Optimization Method for Multi-Metal Single Crystal Purification Based on Learning Model

Abstract

The present invention provides an optimization method for the purification of multi-metal single crystals based on a learning model, which relates to the technical field of data processing. According to the defect characteristics of multi-metals and the impurity-metal repulsive force field parameters, an adsorption energy prediction model is trained. The trained adsorption energy prediction model is used to predict the adsorption energy of multi-metal impurities under different defect configurations, and the optimal defect configuration is selected according to the adsorption energy prediction results. Based on the optimal defect configuration, the adsorption process of each metal within the defect configuration is simulated, the propagation and adsorption behaviors of each metal impurity in the passage are analyzed, and passage simulation data is obtained. According to the purification target, the passage simulation data is optimized to obtain a multi-metal purification scheme, solving the technical problems of low purification efficiency and poor impurity control accuracy existing in the prior art. The technical effect of improving the purification efficiency and accuracy of multi-metal single crystals is achieved by constructing an adsorption energy prediction model and combining with the actual purification target.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an optimization method for multi-metal single crystal purification based on a learning model. Background Art

[0002] With the wide application of high-performance materials in high-end industries such as aerospace, microelectronics, new energy, and precision manufacturing, multi-metal single crystal materials have become one of the key basic materials due to their excellent mechanical properties, electrical and thermal conductivity, and chemical stability. However, during the preparation process of multi-metal single crystal materials, various trace impurities are likely to be mixed in, which will cause problems such as lattice distortion, decreased electron mobility, and deteriorated interface performance, seriously affecting the functional performance and reliability of the materials. Traditional purification methods for multi-metal materials, such as zone melting, electromigration purification, and chemical treatment, have the defects of low purification efficiency and difficulty in precise control.

[0003] The prior art has technical problems of low purification efficiency of multi-metal single crystals and poor impurity control accuracy. Summary of the Invention

[0004] The present application provides an optimization method for multi-metal single crystal purification based on a learning model, which is used to solve the technical problems of low purification efficiency of multi-metal single crystals and poor impurity control accuracy in the prior art.

[0005] In view of the above problems, the present application provides an optimization method for multi-metal single crystal purification based on a learning model. The method includes: training to obtain an adsorption energy prediction model according to the defect characteristics of multi-metals and the impurity-metal repulsive force field parameters; using the trained adsorption energy prediction model to predict the adsorption energy of multi-metal impurities under different defect configurations, and screening the optimal defect configuration according to the adsorption energy prediction results; based on the optimal defect configuration, simulating the adsorption process of each metal within the defect configuration, analyzing the propagation and adsorption behaviors of each metal impurity in the channel, and obtaining channel simulation data; and optimizing the channel according to the purification target for the channel simulation data to obtain a multi-metal purification scheme.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0007] The method provided by the embodiment of the present application trains an adsorption energy prediction model according to the defect characteristics of multi-metals and the impurity-metal repulsive force field parameters; uses the trained adsorption energy prediction model to predict the adsorption energy of multi-metal impurities under different defect configurations, and screens the optimal defect configuration according to the adsorption energy prediction results; based on the optimal defect configuration, simulates the adsorption process of each metal within the defect configuration, analyzes the propagation and adsorption behaviors of each metal impurity in the passage, and obtains passage simulation data; optimizes the passage according to the purification target for the passage simulation data to obtain a multi-metal purification scheme. It achieves the technical effect of improving the purification efficiency and accuracy of multi-metal single crystals by constructing an adsorption energy prediction model and combining with the actual purification target. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0009] Figure 1 Schematic flow chart of the multi-metal single crystal purification optimization method based on a learning model provided by the present application;

[0010] Figure 2 Schematic flow chart of obtaining an adsorption energy prediction model in the multi-metal single crystal purification optimization method based on a learning model provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0011] The present application provides a multi-metal single crystal purification optimization method based on a learning model, which is used to solve the technical problems of low purification efficiency and poor impurity control accuracy in the existing technology for multi-metal single crystals. It achieves the technical effect of improving the purification efficiency and accuracy of multi-metal single crystals by constructing an adsorption energy prediction model and combining with the actual purification target.

[0012] Next, the technical solutions in the present invention will be clearly and completely described with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited by the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. Additionally, it should be noted that for the sake of description, only the parts related to the present invention are shown in the drawings rather than all of them.

[0013] As Figure 1 shown, the present application provides a multi-metal single crystal purification optimization method based on a learning model, and the method includes:

[0014] S100: Based on the defect characteristics of multiple metals and the impurity-metal repulsion field parameters, an adsorption energy prediction model is trained.

[0015] Specifically, a polymetallic single crystal is a long-range ordered single-phase crystal composed of two or more metal elements. Its atoms are arranged in a strict periodic pattern throughout the material. It has high anisotropy and excellent physical properties, such as high-temperature strength and electrical properties. It is widely used in aerospace engine blades, semiconductor substrates, quantum materials and other fields. Such as nickel-based alloys, niobium-titanium superconducting single crystals, semiconductor-doped single crystal silicon, etc. Polymetallic single crystal defects refer to the non-ideal state of atomic arrangement or chemical composition in materials composed of two or more metals, just like a puzzle with a few missing parts or crooked arrangement. The defects in the crystal material are the places where the atomic arrangement is imperfect. Common defect types include vacancies, dislocations, stacking faults, etc. Vacancies are point defects at the atomic scale, such as vacancies formed by the loss of Ni atoms in Ni-based single crystals. Dislocations refer to linear defects generated when the crystal grows or is subjected to force, and stacking faults are errors in the stacking order of atomic layers. In addition, defects in polymetallic materials can adsorb impurities and are natural impurity adsorption sites. Their adsorption capacity depends on the defect type and impurity characteristics, such as impurity size, charge, chemical affinity, etc. By regulating the defect type, the capture of target impurities can be enhanced, thereby improving the purification capacity of multi-metal single crystals. The impurity-metal repulsion field parameter refers to the repulsive force between metal atoms and impurity atoms due to the overlap of electron clouds. Through the repulsive force field, impurities can be pushed from the inside of the metal lattice to specific areas, such as defect channels. For example, in nickel-based alloys, the strong repulsive force between lead (Pb) and nickel (Ni) forces lead atoms to move to the grain boundaries.

[0016] Based on this, collect and extract the structural information of common defects such as vacancies, dislocations, stacking faults, etc. in multiple groups of different multi-metal monomer systems from literature databases or large databases, and extract defect characteristics, including defect category, size, position, defect electron density, etc., as well as impurity-metal repulsive force field parameters. The impurity-metal repulsive force field parameters include the repulsive force field strength and action range in different metal matrices. Use the defect characteristics and impurity-metal repulsive force field parameters as training data, and based on machine learning methods, input the defect characteristics and force field parameters for model training to obtain an adsorption energy prediction model. The models that can be selected include extremely randomized forests, support vector machines, neural network models, etc. Adsorption energy refers to the energy required or released when an impurity atom is adsorbed at a certain defect site, usually expressed in electron volts (eV). The lower the adsorption energy (the larger the negative number), the more stable the adsorption. For example, if the adsorption energy of lead (Pb) at a vacancy is -1.2 eV, it means that it especially likes to stick to the vacancy and is more secure than sticking to a normal position. By training to obtain an adsorption energy prediction model, it is possible to quickly and accurately predict the adsorption energy of a certain impurity in this environment, and then by designing defects to make the defect adsorption energy very low, impurity adsorption can be carried out to improve the purification efficiency and accuracy of multi-metal single crystals.

[0017] S200: Use the trained adsorption energy prediction model to predict the adsorption energy of multi-metal impurities in different defect configurations, and screen the optimal defect configuration according to the adsorption energy prediction results.

[0018] Specifically, the trained adsorption energy prediction model can output the adsorption energy value under this combination by inputting defect characteristics and impurity-metal repulsive force field parameters. Then, use the trained adsorption energy prediction model to batch predict the multi-metal impurities in all alternative defect configurations, quickly obtain the adsorption energy of each impurity-defect combination, and then evaluate the adsorption ability of each defect structure for different impurities, and accordingly screen out the optimal defect configuration in the purification process. For example, prepare a dataset of possible defect configurations in a multi-metal. Each data record includes characteristics such as defect type, structural size, electron density, etc., as well as the corresponding force field parameters. For example, generate the following sample configurations in a Ni-Cu alloy: Configuration A: vacancy (3.5 Å), electron density 0.38 e / ų, Configuration B: dislocation (4.2 Å), electron density 0.40 e / ų. Then, for each impurity, such as Zn, Pb, form sample pairs by combining it with each defect configuration and input them into the model for adsorption energy prediction. For example, the prediction results are as follows:

[0019]

[0020] Next, according to the screening strategy, the configuration with the lowest (most negative) adsorption energy value is selected using the adsorption energy prediction results, or the optimal defect configuration suitable for the purification path is obtained based on comprehensive indicators such as the average adsorption capacity and total adsorption amount of multiple impurities. The optimal defect configuration refers to the defect structure with the best adsorption capacity for various metal impurities. It usually manifests as the configuration with the most negative adsorption energy or the highest comprehensive score of adsorption capacity. For example, if the main goal is to remove Pb, configuration A is selected as the optimal configuration, representing the site where impurities are most easily captured. If Zn needs to be considered as well, a mixed configuration optimization is adopted for comprehensive judgment. By predicting and screening the adsorption energies of different defect configurations, the defect configuration with the strongest adsorption capacity for specific impurities can be quickly screened out, providing a clear direction for subsequent adsorption pathway modeling and crystal structure regulation, improving the adaptability and intelligence level of the purification scheme, thereby enhancing the purification efficiency and quality and reducing the impurity residue in the multi-metal single crystal.

[0021] S300: Based on the optimal defect configuration, simulate the adsorption process of each metal within the defect configuration, analyze the propagation and adsorption behaviors of various metal impurities in the pathway, and obtain pathway simulation data.

[0022] Specifically, after determining the optimal defect configuration, simulate the adsorption process of each metal within the defect configuration based on this optimal defect configuration, and analyze the propagation paths and adsorption behaviors of various impurities within the crystal. Adsorption process simulation refers to reconstructing the entire process of impurities entering the crystal from the material surface, propagating in the defect region, and finally being adsorbed at the atomic scale using computational methods. This process requires simulating the behavior of impurities crossing different energy barriers, i.e., energy obstacles, in the lattice, and molecular dynamics simulation or transition state search algorithms can be used for dynamic path analysis. During the simulation, record the propagation trajectories, jumping paths, residence times, and final adsorption positions of each impurity atom in the defect region, and simultaneously calculate the changes in adsorption energy and migration energy barriers along the path, and form the pathway simulation data. The pathway simulation data includes, but is not limited to: initial position, termination position, path coordinate sequence, adsorption energy, residence point distribution, etc. Through simulation analysis, the diffusion trends and adsorption capacity differences of different impurities are revealed, which helps to identify the types of impurities that need to be preferentially removed or are likely to remain, providing a data basis for subsequent pathway optimization and structure regulation based on real behavior simulation, and significantly improving the accuracy and reliability of the multi-metal single crystal purification path.

[0023] S400: Optimize the pathway according to the purification target for the pathway simulation data to obtain a multi-metal purification scheme.

[0024] Furthermore, the purification target includes a single crystal structure parameter target and a purification threshold target.

[0025] Specifically, after completing the pathway simulation, in order to achieve effective purification of multiple metal impurities, it is necessary to further analyze and optimize the obtained pathway simulation data in combination with the established purification objectives, so as to construct a multi-metal purification scheme that meets the expected performance requirements. First, clarify the purification objectives. The purification objectives include single-crystal structure parameter objectives and purification threshold objectives. Among them, the single-crystal structure parameter objectives refer to the structural property requirements that the crystal should achieve after final purification, such as crystal geometry, lattice stability, surface energy control, etc., which are the key indicators for measuring the structural quality. The purification threshold objective refers to the target value to which the impurity concentration needs to be reduced. For example, the content of a certain metal impurity must be lower than 0.01 at.% (atomic percentage) or the maximum tolerable number of impurities per unit volume, which is the core standard for measuring the purification efficiency and completion. Then, taking the purification objectives as constraints, further analyze and optimize the pathway simulation data to obtain a multi-metal purification scheme. For example, in the scenario of purifying lead (Pb) impurities in nickel metal, since lead particularly likes vacancies and the adsorption energy is -1.2 eV, therefore, by optimizing the propagation adsorption pathway, the obtained purification scheme is to bombard nickel with ions, create a bunch of vacancies by creating defects, utilize the adsorption energy to aggregate and adsorb the lead impurities onto the created vacancy defects for impurity capture. Finally, perform impurity discharge. By heating the material and using the characteristic that the desorption energy barrier of Pb decreases at high temperature, lead is discharged from the vacancies, and then the surface enrichment layer is removed by polishing to achieve the removal of lead impurities. By quantitatively introducing the purification objectives into the optimization process, the transformation from data to application scenarios is realized, and multi-metal purification strategies can be adaptively generated according to different impurity types and concentration requirements, avoiding blind structural adjustment, ensuring that each step of regulation serves the final purification objective, reducing impurity residues, and improving the purification efficiency and quality.

[0026] Furthermore, as Figure 2 shown, step S100 of training an adsorption energy prediction model according to the defect characteristics of multiple metals and the impurity-metal repulsive force field parameters includes: S110: collecting defect characteristic data of multiple metals and impurity-metal repulsive force field parameter data, where the defect characteristic data includes defect category, size, position, and defect electron density, and the impurity-metal repulsive force field parameter data includes force field intensity and action distance; S120: performing adsorption characteristic correlation analysis based on the defect characteristic data and the impurity-metal repulsive force field parameter data to construct a training data set; S130: using the training data set for model training convergence to obtain the adsorption energy prediction model.

[0027] Specifically, first, comprehensive training data are collected, including two types of core parameters closely related to the adsorption behavior: one is the defect characteristic data of multi-metals, and the other is the repulsive force field parameter data between impurities and metals. Among them, the defect characteristic data include the type of defects (such as vacancies, dislocations, stacking faults, etc.), size (i.e., the influence range of defects, with the unit of Å), spatial position (three-dimensional coordinates in the unit cell), and defect electron density (referring to the electron cloud distribution concentration in this region, usually with the unit of e / ų). These characteristics directly determine the adsorption possibility and energy state of impurities at defects. The repulsive force field parameters between impurities and metals mainly measure the repulsive interaction between impurity atoms and matrix atoms, specifically including the force field strength (i.e., the potential energy size of the impurity being repelled) and the action distance (the effective radius affected by the force field). The force field strength and action distance of the repulsive force field can be calculated by molecular dynamics simulation tools such as LAMMPS and modeled using potential function models such as the Lennard-Jones potential or Morse potential. Then, through statistical methods, such as the Pearson correlation coefficient, the adsorption characteristic correlation analysis is carried out on the defect characteristic data and the repulsive force field parameter data between impurities and metals, quantifying the relationship between defect attributes and impurity adsorption behavior, and constructing a training data set. The training data set includes input feature defect types, repulsive force field parameter data between impurities and metals, and label values, that is, the measured or calculated adsorption energy values. The adsorption energy values can be obtained through first-principles calculation software such as VASP or extracted from the literature database. Then, a machine learning algorithm is used for model training. For example, the extremely randomized forest model is selected. During the training process, the data set is divided into a training set and a test set, such as in an 8:2 ratio. Model fitting and hyperparameter tuning are carried out on the training set, and the model performance is evaluated on the test set to ensure its prediction accuracy and generalization ability. When the loss function of the model no longer decreases significantly with the increase of the number of iterations and reaches a stable state, it indicates that the model training converges, and an adsorption energy prediction model after training is obtained. By constructing the adsorption energy prediction model, the prediction speed and accuracy of the adsorption energy can be improved, providing a basis for the subsequent simulation and optimization of the purification path of multi-metals and achieving fast and accurate purification effects.

[0028] Furthermore, according to the defect characteristic data and the repulsive force field parameter data between impurities and metals, an adsorption characteristic correlation analysis is carried out to construct a training data set, including: analyzing the adsorption characteristics of metal impurities and defect configurations according to the defect characteristic data to determine the adsorption energy of the impurity-defect configuration; performing a fusion analysis on the repulsive force field parameter data between impurities and metals and the adsorption energy of the impurity-defect configuration to determine the cooperative adsorption coefficient of the repulsive force field parameters on the adsorption energy of the impurity-defect configuration; constructing a training data set according to the adsorption energy of the impurity-defect configuration and the cooperative adsorption coefficient of the repulsive force field parameters, where the data label in the training data set is the adsorption energy.

[0029] Specifically, using first-principles software or molecular simulation tools, such as VASP and LAMMPS, based on the defect characteristic data, calculate and simulate the adsorption behaviors of different metal impurities in different defect configurations, such as adsorption tendency and adsorption energy change, to identify which defect structures are more friendly or repulsive to impurities, and obtain the corresponding impurity-defect configuration adsorption energy. The impurity-defect configuration adsorption energy refers to the energy released when an impurity atom binds to a specific defect, and a negative value indicates spontaneous adsorption. Then, perform a fusion analysis of the force field parameters and the adsorption energy to analyze the regulatory effect of the repulsive force field of the metal matrix on the adsorption behavior of the impurity atom. Specifically, construct a comparison defect characteristic sample group, that is, under the same defect configuration, adjust the impurity-metal force field strength or interaction distance, record the change values of the adsorption energy respectively, calculate the adsorption energy difference (ΔE), and based on this difference and the basic adsorption energy, obtain the cooperative adsorption coefficient through a normalization function or ratio calculation. And construct a training data set according to the cooperative adsorption coefficient of the impurity-defect configuration adsorption energy and the repulsive force field parameters, and the data label in the training set is the adsorption energy. By fusing the analysis of the structural characteristics and the physical interaction data, the model training data set covers the real defect structures and the actual changes in the repulsive force field, improving the generalization ability and prediction accuracy of the model.

[0030] Further, perform an adsorption characteristic analysis of the metal impurity and the defect configuration according to the defect characteristic data to determine the impurity-defect configuration adsorption energy, including: based on the defect characteristic data, analyze the adsorption probability of different metal impurities between different defect configurations; experimentally measure the measured adsorption energy of different metal impurities under different defect configurations; align the adsorption probability and the measured adsorption energy according to the metal impurity and the defect configuration to determine the impurity-defect configuration adsorption energy.

[0031] Specifically, first, using the collected defect feature data, the adsorption tendencies of various metal impurities in multiple defect configurations are analyzed through first-principles calculations or molecular simulation tools. Among them, the defect feature data includes the type, size, position, and electron density of the defects. According to the defect feature data, the adsorption tendencies of different metal impurities in different defect configurations, such as vacancies and dislocations, are analyzed to predict the impurity adsorption probability, which reflects the influence trend of structural factors on the adsorption behavior. For example, for copper impurities in an aluminum matrix, it is found through calculation that the adsorption probability at vacancy defects is as high as 0.85, while the adsorption probability at the dislocation core is 0.32, indicating that vacancy defects are more likely to adsorb copper impurities. In this stage, a high-throughput calculation platform, such as MaterialsProject, can be used to achieve batch simulations to ensure coverage of the main defect configurations. Then, the actual adsorption energy of the above-mentioned impurity-defect combinations is measured through experimental means. The measurement methods can include first-principles calculations, such as calculating the energy difference before and after adsorption using VASP simulation, and surface science experiments, such as using scanning tunneling microscope (STM) combined with XPS energy spectrum analysis. After obtaining the adsorption probability and adsorption energy, alignment and integration are carried out. Using the metal impurity and defect configuration as the main key, the adsorption probability is paired with the corresponding measured adsorption energy to ensure that each training sample not only has a real physical adsorption energy label but also retains the tendency information of its structure-induced adsorption behavior. In this way, the adsorption energy of each group of impurity-defect configurations is finally determined and used for subsequent model training. By combining experimental data with structural feature data, high-quality data labels are formed, providing accurate and reliable adsorption energy labels for model training, helping the model better learn the causal relationship between the defect structure and adsorption energy, thereby improving the accuracy and generalization ability of the entire adsorption energy prediction model, enhancing the accuracy and reliability of adsorption energy prediction, and further improving the purification and optimization efficiency and accuracy of multi-metal single crystals.

[0032] Furthermore, the fusion analysis is carried out based on the impurity-metal repulsive force field parameter data and the impurity-defect configuration adsorption energy, including: constructing comparison defect feature data according to the impurity-metal repulsive force field parameter data, where the comparison defect feature data includes defect record data of the same metal impurity and the same defect configuration under different impurity-metal repulsive force field parameters; carrying out adsorption energy experimental measurement according to the comparison defect feature data to obtain the adsorption energy difference of the comparison defect feature data; and calculating and obtaining the cooperative adsorption coefficient according to the adsorption energy difference and the impurity-defect configuration adsorption energy under the corresponding non-repulsive force field.

[0033] Specifically, based on the existing parameter data of the impurity-metal repulsive force field, a set of comparison defect feature data is constructed. The construction principle of the comparison defect feature data is as follows: Under the conditions of the same impurity element and defect configuration, different force field strength and action distance parameters are set to form multiple experimental groups to observe the influence of the force field change on the adsorption energy. For example, for Zn impurities in the Ni-Cu alloy grain boundary defect configuration, three groups of force field parameters are set: (1) force field strength = 2.0 eV, action distance = 2.0 Å; (2) strength = 2.5 eV, action distance = 2.2 Å; (3) strength = 3.0 eV, action distance = 2.5 Å. These three groups constitute the comparison samples of this impurity-defect combination. Then, based on the comparison defect feature data, an experimental measurement of the adsorption energy is carried out, that is, under each different set of force field parameters, the adsorption energy value of the impurity at the defect site is measured respectively using the first-principles or molecular simulation calculation, and then the adsorption energy difference of the comparison defect feature data under different force fields is obtained. For example, if the adsorption energies measured in the above three settings are -0.74 eV, -0.69 eV, and -0.65 eV respectively, the adsorption energy differences can be calculated in turn as: ΔE1 = -0.69 - (-0.74) = +0.05 eV, ΔE2 = -0.65 - (-0.74) = +0.09 eV, where -0.74 eV is the basic adsorption energy of this impurity-defect combination under the condition of no external repulsive force field (or the lowest force field). Finally, based on the above adsorption energy differences and the basic adsorption energy in the force field-free state, the cooperative adsorption coefficient is calculated. The cooperative adsorption coefficient is used to quantify the regulation intensity of the force field on the adsorption energy. The formula is: cooperative adsorption coefficient = ΔE / ∣Ebase∣. Substituting the data, we get: cooperative adsorption coefficient 1 = 0.05 / 0.74 ≈ 0.068, cooperative adsorption coefficient 2 = 0.09 / 0.74 ≈ 0.122. The cooperative adsorption coefficient is a dimensionless index, and the larger the value, the stronger the regulation of the force field on the adsorption behavior. The cooperative adsorption coefficient, as one of the input features, is used to train the adsorption energy prediction model, which not only enhances the multi-dimensional expression ability of the model training data but also effectively makes up for the prediction blind area caused by only relying on static structural features, enabling the adsorption energy prediction model to have a sensitive response ability to complex physical force field changes, thus achieving more accurate and reliable predictions and providing reliable data support for the subsequent purification optimization of multi-metal single crystals.

[0034] Furthermore, using the training data set for model training convergence to obtain the adsorption energy prediction model includes: dividing the training data set according to the splitting ratio to construct a training set and a test set; selecting the extremely randomized forest algorithm for training the adsorption energy prediction model. Among them, the adsorption energy prediction model is trained and the model hyperparameters are adjusted using the training set, and the trained model is verified and optimized using the test set until the convergence target is reached to determine the adsorption energy prediction model.

[0035] Specifically, after the construction of the training dataset is completed, the constructed training dataset is subjected to dataset splitting processing. Each sample data in the training dataset contains multiple input features, such as defect type, defect size, defect electron density, impurity-metal repulsive force field parameter, co-adsorption coefficient, etc., and a corresponding output label, that is, the adsorption energy. To ensure the generalization ability of the model on new data, the dataset can be divided into a training set and a test set according to a certain ratio, such as a ratio of training set:test set of 8:2. Among them, the training set is used for the learning process of the model, and the test set is used to verify the prediction performance of the model for unseen samples. The training set is used to fit and train the extremely randomized forest model. The extremely randomized forest is an ensemble model based on decision trees. By constructing multiple trees and introducing more randomness in the feature and sample dimensions, the generalization ability of the model is improved and overfitting is prevented. During the training process, the key hyperparameters of the model are adjusted, such as the number of trees, the maximum tree depth, the minimum number of samples for splitting, etc. Hyperparameter optimization can adopt methods such as grid search or random search. After the training is completed, the model is applied to the test set for prediction. By comparing the predicted value with the true adsorption energy label, the mean squared error is used to evaluate the model performance. If the model does not reach the preset convergence target, that is, the prediction accuracy is insufficient, then the hyperparameters are adjusted and the training is continued for optimization until the model converges stably, and an adsorption energy prediction model is obtained. For example, using a set of parameters in the test set, the input features are: defect type = vacancy, size = 4.0 Å, electron density = 0.32 e / ų, force field strength = 1.8 eV, co-adsorption coefficient = 0.45, and the adsorption energy predicted by the model is -0.82 eV. Compared with the true value of -0.80 eV, the error is only 0.02 eV, indicating that the model has good accuracy. By establishing a non-linear mapping relationship from the structural and force field features to the adsorption energy, the rapid and accurate prediction of the adsorption behavior of the impurity and defect combination is realized, thereby improving the automation and intelligence level of the overall purification path design and enhancing the purification efficiency and accuracy.

[0036] Furthermore, the optimal defect configuration is screened according to the adsorption energy prediction result, including: obtaining the adsorption energy prediction results of the multi-metal impurities under each defect configuration; aiming at maximizing the adsorption amount of the multi-metal impurities, searching for the defect configuration according to the adsorption energy prediction result to obtain the optimal defect configuration.

[0037] Specifically, after predicting the adsorption energies of multi-metal impurities under different defect configurations, the adsorption performances of each defect configuration are further analyzed and compared based on the prediction results to screen out the optimal defect configuration that meets the purification requirements. First, organize the prediction results of the adsorption energies of multi-metal impurities under each defect configuration. The adsorption energy refers to the energy change of a certain impurity at a specific defect position, usually negative, and the more negative the value, the stronger the adsorption. The prediction results are from the adsorption energy prediction model completed in the previous stage of training. By inputting various defect parameters and impurity-metal repulsive force field parameters, the corresponding adsorption energy can be quickly output. Then, based on the obtained multiple adsorption energy prediction results, all defect configurations are screened with the maximization of the adsorption capacity as the optimization goal. The adsorption capacity refers to the comprehensive performance of the adsorption ability of a certain defect configuration to multiple impurities, usually quantified in two ways: the sum of adsorption energies and the average adsorption energy of impurities. The sum of adsorption energies is the sum of the absolute values of the adsorption energies of all impurities in this configuration. The larger the value, the stronger the adsorption ability. The average adsorption energy of impurities refers to calculating the average value of the adsorption energies of each impurity in this configuration, reflecting the general adsorption property of this configuration to impurities. In the actual screening process, the pandas library in Python can be used to aggregate, sort, and screen the adsorption energy prediction table. Through the search and screening of defect configurations, the defect structure that is most conducive to the adsorption of multiple impurities can be efficiently located, providing a high-quality initial structure selection for subsequent adsorption path simulation and purification strategy optimization, achieving precise and targeted purification, significantly enhancing the capture efficiency of target impurities, and improving the quality and efficiency of impurity purification.

[0038] Furthermore, according to the purification target, the path simulation data is optimized for the path, including: obtaining the path simulation data of multi-metal impurities, including adsorption positions, adsorption energies, and propagation and migration paths; using the maximization of the purification target as the evaluation value, and searching for solutions for the geometric shape, size, and surface properties of the single crystal based on the path simulation data to optimize the propagation and adsorption path.

[0039] Specifically, first, obtain the channel simulation data of multi-metal impurities, including adsorption positions, adsorption energies, and propagation and migration paths. The adsorption position refers to the final stopping point or high-adsorption probability region of metal impurities in the crystal structure, usually the coordinate points near specific defects. The adsorption energy is used to describe the energy change of impurity adsorption at a specific position, and the more negative the value, the stronger the adsorption. The propagation and migration path refers to the path trajectory of impurities entering the structure from the surface or internal distribution points, undergoing diffusion, and finally reaching the adsorption point, as well as the energy barriers, jump frequencies, etc. therein. Then, maximize the purification target as the evaluation value. Based on the existing channel simulation data, construct an adjustable parameter space, take the geometric shape and size, and surface properties of the single crystal as optimization variables, and use a multi-objective optimization algorithm to continuously search and evaluate the contributions of different structural combinations to the purification target during the iteration process, realize the overall optimization of the propagation and adsorption channels, and obtain a multi-metal purification scheme with excellent performance and stable structure. By optimizing the propagation and adsorption channels, the purification efficiency and directional control ability can be improved, enabling impurities to aggregate faster and more concentratedly in the controllable region, thereby greatly improving the efficiency and controllability of metal purification.

[0040] Furthermore, optimizing the propagation and adsorption channels includes: performing multi-impurity characteristic quantification analysis based on the channel simulation data of the multi-metal impurities to obtain a multi-impurity adsorption quantification list; performing external field compensation gradient analysis based on the multi-impurity adsorption quantification list, making multi-level divisions according to the compensation gradient, and constructing a multi-level defect compensation network, where the multi-level defect compensation network includes multi-level capture defect parameters; obtaining dynamic external field cooperative regulation parameters; based on the dynamic external field cooperative regulation parameters, performing parameter regulation layer by layer on the multi-level defect compensation network to obtain an external field cooperative regulation strategy; adjusting the evaluation value according to the external field cooperative regulation strategy; based on the adjusted target, searching for geometric shape, size, and surface property schemes for the single crystal to optimize the propagation and adsorption channels.

[0041] Specifically, based on the channel simulation data, the adsorption behaviors of various metal impurities are quantitatively characterized, and the behaviors such as diffusion and adsorption of different metal impurities in the crystal are quantitatively described to form a comparable and sortable index system, including key characteristics such as adsorption energy, migration path length, energy barrier height, residence time, etc., and a multi-impurity adsorption quantification list is formed. The multi-impurity adsorption quantification list is an organized output table of the behavior characteristics of multiple impurities, recording the adsorption capacity and response characteristics of each impurity under specific structural or external field conditions, and is used to describe the adsorption difficulty and behavior trend of each impurity in different defect regions. According to the adsorption quantification data in the multi-impurity adsorption quantification list, an external field compensation gradient analysis is carried out, that is, the degree of change in adsorption efficiency in different regions under the conditions of not applying an external field and applying an external field is evaluated, so as to divide into multiple gradient-level regions, such as high-response regions, medium-response regions, and low-response regions. Based on the gradient division results, different control priorities are assigned respectively to construct a multi-level defect compensation network. The multi-level defect compensation network is a structure-response mapping model that organizes regions with different response levels in the crystal into a multi-layer structure. Each level contains corresponding defect capture parameters, such as the target adsorption energy range, path energy barrier threshold, structure coupling parameters, etc. The multi-level defect compensation network includes multi-level defect capture parameters, that is, the external field regulation conditions and response thresholds adapted to different regions. On this basis, dynamic external field cooperative regulation parameters applicable to each region are designed and obtained. The dynamic external field cooperative regulation parameters refer to physically adjustable parameters, such as magnetic field strength, electric field frequency, electric field direction, etc. The magnetic field strength regulation range is from 0.5 T to 2 T, and the application direction is along the crystal growth direction. The electric field frequency regulation range is from 10 Hz to 100 Hz, and the application direction is perpendicular to the crystal surface. The external field parameters are dynamically adjusted according to the migration path and adsorption behavior of the impurities, with the goal of flexibly configuring in different regions to achieve the optimal impurity guiding effect. For example, an axial static magnetic field (such as 2.0 ± 0.3 T) is preferentially applied to the first-level layer to enhance impurity capture; a rotating magnetic field (1.8 T, 500 rpm) and a low-frequency electric field (5 kV / cm, 10 kHz) are synchronously applied to the second-level layer to promote the migration of impurities to the target defect region; a high-frequency pulsed electric field (12 kV / cm, 100 kHz) is enabled in the third-level layer to inhibit the re-dissolution of impurities and ensure the continuity of purification. Then, based on the dynamic external field cooperative regulation parameters, the parameters of the multi-level defect compensation network are regulated layer by layer, and the external field is applied in stages to gradually optimize the migration path and adsorption behavior of the impurities, forming an external field cooperative regulation strategy. The external field cooperative regulation strategy is an intelligent regulation method that uses multiple external physical fields to hierarchically regulate the behavior of internal impurities in the crystal to achieve the goal of multi-impurity purification, and can achieve precise intervention in the impurity propagation path, adsorption efficiency, re-expansion risk, etc.Finally, the evaluation value is updated and the target is adjusted according to the external field collaborative regulation strategy. Based on the adjusted target, structural search and optimization are carried out on the geometry, size, and surface properties of the single crystal. For example, a genetic algorithm is used to search and optimize the geometry, size, and surface properties of the single crystal. The steps of the genetic algorithm include: initializing the population: generating a set of random single crystal geometry, size, and surface property parameters; fitness function: using the adjusted target as the fitness function to evaluate the fitness of each individual; selection operation: selecting individuals according to the fitness, and individuals with higher fitness have a higher probability of being selected; crossover and mutation: performing crossover and mutation operations on the selected individuals to generate a new population; iterative optimization: repeating the selection, crossover, and mutation operations until the preset number of iterations is reached or the fitness converges. Through optimization search, the impurity guiding property and adsorption efficiency are further improved to obtain the optimal solution for the overall propagation adsorption path. By combining data quantification, external field response, and hierarchical regulation, multi-dimensional factors are comprehensively integrated to further enhance the adaptability and accuracy of the purification strategy, strengthen the adsorption effect between impurities and metals, and improve the purification effect.

[0042] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0043] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and deformations to the present application without departing from the scope of the present application. Thus, if these modifications and deformations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and deformations.

Claims

1. A multi-metal single crystal purification optimization method based on a learning model, characterized in that: include: According to the defect characteristics of multi-metal and the impurity-metal repulsion field parameters, the adsorption energy prediction model is trained and obtained; The adsorption energy of multi-metal impurities under different defect configurations is predicted using the trained adsorption energy prediction model, and the optimal defect configuration is selected according to the adsorption energy prediction results; Based on the optimal defect configuration, simulating the adsorption process of each metal in the defect configuration, analyzing the propagation and adsorption behavior of each metal impurity in the pathway, and obtaining pathway simulation data, wherein the pathway simulation data includes adsorption position, adsorption energy, and propagation migration path; The pathway simulation data is optimized according to the purification target to obtain a multi-metal purification solution.

2. The multi-metal single crystal purification optimization method based on the learning model according to claim 1 is characterized in that: The method of training and obtaining an adsorption energy prediction model based on the defect characteristics of multiple metals and the impurity-metal repulsion field parameters includes: Collect defect feature data of multiple metals and impurity-metal repulsion field parameter data, wherein the defect feature data includes defect type, size, position, and defect electron density, and the impurity-metal repulsion field parameter data includes field strength and action distance; Perform adsorption feature correlation analysis based on the defect feature data and impurity-metal repulsion field parameter data to construct a training data set; The training data set is used to perform model training convergence to obtain the adsorption energy prediction model.

3. The multi-metal single crystal purification optimization method based on the learning model according to claim 2 is characterized in that: According to the defect feature data and the impurity-metal repulsion field parameter data, adsorption feature correlation analysis is performed to construct a training data set, including: Analyzing the adsorption characteristics of metal impurities and defect configurations according to the defect characteristic data to determine the impurity-defect configuration adsorption energy; According to the impurity-metal repulsive force field parameter data and the impurity-defect configuration adsorption energy, a synergistic adsorption coefficient of the repulsive force field parameter on the impurity-defect configuration adsorption energy is determined; A training data set is constructed according to the impurity-defect configuration adsorption energy and the synergistic adsorption coefficient of the repulsive force field parameters, wherein the data label in the training data set is the adsorption energy.

4. The multi-metal single crystal purification optimization method based on the learning model according to claim 3 is characterized in that: The adsorption characteristics of metal impurities and defect configurations are analyzed according to the defect characteristic data to determine the impurity-defect configuration adsorption energy, including: Based on the defect characteristic data, analyzing the adsorption probability of different metal impurities between different defect configurations; The adsorption energies of different metal impurities in different defect configurations were experimentally determined; The adsorption probability is performed according to the metal impurities and defect configurations, and the adsorption energy is aligned to determine the impurity-defect configuration adsorption energy.

5. The multi-metal single crystal purification optimization method based on the learning model according to claim 4 is characterized in that: A fusion analysis is performed based on the impurity-metal repulsion field parameter data and the impurity-defect configuration adsorption energy, including: According to the impurity-metal repulsion field parameter data, constructing comparison defect feature data, the comparison defect feature data including defect record data of the same metal impurity and the same defect configuration under different impurity-metal repulsion field parameters; According to the compared defect characteristic data, an adsorption energy experimental measurement is performed to obtain an adsorption energy difference of the compared defect characteristic data; The synergistic adsorption coefficient is calculated based on the adsorption energy difference and the adsorption energy of the impurity-defect configuration under the corresponding non-repulsive force field.

6. The multi-metal single crystal purification optimization method based on learning model according to claim 3, characterized in that: The training data set is used to perform model training convergence to obtain the adsorption energy prediction model, including: Splitting the training data set according to the splitting ratio to construct a training set and a test set; The extreme random forest algorithm is selected to train the adsorption energy prediction model, wherein the adsorption energy prediction model is trained and the model hyperparameters are adjusted using the training set, and the trained model is verified and optimized using the test set until the convergence target is reached to determine the adsorption energy prediction model.

7. The multi-metal single crystal purification optimization method based on learning model according to claim 1, characterized in that: Screen the optimal defect configuration based on the adsorption energy prediction results, including: Obtain the adsorption energy prediction results of multi-metal impurities under various defect configurations; With the goal of maximizing the adsorption amount of the multi-metal impurities, a defect configuration search is performed based on the adsorption energy prediction result to obtain the optimal defect configuration.

8. The multi-metal single crystal purification optimization method based on learning model according to claim 1, characterized in that: Optimizing the pathway simulation data according to the purification target, including: Taking the maximization of the purification target as the evaluation value, the geometric shape, size, and surface property scheme of the single crystal are searched based on the pathway simulation data to optimize the propagation adsorption pathway.

9. The multi-metal single crystal purification optimization method based on learning model according to claim 8, characterized in that: Optimize the propagation adsorption pathway, including: According to the channel simulation data of the multi-metal impurities, a multi-impurity feature quantitative analysis is performed to obtain a multi-impurity adsorption quantitative list; Based on the multi-impurity adsorption quantification list, an external field compensation gradient analysis is performed, and a multi-level division is performed according to the compensation gradient to construct a multi-level defect compensation network, wherein the multi-level defect compensation network includes multi-level capture defect parameters; Obtain dynamic external field coordinated control parameters; Based on the dynamic external field coordinated control parameters, the parameters of the multi-level defect compensation network are controlled layer by layer to obtain an external field coordinated control strategy; Performing target adjustment on the evaluation value according to the external field coordinated control strategy; Based on the adjustment target, the geometric shape, size and surface property solutions of the single crystal are searched to optimize the propagation adsorption path.

10. The multi-metal single crystal purification optimization method based on learning model according to claim 1, characterized in that: The purification targets include single crystal structure parameter targets and purification threshold targets.

Images