-layer composite medium module that divides the crystalline silicon, amorphous silicon, and strained silicon regions according to the material measurement data, obtains the complex dielectric constants of different regions through an optimization method, and constructs a composite medium model;

[0115] A material layer analysis module that performs depth-layer Fourier transform, analyzes the complex dielectric constant distribution of different depth layers, calculates the normalized weights of different layers, and finally obtains the material proportion of different layers;

[0116] A deep learning defect prediction module that uses DNN for defect distribution prediction, identifies the damaged areas in the material, and adjusts the model parameters through Bayesian optimization;

[0117] A matching degree verification module that calculates the matching degree between the DNN prediction result and the experimental measurement data and judges the reliability of the prediction result;

[0118] The data security storage module securely stores measurement data, calculation data, and prediction data.

[0119] This embodiment also provides a computer device applicable to the case of the high-precision non-destructive detection method for ion implantation defects, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the high-precision non-destructive detection method for ion implantation defects proposed in the above embodiment.

[0120] The computer device can be a terminal. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0121] This embodiment also provides a storage medium with a computer program stored thereon. When the program is executed by a processor, it implements the high-precision non-destructive detection method for ion implantation defects proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0122] In summary, by calculating the Stokes parameters, the present invention can reveal the microscopic inhomogeneity inside the material through numerical analysis, improve the resolution of the damage area, and accurately distinguish the distributions of crystalline silicon, amorphous silicon, and strained silicon after ion implantation by dividing the crystal material region, thereby enhancing the analytical ability for changes in the microscopic structure of the material. By combining spectroscopic ellipsometry with depth-stratified Fourier transform, material region division, and optimized calculation, the solution can achieve accurate stratification analysis of the microscopic structure and defects of the material based on high-precision polarization optical measurement. By calculating the refractive index and extinction coefficient through the Stokes vector, ellipsometric parameter calculation, and Fresnel formula derivation, the optical response of the material can be quantitatively described and directly correlated with the internal structural changes of the material after ion implantation, realizing non-destructive, high-precision, and depth-resolvable ion implantation defect detection.

[0123] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A high-precision non-destructive detection method for ion implantation defects, characterized in that, Including: Select a measurement light source to collect reflected light intensity data, use the Stokes vector to represent the intensity distribution of polarized light, calculate the ellipsometry parameters and determine the refraction angle data, calculate the complex refractive index through the Fresnel formula and determine the refractive index and extinction coefficient, and analyze the overall complex dielectric constant of the material; Conduct crystal material region division, use the nonlinear regression method to obtain the complex dielectric constant of the optimized region, construct a three-layer composite medium model in combination with the region occupancy ratio, optimize and calculate the optimal solution of the region occupancy ratio through the Lagrange optimization method, perform depth-layer Fourier transform, obtain the complex dielectric constant distribution in the depth domain, and perform integral calculation on the complex dielectric constant distribution data of different layers to confirm the normalized weight and calculate the material occupancy ratio of different layers; Build a deep learning model based on DNN, use the Bayesian optimization method to optimize the model parameters, predict the material defect distribution, and perform prediction matching degree prediction; Encrypt the prediction data and perform secure storage; The conduct of crystal material region division, calculation of the optimal solution of the region occupancy ratio, confirmation of the normalized weight and calculation of the material occupancy ratio of different layers includes: During the ion implantation process, the crystal structure of silicon material is divided into crystalline silicon region, amorphous silicon region and strained silicon region. Based on experimental data, the complex dielectric constants corresponding to crystalline silicon, amorphous silicon and strained silicon are determined respectively. Among them, the initial complex dielectric constant values of crystalline silicon, amorphous silicon and strained silicon at different wavelengths are obtained from the database respectively. The nonlinear regression method is used to set the optimization goal as the minimum error between the three complex dielectric constants and the overall complex dielectric constant, and the Newton iteration method is used for optimization iteration until the iteration error no longer changes significantly, then stop the iteration and determine the three complex dielectric constants; Construct a three-layer composite dielectric model by taking the sum of the products of three complex dielectric constants and the corresponding proportion values of crystalline silicon, amorphous silicon, and strained silicon as the overall complex dielectric constant, where , and represent the proportion values of crystalline silicon, amorphous silicon, and strained silicon, respectively; By using the Lagrangian optimization method, set the goal of maximizing the goodness of fit of the complex permittivity model and optimize the calculation , and for the optimal solution; Perform depth-layer Fourier transform A-FFT on the overall complex dielectric constant to obtain the complex dielectric constant distribution in the depth domain; Based on the complex dielectric constant after Fourier transform, extract the data corresponding to short wavelength, medium wavelength and long wavelength respectively for inverse Fourier transform, obtain the complex dielectric constant distributions of the surface layer, medium depth layer and deep layer materials respectively, and determine the normalized weight of each layer according to the integral values of the complex dielectric constant distribution data of different layers; Calculate the product of the optimal occupancy ratios obtained by the Lagrangian optimization method according to the normalization weights of different layers and as well as to obtain the material occupancy ratios of different layers respectively, and perform normalization verification.

2. The high-precision non-destructive detection method for ion implantation defects according to claim 1, characterized in that: The analysis of the overall complex dielectric constant of the material includes: Based on spectroscopic ellipsometry for ion implantation defect detection, measure the reflected light intensities polarized in the parallel p and perpendicular s directions respectively. Place a quarter-wave plate QWP in the reflected light beam path to convert circularly polarized light into linearly polarized light, and sequentially adjust the polarizer to make it transmit right-handed circularly polarized and left-handed circularly polarized light, and measure the reflected light intensities after passing through right-handed and left-handed circularly polarized light; Collect the reflected light intensity data after the polarizer by placing a wavelength modulation type wave plate in the light beam path; Use the Stokes vector to represent the intensity distribution of polarized light, and calculate the amplitude ratio and phase difference of the ellipsometry parameters; Determine the reflection coefficient and determine the refraction angle data through Snell's law, and calculate the complex refractive index through the Fresnel formula and determine the refractive index and extinction coefficient; Analyze and calculate the response value of the material to the electric field and the light absorption ability value of the material according to the refractive index and extinction coefficient, and calculate the overall complex dielectric constant.

3. The high-precision non-destructive detection method for ion implantation defects according to claim 2, characterized in that: The deep learning model is constructed based on DNN, and the Bayesian optimization method is used to optimize the model parameters for predicting the distribution of material defects, including: Construct a deep learning model based on DNN, including an input layer, a hidden layer, and an output layer. Combine the complex permittivity data of different depth layers and the corresponding material occupancy ratio data of each depth layer to form a data matrix, and input it into the deep learning model through the input layer. Predict the distribution of material defects through the hidden layer, including the predicted values of the material defect distribution area and the depth prediction value, and output them through the output layer. Use the Bayesian optimization method to adjust the neural network parameters. Set the optimization goal to maximize the model prediction accuracy. Through Bayesian optimization iteration and update, stop the iteration when the update change is no longer obvious, and obtain the optimized model parameters.

4. The high-precision non-destructive detection method for ion implantation defects according to claim 3, characterized in that: The prediction of the prediction matching degree refers to using the calibrated experimental measurement defect distribution data to calculate the matching degree between the predicted material defect distribution data of the deep learning model. Set an error threshold based on the experimental data. If the matching degree is greater than or equal to the error threshold, it is judged that the prediction is accurate.

5. The high-precision non-destructive detection method for ion implantation defects according to claim 4, characterized in that: The selection of the measurement light source to collect the reflected light intensity data refers to selecting a supercontinuum laser source as the measurement light source, using a femtosecond pulse light source, using a polarization modulator to generate incident light with a specific polarization state, allowing the incident light to pass through an adjustable polarizer to adjust the polarization direction, and collecting the reflected light intensity data through a high-sensitivity photodetector.

6. The high-precision non-destructive detection method for ion implantation defects according to claim 5, wherein: The data encryption and secure storage of the predicted data refer to encrypting the predicted defect distribution data using the AES-256 encryption standard and performing encrypted storage through the KMS secure key management.

7. A system for a high-precision non-destructive detection method of ion implantation defects, based on the high-precision non-destructive detection method of ion implantation defects described in any one of claims 1 to 6, characterized in that: Including: An optical data acquisition module that selects and controls the measurement light source, adjusts the polarization state of the incident light, collects the reflected light intensity data, and measures the response of the material to different polarized lights. An optical parameter calculation module that analyzes the polarization information of light, calculates the response parameters of the material to light, calculates the ellipsometry parameters, determines the refraction angle data, and further obtains the refractive index and extinction coefficient. A three-layer composite dielectric module that divides the crystalline silicon, amorphous silicon, and strained silicon regions according to the material measurement data, obtains the complex permittivity of different regions through an optimization method, and constructs a composite dielectric model. A material layer analysis module that performs depth-layer Fourier transform, analyzes the complex permittivity distribution of different depth layers, calculates the normalized weights of different layers, and finally obtains the material occupancy ratio of different layers. A deep learning defect prediction module that uses DNN to predict the defect distribution, identifies the damaged areas in the material, and adjusts the model parameters through Bayesian optimization. A matching degree verification module that calculates the matching degree between the DNN prediction result and the experimental measurement data, and judges the reliability of the prediction result. A data security storage module that securely stores the measurement data, calculation data, and prediction data.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the high-precision non-destructive detection method for ion implantation defects described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the high-precision non-destructive detection method for ion implantation defects described in any one of claims 1 to 6.

Images