VASP-based metal surface adsorption energy differential charge analysis method and system

Abstract

The invention discloses a VASP-based metal surface adsorption energy differential charge analysis method and system, and relates to the technical field of metal surface adsorption energy differential charge analysis, and the method comprises the steps: carrying out the modeling and standardization processing of a metal surface-adsorbate system, and obtaining a standard data set; training the standard data set by adopting a deep neural network model to obtain a prediction model, inputting the standard data set into the prediction model, and outputting a prediction result; dynamically adjusting VASP calculation parameters by adopting an environmental effect correction method, and generating a VASP input file based on a prediction result and an actual operation condition; performing high-precision energy and electron density calculation on the input file by adopting a first principle calculation tool VASP to obtain electron density distribution; performing differential operation on the electron density distribution to obtain a differential charge density map; and performing three-dimensional rendering on the differential charge density map by adopting a visual analysis tool to obtain a visual analysis result.

Description

Technical Field

[0001] The present invention relates to the technical field of metal surface adsorption energy differential charge analysis, and in particular to a VASP-based metal surface adsorption energy differential charge analysis method and system. Background Art

[0002] Differential charge analysis of metal surface adsorption energy is a method used to study the adsorption behavior of adsorbates (such as small molecules) on metal surfaces. It determines the adsorption energy by calculating the energy change in the system before and after adsorption, and analyzes the change in electron density to understand the charge transfer phenomenon that occurs during the adsorption process. Therefore, utilizing advanced technologies to improve the intelligence and safety of differential charge analysis of metal surface adsorption energy has become a pressing issue.

[0003] In the field of differential charge analysis of metal surface adsorption energy, traditional methods rely on intensive first-principles calculations (DFT), which are time-consuming and require high computing resources. There is no effective pre-screening mechanism, resulting in a large number of unnecessary calculations. In addition, the calculation parameters in first-principles software such as VASP are usually fixed or based on empirical settings, and cannot be dynamically adjusted according to specific experimental conditions. There is a lack of the ability to automatically optimize parameters under different environmental conditions, which affects the authenticity and accuracy of the simulation 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 VASP-based metal surface adsorption energy differential charge analysis method to solve the problem that the calculation parameters in first-principles software such as VASP are usually fixed or set based on experience, fail to be dynamically adjusted according to specific experimental conditions, and lack the ability to automatically optimize parameters under different environmental conditions, which affects the authenticity and accuracy of the simulation results.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a VASP-based metal surface adsorption energy differential charge analysis method, which comprises: Model and standardize the metal surface-adsorbate system to obtain a standard data set; Use a deep neural network model to train the standard data set to obtain a prediction model, input the standard data set into the prediction model, and output the prediction results; The environmental effect correction method is used to dynamically adjust the VASP calculation parameters and generate VASP input files based on the prediction results and actual operating conditions; The first-principles calculation tool VASP is used to perform high-precision energy and electron density calculations on the input file to obtain the electron density distribution; Perform differential operation on the electron density distribution to obtain the differential charge density map; Use visualization analysis tools to perform three-dimensional rendering of the differential charge density map to obtain visualization analysis results; Integrate adsorption energy predictions, VASP calculation results, differential charge density maps, and visualization analysis results to form a complete analysis report.

[0007] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the metal surface-adsorbate system is modeled and standardized to obtain a standard data set, specifically comprising the following steps: The metal surface structure information in the materials science database is retrieved, and the ASE modeling tool is used to perform surface cutting and vacuum layer construction on the original crystal structure to obtain the initial surface model; Active site analysis is performed on the initial surface model to obtain the adsorption sites, and the molecular structure of the adsorbate is modeled and placed on the adsorption sites to obtain the adsorption system structure; The adsorbents include typical small molecules of H2, CO and CH3OH; Perform local energy minimization on the adsorption system structure to obtain the optimized structural model; Perform graph structure extraction on the optimized structural model to obtain graph node and edge sets; A multi-scale graph convolution embedding function is used to extract features from the graph structure and obtain a structure representation vector; The normalization method is used to normalize the structure representation vector to obtain the standardized structure feature vector; The standardized structural feature vectors, physical parameters and environmental variables are combined to obtain a standard data set.

[0008] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the method comprises the following steps: training a standard data set with a deep neural network model to obtain a prediction model, inputting the standard data set into the prediction model, and outputting a prediction result. Divide the standard data set into training set, validation set and test set; Use the graph convolutional network architecture to model the structural feature vectors in the training set and initialize the graph neural network model GCN; The graph neural network model GCN is trained on the training set using a loss function optimization algorithm. The model parameters are adjusted by minimizing the loss function to obtain a preliminary training model. Among them, the mean square error MSE is defined as the loss function; Further optimize the preliminary training model to avoid overfitting and obtain the optimized prediction model for actual prediction; Input the test set into the optimized prediction model to generate prediction results; Calculate prediction errors, including root mean square error and coefficient of determination; The new standard data set is input into the optimized prediction model to obtain new adsorption energy prediction results.

[0009] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the VASP calculation parameters are dynamically adjusted using the environmental effect correction method, and a VASP input file is generated based on the prediction results and actual operating conditions. The specific steps are as follows: Measure or query actual operating conditions to obtain specific environmental parameter values; The actual operating conditions include temperature and pressure; Define a correction function based on the ideal gas law and thermodynamic relations to adjust VASP input parameters; Adjust the key parameters in the VASP input file according to the obtained correction coefficient to obtain the preliminarily adjusted VASP input file The INCAR part of the adjustment VASP input file includes a cutoff energy adjustment formula and a k-point grid density adjustment formula; The adsorption energy output by the prediction model was compared with the VASP calculation results after preliminary adjustment; Setting thresholds , when the difference between the two exceeds the threshold , then adjust the cutoff energy and k-point grid density proportionally according to the error size; Update the relevant parameters in the INCAR file and regenerate the VASP input file to obtain the adjusted VASP input file.

[0010] As a preferred solution of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, wherein: the first-principles calculation tool VASP is used to perform high-precision energy and electron density calculations on the input file to obtain the electron density distribution, and the specific steps are as follows: The geometric optimization algorithm is used to minimize the energy of the initial structural model; Perform geometric optimization calculations and use the quasi-Newton algorithm for structural optimization; Set the maximum force convergence criterion to During the calculation process, the program will automatically adjust the atomic positions to find the lowest energy configuration, and after completion, it will output the optimized structural model and the corresponding total energy. ; Based on the optimized structural model, perform self-consistent field (SCF) calculations; The SCF process obtains the electron density of the system by iteratively solving the Kohn-Sham equation ; After the self-consistent field calculation is completed, the existing wave function is used to perform the non-self-consistent field NSCF calculation; Set the energy grid and k-point path to calculate the state density; For the projected density of states PDOS, the contribution of specific atoms or orbitals is considered; The electron density distribution calculated above is saved as a standardized format file.

[0011] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the differential operation of the electron density distribution is performed to obtain a differential charge density map, and the specific steps are as follows: Self-consistent field calculations are performed on the substrate and adsorbate separately to obtain their respective electron density distributions; Construct composite architectures containing substrates and adsorbates and perform self-consistent field calculations; Output the total electron density distribution of the adsorption system; The differential charge density formula is used to perform differential operation on the obtained electron density distribution to obtain the differential charge density ; Based on the original calculation, the spatial resolution can be improved by increasing the number of grid points or reducing the grid spacing; Set a new number of grid points , so that the number of grid points in each direction is doubled; The differential charge density data are resampled using linear interpolation to generate a high-resolution differential charge density map; Differential charge density Z-score normalization was performed to obtain a normalized differential charge density map.

[0012] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the following steps are used to perform three-dimensional rendering of the differential charge density map using a visualization analysis tool to obtain a visualization analysis result: Use a text editor to process the CHGCAR file generated by VASP calculation and extract the required differential charge density data from it; Use the VESTA tool to convert the extracted data into a CUBE format file. In the VESTA visualization software, adjust the Isosurface parameters, set an appropriate charge density threshold, and display the charge accumulation and depletion regions; By modifying the color mapping scheme, the differential charge density of different intensities is presented in different colors; Apply the clipping plane function to cut specific parts of the model to observe the charge distribution in the internal structure; Use the animation function to record rotation, zoom, and translation actions to create dynamic demonstration videos to help you more intuitively understand the differential charge distribution and its changing trends; Export the final visualization as a high-quality image or video file.

[0013] As a preferred embodiment of the VASP-based metal surface adsorption energy differential charge analysis method of the present invention, the steps of integrating the adsorption energy prediction value, VASP calculation results, differential charge density map and visual analysis results to form a complete analysis report are as follows: Extract the adsorption energy prediction values ​​of all relevant adsorption systems from the deep learning model, classify and organize the prediction values ​​according to the adsorbate type, substrate material and experimental conditions, and form an adsorption energy prediction list; According to the VASP calculation output file, the total energy, adsorption energy and other key physical quantities of each adsorption system are extracted; For each adsorption system, its detailed electronic structure information is recorded, including density of states and projected density of states; Integrate the above information to form a VASP calculation result set ; Use VESTA or similar tools to load the CUBE format differential charge density file; Adjust isosurface parameters, color mapping schemes, and transparency settings to optimize visual effects; Apply the clipping plane function to cut the model and highlight the area of ​​interest; Export the final high-quality 3D rendered image ; In VESTA or other visualization software that supports animation, set the rotation angle, zoom ratio and translation path; Record a series of continuous actions to create dynamic demonstration videos; Generate a separate dynamic demonstration video for each adsorption system ; Compare the adsorption energy prediction value with the adsorption energy calculated by VASP and calculate the error between the two ; Analyzing differential charge density maps Comparison with VASP calculation results Electronic structure information in; Integrate all comparative analysis results to form a consistency test report; Use LaTeX, Word or other document editing software to create new document templates; Insert the adsorption energy prediction list and VASP calculation result set into the document in sequence , high-quality 3D rendered images , Visual Analysis Video Collection And the consistency check report, organize the content in a logical order to get a complete analysis report.

[0014] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the VASP-based metal surface adsorption energy differential charge analysis method as described in the first aspect of the present invention is implemented.

[0015] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the VASP-based metal surface adsorption energy differential charge analysis method as described in the first aspect of the present invention is implemented.

[0016] The beneficial effects of the present invention are as follows: by adopting a deep neural network model to train a standard data set, a prediction model is obtained, which realizes rapid and accurate prediction of adsorption energy, significantly shortens the time required for traditional DFT calculations, and improves research efficiency; by adopting an environmental effect correction method to dynamically adjust the VASP calculation parameters and generate a VASP input file, the consistency of simulation conditions and actual operations is achieved, the consistency of simulation results and experimental conditions is improved, the credibility is enhanced, and a seamless connection from theoretical modeling to real scenarios is achieved; by performing differential operations on the electron density distribution, a differential charge density map is obtained, which realizes the visualization and quantitative analysis of the charge transfer behavior during the adsorption process, and can be used to identify microscopic mechanisms such as the position of chemical bond formation and the direction of charge polarization. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 Flowchart of the VASP-based metal surface adsorption energy differential charge analysis method in Example 1. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0020] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0022] Example 1, with reference to Figure 1 , which is the first embodiment of the present invention, provides a VASP-based metal surface adsorption energy differential charge analysis method, comprising the following steps: S1. Model and standardize the metal surface-adsorbate system to obtain a standard data set; Furthermore, the metal surface structure information in the materials science database was retrieved, and the ASE modeling tool was used to perform surface cutting and vacuum layer construction on the original crystal structure to obtain the initial surface model; Active site analysis is performed on the initial surface model to obtain the adsorption sites, and the molecular structure of the adsorbate is modeled and placed on the adsorption sites to obtain the adsorption system structure; Adsorbates include typical small molecules such as H2, CO and CH3OH; Perform local energy minimization on the adsorption system structure to obtain the optimized structural model; Perform graph structure extraction on the optimized structural model to obtain graph node and edge sets; The multi-scale graph convolution embedding function is used to extract the features of the graph structure and obtain the structure representation vector, which is expressed as: ; in, is the number of graph convolution layers, is the weighting coefficient, For the Number of layer nodes; The normalization method is used to normalize the structure representation vector to obtain the standardized structure feature vector; The standardized structural feature vectors, physical parameters and environmental variables are combined to obtain a standard data set; It should be noted that the accuracy and applicability of the model are ensured by searching the metal surface structure information in the materials science database and using the ASE modeling tool to perform surface cutting and vacuum layer construction on the original crystal structure. It not only provides an accurate initial geometric configuration for subsequent calculations, but also determines the adsorption position through active site analysis, thereby improving the reliability of adsorption energy prediction. In addition, through steps such as local energy minimization, graph structure extraction and feature embedding, an effective conversion from complex three-dimensional structure to low-dimensional vector representation is achieved, which facilitates the understanding and application of machine learning models.

[0023] S2. Use a deep neural network model to train the standard data set to obtain a prediction model, input the standard data set into the prediction model, and output the prediction results; Furthermore, the standard dataset is divided into training set, validation set and test set; The graph convolutional network architecture is used to model the structural feature vectors in the training set and initialize the graph neural network model GCN, where each layer is defined as follows: ; in, is the normalized adjacency matrix, It is The node feature matrix of the layer, It is The weight matrix of the layer, Represents the ReLU activation function; The graph neural network model GCN is trained on the training set using a loss function optimization algorithm. The model parameters are adjusted by minimizing the loss function to obtain a preliminary training model. Among them, the mean square error MSE is defined as the loss function, and the expression is: ; in, is the true value, is the model prediction value, is the sample size; Further optimize the preliminary training model to avoid overfitting and obtain the optimized prediction model for actual prediction; Input the test set into the optimized prediction model to generate prediction results; Calculate the prediction error, including the root mean square error and the coefficient of determination, as follows:

[0024] ; in, is the number of test set samples, and are the predicted value and the true value respectively, is the mean of the true values; The new standard data set is input into the optimized prediction model to obtain new adsorption energy prediction results; It should be noted that dividing the standard dataset into training set, validation set and test set, and using the graph convolutional network architecture to model the structural feature vectors in the training set can effectively capture the topological relationship between atoms and improve the generalization ability of the model. By defining the mean square error (MSE) as the loss function and combining the early stopping method to optimize the model parameters, the occurrence of overfitting is avoided, making the prediction model more stable and reliable in practical applications. The final prediction results can not only be used to guide VASP parameter settings, but also provide fast and reliable initial values ​​for high-throughput screening.

[0025] S3. Dynamically adjust VASP calculation parameters using environmental effect correction methods and generate VASP input files based on prediction results and actual operating conditions; Furthermore, actual operating conditions are measured or queried to obtain specific environmental parameter values; Actual operating conditions include temperature and pressure; Define a correction function based on the ideal gas law and thermodynamic relations to adjust the VASP input parameters. The expression is: ; in, is the temperature influence coefficient, is the temperature attenuation coefficient, is the pressure influence coefficient, is the reference temperature, is standard atmospheric pressure; Adjust the key parameters in the VASP input file according to the obtained correction coefficient to obtain the preliminarily adjusted VASP input file Adjusting the INCAR part of the VASP input file includes the cutoff energy adjustment formula and the k-point grid density adjustment formula, which are expressed as follows: ; ; in, is the original cutoff energy, is the adjusted cutoff energy, It's original Point grid density, It is adjusted Point grid density; The adsorption energy output by the prediction model was compared with the VASP calculation results after preliminary adjustment; Setting thresholds , when the difference between the two exceeds the threshold , then the cutoff energy and k-point grid density are adjusted proportionally according to the error size, and the expression is: ; ; in, is the error between the predicted value and the calculated value, , are the adjustment coefficients for the cutoff energy and k-point grid density, respectively; Update the relevant parameters in the INCAR file and regenerate the VASP input file to obtain the adjusted VASP input file; It should be noted that by measuring or querying the actual operating conditions and designing correction functions based on the ideal gas law and thermodynamic relationships to adjust the VASP input parameters, the consistency of the simulation conditions and the experimental environment is ensured. This method not only solves the error problem caused by traditional fixed parameters, but also further fine-tunes key parameters through predictive model output feedback, thereby improving the calculation accuracy. The final generated VASP input file can more realistically reflect the physical and chemical behavior under actual operating conditions, thereby enhancing the credibility of the simulation results.

[0026] S4. Use the first-principles calculation tool VASP to perform high-precision energy and electron density calculations on the input file to obtain the electron density distribution; Furthermore, the geometric optimization algorithm is used to minimize the energy of the initial structural model; Perform geometric optimization calculations and use the quasi-Newton algorithm for structural optimization; Set the maximum force convergence criterion to During the calculation process, the program will automatically adjust the atomic positions to find the lowest energy configuration, and after completion, it will output the optimized structural model and the corresponding total energy. ; Based on the optimized structural model, perform self-consistent field (SCF) calculations; The SCF process obtains the electron density of the system by iteratively solving the Kohn-Sham equation , the expression is: ; in, is the reduced Planck constant, is the mass of the electron, is the effective potential, including ionic potential, Hartree potential and exchange correlation potential, is the wave function, is the intrinsic energy; After the self-consistent field calculation is completed, the existing wave function is used to perform the non-self-consistent field NSCF calculation; Set the energy grid and k-point path, calculate the state density, the expression is: ; in, represents the band index, represents the k-point index, corresponds to the energy band and k points The intrinsic energy of For the projected density of states PDOS, considering the contribution of specific atoms or orbitals, the expression is: ; in, represents a specific atom or orbital, is the basis function of the atom or orbital, is the wave function; Save the electron density distribution calculated above as a standardized format file; It should be noted that by minimizing the energy of the initial structural model through the geometric optimization algorithm and performing self-consistent field SCF calculations based on the optimized structural model, the accurate electron density distribution of the system can be obtained. In this process, the iterative solution of the Kohn-Sham equation ensures the accuracy of the electronic structure description, while the non-self-consistent field NSCF calculation further provides important information such as the density of states DOS and the projected density of states PDOS, which helps to deeply understand the behavior of electrons. The standardized format file finally saved provides a high-quality data basis for the subsequent differential charge density analysis.

[0027] S5. performing a differential operation on the electron density distribution to obtain a differential charge density map; Furthermore, self-consistent field calculations are performed on the substrate and adsorbate separately to obtain their respective electron density distributions; Construct a composite architecture consisting of the substrate and the adsorbate and perform a self-consistent field calculation, expressed as: ; in, is the effective potential of the composite system, is the wave function of the composite system, is the corresponding eigenenergy; Output the total electron density distribution of the adsorption system; The differential charge density formula is used to perform differential operation on the obtained electron density distribution to obtain the differential charge density , the expression is: ; in, is the total electron density of the composite system, is the electron density of the substrate, is the electron density of the adsorbate; Based on the original calculation, the spatial resolution can be improved by increasing the number of grid points or reducing the grid spacing; Set a new number of grid points , so that the number of grid points in each direction is doubled, the expression is: ; in, Indicates one of the three directions x, y, and z. is the original grid point number; The differential charge density data are resampled using linear interpolation to generate a high-resolution differential charge density map; Differential charge density Perform Z-score normalization to obtain a normalized differential charge density map; It should be noted that by performing self-consistent field calculations on the substrate and adsorbate separately and processing the results using the differential charge density formula, the specific circumstances of charge transfer during the adsorption process can be intuitively displayed. By increasing the number of grid points or reducing the grid point spacing to improve spatial resolution, and using Z-score normalization processing, image contrast and clarity are ensured, making it easier to identify charge accumulation and depletion areas. The quantitative analysis method not only reveals the essence of the adsorption mechanism, but also provides an important theoretical basis for catalyst design.

[0028] S6. Use a visualization analysis tool to perform three-dimensional rendering on the differential charge density map to obtain a visualization analysis result; Furthermore, a text editor was used to process the CHGCAR file generated by the VASP calculation to extract the required differential charge density data; Use the VESTA tool to convert the extracted data into a CUBE format file. In the VESTA visualization software, adjust the Isosurface parameters, set an appropriate charge density threshold, and display the charge accumulation and depletion regions; By modifying the color mapping scheme, the differential charge density of different intensities is presented in different colors; Apply the clipping plane function to cut specific parts of the model to observe the charge distribution in the internal structure; Use the animation function to record rotation, zoom, and translation actions to create dynamic demonstration videos to help you more intuitively understand the differential charge distribution and its changing trends; Export the final visualization image or video as a high-quality image or video file; It should be noted that the three-dimensional rendering of differential charge density data through VESTA visualization software can not only intuitively display the spatial characteristics of the charge distribution, but also enhance understanding and interpretation through color mapping schemes and clipping plane functions. The animation function records continuous actions and creates dynamic demonstration videos to help scientific researchers more comprehensively understand the differential charge distribution and its changing trends. High-quality visualization results are not only an important tool for scientific research, but also an effective means of presenting research results to the outside world.

[0029] S7. Integrate the adsorption energy prediction value, VASP calculation results, differential charge density map and visualization analysis results to form a complete analysis report; Furthermore, the adsorption energy prediction values ​​of all relevant adsorption systems are extracted from the deep learning model, and the prediction values ​​are classified and sorted according to the adsorbate type, substrate material and experimental conditions to form an adsorption energy prediction list; According to the VASP calculation output file, the total energy, adsorption energy and other key physical quantities of each adsorption system are extracted; For each adsorption system, its detailed electronic structure information is recorded, including density of states and projected density of states; Integrate the above information to form a VASP calculation result set ; Use VESTA or similar tools to load the CUBE format differential charge density file; Adjust isosurface parameters, color mapping schemes, and transparency settings to optimize visual effects; Apply the clipping plane function to cut the model and highlight the area of ​​interest; Export the final high-quality 3D rendered image ; In VESTA or other visualization software that supports animation, set the rotation angle, zoom ratio and translation path; Record a series of continuous actions to create dynamic demonstration videos; Generate a separate dynamic demonstration video for each adsorption system ; Compare the adsorption energy prediction value with the adsorption energy calculated by VASP and calculate the error between the two , the expression is: ; in, is the adsorption energy calculated by VASP, is the predicted value of the deep learning model; Analyzing differential charge density maps Comparison with VASP calculation results Electronic structure information in; Integrate all comparative analysis results to form a consistency test report; Use LaTeX, Word or other document editing software to create new document templates; Insert the adsorption energy prediction list and VASP calculation result set into the document in sequence , high-quality 3D rendered images , Visual Analysis Video Collection and consistency check report, organize the content in a logical order to obtain a complete analysis report; It should be noted that by integrating the adsorption energy prediction values, VASP calculation results, differential charge density maps and visual analysis results, a complete analysis report with clear logic and detailed content is formed. This not only provides comprehensive technical support for researchers, but also promotes interdisciplinary knowledge sharing and technical exchanges. By displaying the research results of each link, the report helps readers better understand the design concept and implementation details of the entire process, and at the same time provides reference suggestions for further research directions.

[0030] This embodiment also provides a computer device, which is suitable for the VASP-based metal surface adsorption energy differential charge analysis method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the VASP-based metal surface adsorption energy differential charge analysis method proposed in the above embodiment.

[0031] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0032] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the VASP-based metal surface adsorption energy differential charge analysis method 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 read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0033] In summary, the present invention uses a deep neural network model to train a standard data set to obtain a prediction model, thereby achieving rapid and accurate prediction of adsorption energy, significantly shortening the time required for traditional DFT calculations, and improving research efficiency. By dynamically adjusting the VASP calculation parameters using an environmental effect correction method and generating a VASP input file, the simulation conditions are matched consistently with actual operations, the consistency of the simulation results with the experimental conditions is improved, the credibility is enhanced, and a seamless connection from theoretical modeling to real-world scenarios is achieved. By performing differential operations on the electron density distribution, a differential charge density map is obtained, which realizes the visualization and quantitative analysis of the charge transfer behavior during the adsorption process, and can be used to identify microscopic mechanisms such as the position of chemical bond formation and the direction of charge polarization.

[0034] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A VASP-based metal surface adsorption energy differential charge analysis method, characterized by: include: Model and standardize the metal surface-adsorbate system to obtain a standard data set; Use a deep neural network model to train the standard data set to obtain a prediction model, input the standard data set into the prediction model, and output the prediction results; The environmental effect correction method is used to dynamically adjust the VASP calculation parameters and generate VASP input files based on the prediction results and actual operating conditions; The first-principles calculation tool VASP is used to perform high-precision energy and electron density calculations on the input file to obtain the electron density distribution; Perform differential operation on the electron density distribution to obtain the differential charge density map; Use visualization analysis tools to perform three-dimensional rendering of the differential charge density map to obtain visualization analysis results; Integrate adsorption energy predictions, VASP calculation results, differential charge density maps, and visualization analysis results to form a complete analysis report.

2. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 1, characterized in that: The metal surface-adsorbate system is modeled and standardized to obtain a standard data set. The specific steps are: The metal surface structure information in the materials science database is retrieved, and the ASE modeling tool is used to perform surface cutting and vacuum layer construction on the original crystal structure to obtain the initial surface model; Active site analysis is performed on the initial surface model to obtain the adsorption sites, and the molecular structure of the adsorbate is modeled and placed on the adsorption sites to obtain the adsorption system structure; The adsorbents include typical small molecules of H2, CO and CH3OH; Perform local energy minimization on the adsorption system structure to obtain the optimized structural model; Perform graph structure extraction on the optimized structural model to obtain graph node and edge sets; A multi-scale graph convolution embedding function is used to extract features from the graph structure and obtain a structure representation vector; The normalization method is used to normalize the structure representation vector to obtain the standardized structure feature vector; The standardized structural feature vectors, physical parameters and environmental variables are combined to obtain a standard data set.

3. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 2, characterized in that: The deep neural network model is used to train the standard data set to obtain a prediction model, and the standard data set is input into the prediction model to output the prediction results. The specific steps are as follows: Divide the standard data set into training set, validation set and test set; Use the graph convolutional network architecture to model the structural feature vectors in the training set and initialize the graph neural network model GCN; The graph neural network model GCN is trained on the training set using a loss function optimization algorithm. The model parameters are adjusted by minimizing the loss function to obtain a preliminary training model. Among them, the mean square error MSE is defined as the loss function; Further optimize the preliminary training model to avoid overfitting and obtain the optimized prediction model for actual prediction; Input the test set into the optimized prediction model to generate prediction results; Calculate prediction errors, including root mean square error and coefficient of determination; The new standard data set is input into the optimized prediction model to obtain new adsorption energy prediction results.

4. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 3, characterized in that: The environmental effect correction method is used to dynamically adjust the VASP calculation parameters and generate a VASP input file based on the prediction results and actual operating conditions. The specific steps are as follows: Measure or query actual operating conditions to obtain specific environmental parameter values; The actual operating conditions include temperature and pressure; Define a correction function based on the ideal gas law and thermodynamic relations to adjust VASP input parameters; Adjust the key parameters in the VASP input file according to the obtained correction coefficient to obtain the preliminarily adjusted VASP input file The INCAR part of the adjustment VASP input file includes a cutoff energy adjustment formula and a k-point grid density adjustment formula; The adsorption energy output by the prediction model was compared with the VASP calculation results after preliminary adjustment; Setting thresholds , when the difference between the two exceeds the threshold , then adjust the cutoff energy and k-point grid density proportionally according to the error size; Update the relevant parameters in the INCAR file and regenerate the VASP input file to obtain the adjusted VASP input file.

5. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 4, characterized in that: The first-principles calculation tool VASP is used to perform high-precision energy and electron density calculations on the input file to obtain the electron density distribution. The specific steps are as follows: The geometric optimization algorithm is used to minimize the energy of the initial structural model; Perform geometric optimization calculations and use the quasi-Newton algorithm for structural optimization; Set the maximum force convergence criterion to During the calculation process, the program will automatically adjust the atomic positions to find the lowest energy configuration, and after completion, it will output the optimized structural model and the corresponding total energy. ; Based on the optimized structural model, perform self-consistent field (SCF) calculations; The SCF process obtains the electron density of the system by iteratively solving the Kohn-Sham equation ; After the self-consistent field calculation is completed, the existing wave function is used to perform the non-self-consistent field NSCF calculation; Set the energy grid and k-point path to calculate the state density; For the projected density of states PDOS, the contribution of specific atoms or orbitals is considered; The electron density distribution calculated above is saved as a standardized format file.

6. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 5, characterized in that: The differential operation of the electron density distribution is performed to obtain a differential charge density map, and the specific steps are as follows: Self-consistent field calculations are performed on the substrate and adsorbate separately to obtain their respective electron density distributions; Construct composite architectures containing substrates and adsorbates and perform self-consistent field calculations; Output the total electron density distribution of the adsorption system; The differential charge density formula is used to perform differential operation on the obtained electron density distribution to obtain the differential charge density ; Based on the original calculation, the spatial resolution can be improved by increasing the number of grid points or reducing the grid spacing; Set a new number of grid points , so that the number of grid points in each direction is doubled; The differential charge density data are resampled using linear interpolation to generate a high-resolution differential charge density map; Differential charge density Z-score normalization was performed to obtain a normalized differential charge density map.

7. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 6, characterized in that: The visualization analysis tool is used to perform three-dimensional rendering on the differential charge density map to obtain the visualization analysis results. The specific steps are as follows: Use a text editor to process the CHGCAR file generated by VASP calculation and extract the required differential charge density data from it; Use the VESTA tool to convert the extracted data into a CUBE format file. In the VESTA visualization software, adjust the Isosurface parameters, set an appropriate charge density threshold, and display the charge accumulation and depletion regions; By modifying the color mapping scheme, the differential charge density of different intensities is presented in different colors; Apply the clipping plane function to cut specific parts of the model to observe the charge distribution in the internal structure; Use the animation function to record rotation, zoom, and translation actions to create dynamic demonstration videos to help you more intuitively understand the differential charge distribution and its changing trends; Export the final visualization as a high-quality image or video file.

8. The VASP-based metal surface adsorption energy differential charge analysis method according to claim 7, characterized in that: The integration of adsorption energy prediction values, VASP calculation results, differential charge density maps and visual analysis results to form a complete analysis report is as follows: Extract the adsorption energy prediction values ​​of all relevant adsorption systems from the deep learning model, classify and organize the prediction values ​​according to the adsorbate type, substrate material and experimental conditions, and form an adsorption energy prediction list; According to the VASP calculation output file, the total energy, adsorption energy and other key physical quantities of each adsorption system are extracted; For each adsorption system, its detailed electronic structure information is recorded, including density of states and projected density of states; Integrate the above information to form a VASP calculation result set ; Use VESTA or similar tools to load the CUBE format differential charge density file; Adjust isosurface parameters, color mapping schemes, and transparency settings to optimize visual effects; Apply the clipping plane function to cut the model and highlight the area of ​​interest; Export the final high-quality 3D rendered image ; In VESTA or other visualization software that supports animation, set the rotation angle, zoom ratio and translation path; Record a series of continuous actions to create dynamic demonstration videos; Generate a separate dynamic demonstration video for each adsorption system ; Compare the adsorption energy prediction value with the adsorption energy calculated by VASP and calculate the error between the two ; Analyzing differential charge density maps Comparison with VASP calculation results Electronic structure information in; Integrate all comparative analysis results to form a consistency test report; Use LaTeX, Word or other document editing software to create new document templates; Insert the adsorption energy prediction list and VASP calculation result set into the document in sequence , high-quality 3D rendered images , Visual Analysis Video Collection And the consistency check report, organize the content in a logical order to get a complete analysis report.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the VASP-based metal surface adsorption energy differential charge analysis method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the VASP-based metal surface adsorption energy differential charge analysis method according to any one of claims 1 to 8 are implemented.

Images