Data-driven low-dimensional energy material high-throughput screening method, device and storage medium

Abstract

The present invention discloses a data-driven high-throughput screening method for low-dimensional energy materials, a computer device, and a storage medium, including the steps of obtaining substrate structure information, constructing a two-dimensional electrocatalyst structure model, simulating the electrochemical reaction mechanism based on the substrate structure information and the two-dimensional electrocatalyst structure model, calculating the overpotential based on the electrochemical reaction mechanism, using a machine learning model as a judgment model, and predicting the electrocatalyst based on the two-dimensional electrocatalyst structure model and the overpotential. The present invention provides the user with the most appropriate material model through the experimental sample characteristics provided by the user or the model characteristics adopted according to previous literature, and directly provides the models of various intermediate products and the final electrochemical evaluation index values. The user can directly obtain the relevant calculation model and the functions of submitting calculation tasks, processing data, etc. without first-principles calculations or supercomputing experience. The present invention is widely used in the field of materials technology.

Description

Technical Field

[0001] The present invention relates to the field of material technology, and in particular to a data-driven high-throughput screening method for low-dimensional energy materials, a computer device, and a storage medium. Background Art

[0002] The design and development of energy materials is one of the core tasks in solving energy and environmental problems. Energy materials include energy storage materials and energy conversion materials, represented by secondary metal ion battery electrode materials and electrocatalyst materials respectively. For secondary batteries, the properties of efficient electrode materials include but are not limited to electrical conductivity, thermal conductivity, and excellent ion adsorption, embedding, extraction and migration properties. For electrocatalytic materials, the electrochemical reaction mechanism at the solid-liquid interface is one of the core issues to be solved. These reactions include nitrogen reduction (NRR), carbon dioxide reduction (CO2RR), water electrolysis (HER and OER), oxygen reduction reaction (ORR) and sulfur redox reaction (SRR / SER). The efficiency of these electrochemical reactions depends largely on the performance of the catalyst material.

[0003] In order to improve material performance, in addition to the design and development of new materials and new structures, the modification and design of existing materials is also an important means. Taking the design of electrocatalyst materials as an example, the core task is to improve their activity, selectivity, stability and cost-effectiveness. Traditional catalyst materials are mainly based on precious metals. Although these catalysts have excellent performance, they are expensive and have limited reserves. Currently, there are a lot of research on their alternative materials. These alternative materials include catalyst materials containing non-precious metals as well as catalyst materials that are completely metal-free. There are both three-dimensional bulk materials and two-dimensional, one-dimensional and zero-dimensional materials. Intrinsic materials generally do not have excellent catalytic properties, so defect engineering or doping strategies are effective means to design new and efficient catalysts. However, there are many types of catalysts prepared according to this scheme. For single-atom and diatomic catalysts supported on carbon-based two-dimensional materials alone, the possible materials may reach tens of thousands. According to current reports, there are dozens of metals that can be screened (mainly transition metals). Considering that the coordination configuration may vary depending on the type and possible position of the heterogeneous atoms (C, O, N, P, B, S, Cl, F, etc.), and considering a variety of intrinsic defects and modifications of oxygen-containing functional groups, the possible configurations of single-atom catalysts will be even more.

[0004] Materials science has entered a new era driven by big data. Data generation, database establishment, data mining, and machine learning have become highly effective means of designing novel functional materials. Fundamentally, materials design relies on an understanding of their structure. Through experimental work, researchers have acquired a wealth of structural information and established corresponding databases. While some high-throughput computational materials design software packages exist, no algorithms have yet been developed for the design of low-dimensional secondary battery electrode materials and electrocatalysts.

[0005] Explanation of terms:

[0006] Energy materials: Materials required for energy storage and conversion devices. Common energy storage and conversion devices include lithium-ion secondary batteries and metal-air batteries.

[0007] First principles calculation: refers to quantum calculation that does not use empirical parameters and only requires a small amount of basic experimental data.

[0008] High-throughput computing: A scientific research method that uses a small amount of computing resources to quickly and accurately calculate the various properties of material systems in large quantities, thereby exploring and predicting the properties of materials.

[0009] Low-dimensional materials: These refer to two-dimensional, one-dimensional, or zero-dimensional materials, as opposed to three-dimensional bulk materials. Low-dimensional materials lack spatial periodicity in one, two, or even three dimensions. Examples include two-dimensional materials like graphene, one-dimensional materials like carbon nanotubes, and zero-dimensional materials like fullerenes.

[0010] Single-atom catalysts: Single metal atoms dispersed and riveted on special substrate materials, possessing the efficient catalytic properties of metal surfaces or clusters. Summary of the Invention

[0011] In response to the technical problems that there is currently no algorithm for the design of low-dimensional secondary battery electrode materials and electrocatalyst materials, the purpose of the present invention is to provide a data-driven high-throughput screening method, computer device and storage medium for low-dimensional energy materials.

[0012] In one aspect, an embodiment of the present invention includes a data-driven high-throughput screening method for low-dimensional energy materials, comprising:

[0013] Obtaining substrate structure information;

[0014] Construct a two-dimensional electrocatalyst structure model;

[0015] simulating an electrochemical reaction mechanism based on the substrate structure information and the two-dimensional electrocatalyst structure model;

[0016] Calculate the overpotential based on the electrochemical reaction mechanism;

[0017] A machine learning model is used as a judgment model to perform electrocatalyst prediction based on the two-dimensional electrocatalyst structure model and the overpotential.

[0018] Furthermore, the obtaining of substrate structure information includes:

[0019] Establish user interface;

[0020] The substrate structure information is read through the user operation interface.

[0021] Furthermore, the reading of the substrate structure information includes:

[0022] Extract substrate structure information from the Materials Project database.

[0023] Furthermore, the constructing of the two-dimensional electrocatalyst structure model comprises:

[0024] Reading the substrate structure information through a Python program;

[0025] generating several types of defect information based on the substrate structure information;

[0026] Heterogeneous atom information is introduced into the substrate structure information to construct the metal atom coordination environment information and obtain the two-dimensional electrocatalyst structure model.

[0027] Furthermore, the constructing of the two-dimensional electrocatalyst structure model comprises:

[0028] Obtaining a lattice vector of a unit cell of a heterojunction upper structure and a lattice vector of a unit cell of a heterojunction lower structure;

[0029] Perform several iterative processes; in any iterative process, within the lattice period of this iterative process, all construction schemes between the heterojunction upper structure and the heterojunction lower structure are traversed and calculated according to the lattice vector of the unit cell of the heterojunction upper structure and the lattice vector of the unit cell of the heterojunction lower structure, and the mismatch rate and mismatch angle corresponding to each construction scheme are calculated; when a feasible construction scheme is screened out in this iterative process, this iterative process is used as the last iterative process, and otherwise the next iterative process is performed; wherein the mismatch rate corresponding to the feasible construction scheme is less than a preset mismatch rate threshold, and the mismatch angle corresponding to the feasible construction scheme is less than a preset mismatch angle threshold; the lattice period of the first iterative process is the initial lattice period, and the lattice period of other iterative processes is obtained by expanding the lattice period of the previous iterative process;

[0030] Among all the feasible construction schemes obtained in the last iterative process, the feasible construction scheme with the smallest sum of the number of atoms of the heterojunction upper structure and the heterojunction lower structure is selected to generate the two-dimensional electrocatalyst structure model.

[0031] Furthermore, simulating the electrochemical reaction mechanism according to the substrate structure information and the two-dimensional electrocatalyst structure model includes:

[0032] Select catalyst model and reaction type;

[0033] Determine the adsorption site based on the structural characteristics of the substrate;

[0034] Implement the adsorption model of intermediate products.

[0035] Furthermore, the calculating of the overpotential according to the electrochemical reaction mechanism includes:

[0036] For the nitrogen reduction reaction, perform the following steps:

[0037] Preliminary screening of catalyst performance based on N2 molecule adsorption energy;

[0038] The catalytic performance was evaluated based on the free energy changes of the 1st, 2nd, and 6th hydrogenation reactions of adsorbed N2 molecules.

[0039] If the free energy change of hydrogenation in the above three steps is less than 0.50 eV, the performance is judged to be excellent, and all hydrogenation steps are supplemented;

[0040] For the oxygen reduction reaction, the following steps are performed:

[0041] Select catalyst model and reaction type;

[0042] Calculate the overpotential based on the adsorption free energy of the intermediate products OOH, O, and OH;

[0043] The adsorption free energy of the intermediate product OH is correlated with the overpotential to generate a two-dimensional volcano plot;

[0044] For sulfur oxidation / reduction reactions, the following steps are performed:

[0045] Select catalyst model and reaction type;

[0046] According to the free energy change of the reduction process, a lithium-sulfur long chain adsorption structure model was constructed;

[0047] Determine the maximum energy change;

[0048] Calculate the dissociation energy barrier of Li2S clusters.

[0049] Furthermore, the machine learning model is a regression function model, a decision tree model, a random forest model or a neural network model.

[0050] On the other hand, an embodiment of the present invention also includes a computer device comprising a memory and a processor, wherein the memory is used to store at least one program, and the processor is used to load the at least one program to execute the data-driven low-dimensional energy material high-throughput screening method in the embodiment.

[0051] On the other hand, an embodiment of the present invention also includes a storage medium storing a processor-executable program, which, when executed by the processor, is used to execute the data-driven low-dimensional energy material high-throughput screening method in the embodiment.

[0052] The beneficial effects of the present invention are as follows: the data-driven high-throughput screening method for low-dimensional energy materials in the embodiment provides the user with the most appropriate material model through the experimental sample characteristics provided by the user or the model characteristics adopted according to previous literature, and directly provides the models of various intermediate products and the final electrochemical evaluation index values. The user does not need first-principles calculations or supercomputing experience, and can directly obtain relevant calculation models and submit calculation tasks, process data, and other functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the steps of the data-driven high-throughput screening method for low-dimensional energy materials in the embodiment;

[0054] Figure 2 Schematic diagram of the process of the data-driven high-throughput screening method for low-dimensional energy materials in the embodiment;

[0055] Figure 3 Schematic diagram of the process of constructing a two-dimensional electrocatalyst structure model for a heterogeneous structure in the embodiment. DETAILED DESCRIPTION

[0056] Existing high-throughput computing software has the following main problems:

[0057] (1) The databases provided are mostly for intrinsic materials. There is no corresponding database for the design of electrochemical energy storage materials, especially for the high-throughput design of electrocatalyst materials related to doping or defect engineering. Even if there are corresponding model plug-ins for generating defect structures, their single function cannot meet the special requirements of current energy material design.

[0058] (2) It is not friendly to beginners in the field of computational materials science or users without computational simulation experience. These software require users to have a relatively good grasp of computational software such as VASP and Quantum-espresso; users are required to be able to determine which model to use and establish a model based on the morphology or performance of the experimental sample, and to be able to independently calculate evaluation indicators related to electrochemical performance.

[0059] (3) There is no “workstation-like” energy material design function that integrates task generation, submission, and data processing based on structure-activity relationships, such as being able to directly provide the free energy change curve related to the electrochemical reaction.

[0060] In view of the above defects, in this embodiment, Figure 1 As shown in FIG, the data-driven high-throughput screening method for low-dimensional energy materials includes the following steps:

[0061] S1. Obtain substrate structure information;

[0062] S2. Construct a two-dimensional electrocatalyst structure model;

[0063] S3. Simulate the electrochemical reaction mechanism based on substrate structure information and a 2D electrocatalyst structure model;

[0064] S4. Calculate the overpotential based on the electrochemical reaction mechanism;

[0065] S5. Use a machine learning model as a judgment model to predict electrocatalysts based on the two-dimensional electrocatalyst structure model and overpotential.

[0066] In this embodiment, the overall process of steps S1-S5 is as follows: Figure 2 As shown. Figure 2 When executing steps S1-S5, the user interface first selects the substrate, coordination environment, and reaction type. The user then packages the required input files for the first-principles calculation software package and sends them to the supercomputer for batch submission of calculation tasks and extraction of results. The basic physical properties of the substrate are first extracted, including information such as total energy, magnetic moment, density of states, bond length, and density of states. The adsorption free energy of the intermediate product is then calculated, yielding parameters such as overpotential that measure electrocatalytic performance. Finally, the basic physical properties of the substrate and electrochemical indicators such as overpotential are plotted with a single click.

[0067] In this embodiment, when executing step S1, that is, the step of obtaining substrate structure information, the following steps are specifically performed:

[0068] S101. Establish a user interface;

[0069] S102. Read substrate structure information through the user operation interface.

[0070] The user operation interface established by step S101 can be used by the user to operate, and the background program is controlled based on the operation interface to realize the automation of functions such as substrate structure reading, construction of electrochemical reaction related models, submission of calculations and post-processing data. Through the user operation interface, the user can not only build an energy structure model related to common electrochemistry, but also can intuitively display the relevant electrochemical properties of the material. In addition to considering the model and calculation details required for realizing electrochemical performance calculations, the premise that the user is not familiar with theoretical calculations is also considered. Through the user operation interface, there is no need or as little as possible to master the use of supercomputing skills, and it is possible to quickly make a preliminary judgment on the catalyst active sites and reaction mechanisms from the atomic scale, and even do further high-throughput screening. In the present embodiment, the user operation interface is mainly built based on the pyside2 module of the python language, and the interface is designed by the pyside2-designer program.

[0071] In this embodiment, when executing step S201 , ie, the step of reading substrate structure information, the substrate structure information may be captured from the Materials Project database through the user operation interface.

[0072] In this embodiment, step S2, that is, the step of constructing a two-dimensional electrocatalyst structure model, can realize batch construction of two-dimensional electrocatalyst structure models and establishment of a database.

[0073] In this embodiment, when executing step S2, the following steps may be specifically performed:

[0074] S201A. Read substrate structure information through Python program;

[0075] S202A generates several types of defect information for the substrate structure information;

[0076] S203A. Introduce heterogeneous atom information into the substrate structure information, construct the metal atom coordination environment information, and obtain a two-dimensional electrocatalyst structure model.

[0077] Steps S201A-S203A can construct a two-dimensional electrocatalyst structure model for two-dimensional materials and the single-atom catalysts they carry. Single-atom catalysts generally refer to metal atoms dispersedly embedded in substrate materials, which have high atomic utilization and high electrocatalytic activity. In addition to the above-mentioned graphene, common substrate materials also include two-dimensional carbon nitrogen compounds and MXene, transition metal sulfide compounds, and surfaces of different bulk materials, etc. The above-mentioned substrates can provide different coordination environments, so each metal single atom may have a large number of different coordination situations. Therefore, the basic substrate structure can be read in using a Python program first. These substrate structures can be captured from the Materials Project database, generating different types of defects and introducing heterogeneous atoms to construct the coordination environment of metal atoms. Metal atoms mainly select transition metals and some main group metal elements.

[0078] In this embodiment, when executing step S2, the following steps may be specifically performed:

[0079] S201B obtains the heterojunction upper structure unit cell lattice vector and the heterojunction lower structure unit cell lattice vector;

[0080] S202B. Execute several iterations; in any iteration, within the lattice period of this iteration, all construction schemes between the heterojunction upper structure and the heterojunction lower structure are traversed and calculated based on the lattice vectors of the unit cell of the heterojunction upper structure and the lattice vectors of the unit cell of the heterojunction lower structure, and the mismatch rate and mismatch angle corresponding to each construction scheme are calculated. When a feasible construction scheme is screened out in this iteration, this iteration is considered the last iteration, and if not, the next iteration is executed; wherein the mismatch rate corresponding to the feasible construction scheme is less than a preset mismatch rate threshold, and the mismatch angle corresponding to the feasible construction scheme is less than a preset mismatch angle threshold; the lattice period of the first iteration is the initial lattice period, and the lattice period of the other iterations is obtained by expanding the lattice period of the previous iteration;

[0081] S203B. Among all feasible construction schemes obtained in the last iterative process, select the feasible construction scheme with the smallest sum of the number of atoms in the heterojunction upper structure and the heterojunction lower structure to generate a two-dimensional electrocatalyst structure model.

[0082] Steps S201B-S203B can construct a two-dimensional electrocatalyst structure model for the heterogeneous structure. The process of steps S201B-S203B is as follows: Figure 3 shown.

[0083] Reference Figure 3In the field of electrochemical materials, heterostructures based on two-dimensional materials are widely used in the fields of secondary batteries and electrocatalysis. Common two-dimensional materials include graphene and related materials (such as graphyne, etc.), boron nitride, two-dimensional carbon and nitrogen compounds (such as BN, g-CN, C2N, C3N4 and C3N, etc.), Mxene and some other theoretically predicted two-dimensional or low-dimensional systems. In this embodiment, these materials and the heterostructures they constitute are summarized into a special database. Since the unit cell parameters and basis vector angles of these two-dimensional materials are not strictly matched, the lattice mismatch will cause the two layers of materials in the heterostructure to have tensile or compressive deformation and generate stress. The lattice mismatch rate and mismatch angle are defined as follows:

[0084]

[0085] θ=|θ upper -θ lower |

[0086] Where δ is the mismatch rate, θ is the mismatch angle, and v is the supercell lattice vector, where v upper Represents the unit cell vector of the heterojunction superstructure, v lower The unit cell vector represents the lower structure of the heterojunction; θupper and θlower are the angles between the unit cell vectors of the upper and lower structures of the heterojunction, respectively. Given the competitive relationship between the number of atoms in a supercell and stress, the minimum supercell is usually sought under the premise of setting the maximum lattice mismatch rate and the maximum mismatch angle. For a single-layer two-dimensional material, the plane basis vectors of its primitive cell are a1 and a2. Since the primitive cell is the smallest repeating unit that satisfies the crystal translation periodicity, the primitive cell can still satisfy the periodicity after being translated by an integer multiple of the lattice along the basis vector direction. That is, the translation period of the unit cell can be an integer multiple of the basis vector. Therefore, after the primitive cell is expanded by n times, the periodicity of its supercell also becomes n times that of the primitive cell. The lattice vectors of the unit cell are defined as w = n1a1 + n2a2 and v = m1a1 + m2a2. The idea of constructing a supercell heterojunction is to find all the supercell index parameters [n1, n2, m1, m2] and [u1, u2, v1, v2] of the upper and lower layer materials in several lattice periods respectively. The lattice period is initially set to cycle = 5. Each integer variable in [n1, n2, m1, m2] can be selected from a total of 11 integers in [-5, 5]. The unit cell redefinition method of the single-layer material can be initialized to 11 4 = 14641 kinds (the case of a zero vector after redefinition is not excluded), traverse the mismatch rate and mismatch angle between the upper and lower crystal supercells, set the maximum mismatch rate and maximum mismatch angle as thresholds, screen out all feasible construction schemes, and find the heterojunction scheme with the least number of atoms among these construction schemes to generate the initial structure file; if there is no construction scheme that meets the mismatch rate and mismatch angle, expand the lattice period cycle = 6, 7, 8... and find a suitable heterojunction construction scheme again.

[0087] In this embodiment, when executing step S3, that is, simulating the electrochemical reaction mechanism based on the substrate structure information and the two-dimensional electrocatalyst structure model, the following steps may be specifically performed:

[0088] S301. Select catalyst model and reaction type;

[0089] S302. Determine the adsorption site according to the structural characteristics of the substrate;

[0090] S303. Implement an adsorption model for intermediate products.

[0091] By executing steps S301-S303, we can batch-generate adsorption models for intermediate products of different reaction types and embed reaction descriptors. Specifically, steps S301-S303 simulate various electrochemical reaction mechanisms, including nitrogen reduction, oxygen reduction, carbon dioxide reduction, oxygen evolution, hydrogen evolution, and sulfur oxidation / reduction, based on different substrates and single-atom catalysts, as well as analyze the associated data.

[0092] After executing steps S301 to S303 to obtain the adsorption model, an input file may be generated according to the adsorption model and the input file may be submitted for calculation.

[0093] In this embodiment, by executing step S4, that is, calculating the overpotential based on the electrochemical reaction mechanism, the overpotential can be calculated for different reactions. This involves the adsorption structure of each intermediate product, adsorption free energy, etc. Specifically, step S4 is designed with corresponding steps for nitrogen reduction reaction, oxygen reduction reaction, and sulfur oxidation / reduction reaction:

[0094] Nitrogen reduction reaction:

[0095] (1) Preliminary screening of catalyst performance based on N2 molecular adsorption energy;

[0096] (2) Evaluate the catalytic performance based on the free energy changes of the first, second, and sixth hydrogenation reactions of adsorbed N2 molecules;

[0097] (3) If the free energy change of the above three steps of hydrogenation is less than 0.50 eV, the performance is judged to be excellent, and all hydrogenation steps are supplemented.

[0098] Oxygen reduction reaction:

[0099] (1) The user selects the catalyst model and reaction type;

[0100] (2) Calculate the adsorption free energy of the three intermediate products OOH, O, and OH and the overpotential;

[0101] (3) Correlate the adsorption free energy of the intermediate product OH with the overpotential and draw a two-dimensional volcano plot.

[0102] Sulfur oxidation / reduction reaction:

[0103] (1) The user selects the catalyst model and reaction type;

[0104] (2) Free energy change during the reduction process and construction of a lithium-sulfur long chain adsorption structure model;

[0105] (3) Determination of maximum energy change;

[0106] (4) The decomposition energy barrier of Li2S clusters.

[0107] In step S5, a machine learning model is used for catalyst prediction. This step uses algorithms such as regression functions, decision trees, random forests, and neural networks to explore the relationship between the catalyst's basic physical and chemical properties and its overpotential (an indicator for evaluating catalytic performance).

[0108] In this embodiment, when executing step S5, the following steps may be specifically performed:

[0109] S501. Use the existing data in the database and the user-generated data to train the best model;

[0110] S502. Use this model as a judgment model for this type of catalyst to predict more electrocatalysts with excellent performance.

[0111] Step S501 is the process of training a machine learning model. The trained machine learning model processes the catalyst's physical and chemical properties and overpotential data to output catalyst prediction data. This catalyst prediction data can convey information such as the catalyst's potential chemical structure, thereby helping guide the production and use of new catalysts.

[0112] In summary, the low-dimensional energy material screening technology implemented by executing steps S1-S5 provides the user with the most appropriate material model based on the experimental sample characteristics provided by the user or the model characteristics adopted based on previous literature, and directly provides the models of various intermediate products and the final electrochemical evaluation index values. The data-driven low-dimensional energy material high-throughput screening method in this embodiment has the following advantages:

[0113] (1) Users can directly obtain relevant computing models and submit computing tasks, process data, and other functions without first-principles calculations or supercomputing experience.

[0114] (2) Provide a large number of defect structure models that have been proven to be feasible, and provide experimental workers with electrochemical related curves and data that can be generated with one click.

[0115] (3) Apply the latest descriptors in the field without user input to guide high-throughput calculations for screening.

[0116] A computer program that executes the data-driven high-throughput screening method for low-dimensional energy materials in this embodiment can be written and written into a storage medium or a computer device. When the computer program is read out and run, the data-driven high-throughput screening method for low-dimensional energy materials in this embodiment is executed, thereby achieving the same technical effect as the data-driven high-throughput screening method for low-dimensional energy materials in the embodiment.

[0117] It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to the other feature, or it may be indirectly fixed or connected to the other feature. In addition, the descriptions of up, down, left, right, etc. used in this disclosure are only relative to the relative positional relationship of the components of the present disclosure in the accompanying drawings. The singular forms of "a", "said" and "the" used in this disclosure are also intended to include the plural forms, unless the context clearly indicates otherwise. In addition, unless otherwise defined, all technical and scientific terms used in this embodiment have the same meaning as those generally understood by those skilled in the art. The terms used in the description of this embodiment are only for describing specific embodiments and are not intended to limit the present invention. The term "and / or" used in this embodiment includes any combination of one or more related listed items.

[0118] It should be understood that, although the present disclosure may adopt the term first, second, third etc. to describe various elements, these elements should not be limited to these terms.These terms are only used to distinguish the elements of the same type from each other.For example, without departing from the scope of the present disclosure, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element.The use of any and all examples or exemplary language ("for example", "such as" etc.) provided by the present embodiment is only intended to better illustrate embodiments of the present invention, and unless otherwise required, the scope of the present invention will not be limited.

[0119] It should be appreciated that embodiments of the present invention can be implemented or practiced by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory. The methods can be implemented in a computer program using standard programming techniques, including a non-transitory computer-readable storage medium configured with a computer program, wherein the storage medium so configured causes the computer to operate in a specific and predefined manner, according to the methods and figures described in the specific embodiments. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if desired, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can be run on a programmed application-specific integrated circuit for this purpose.

[0120] In addition, the operations of the processes described in this embodiment may be performed in any suitable order, unless otherwise indicated in this embodiment or otherwise clearly contradicted by the context. The processes described in this embodiment (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed collectively on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions that can be executed by one or more processors.

[0121] Furthermore, the methods can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Various aspects of the present invention can be implemented as machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, RAM, ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein. In addition, the machine-readable code, or portions thereof, can be transmitted over a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention also includes the computer itself.

[0122] The computer program can be applied to input data to perform the functions described in the present embodiment, thereby converting the input data to generate output data that is stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents a physical and tangible object, including a specific visual depiction of the physical and tangible object produced on the display.

[0123] The above description is merely a preferred embodiment of the present invention. The present invention is not limited to the aforementioned embodiments. As long as the technical effects of the present invention are achieved by the same means, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention. Within the scope of protection of the present invention, various modifications and variations of the technical solutions and / or implementation methods are possible.

Claims

1. A data-driven high-throughput screening method for low-dimensional energy materials, characterized in that: The data-driven high-throughput screening method for low-dimensional energy materials includes: Obtaining substrate structure information; Construct a two-dimensional electrocatalyst structure model; simulating an electrochemical reaction mechanism based on the substrate structure information and the two-dimensional electrocatalyst structure model; Calculate the overpotential based on the electrochemical reaction mechanism; Using a machine learning model as a judgment model, predicting the electrocatalyst based on the two-dimensional electrocatalyst structure model and the overpotential; The constructing of the two-dimensional electrocatalyst structure model comprises: Obtaining a lattice vector of a unit cell of a heterojunction upper structure and a lattice vector of a unit cell of a heterojunction lower structure; Perform several iterative processes; in any iterative process, within the lattice period of this iterative process, all construction schemes between the heterojunction upper structure and the heterojunction lower structure are traversed and calculated based on the lattice vectors of the unit cell of the heterojunction upper structure and the lattice vectors of the unit cell of the lower structure, and the mismatch rate and mismatch angle corresponding to each construction scheme are calculated; when a feasible construction scheme is screened out in this iterative process, this iterative process is taken as the last iterative process, and otherwise the next iterative process is performed; wherein the mismatch rate corresponding to the feasible construction scheme is less than a preset mismatch rate threshold, and the mismatch angle corresponding to the feasible construction scheme is less than a preset mismatch angle threshold; the lattice period of the first iterative process is the initial lattice period, and the lattice period of other iterative processes is obtained by expanding the lattice period of the previous iterative process; Among all the feasible construction schemes obtained in the last iteration process, the feasible construction scheme with the smallest sum of the number of atoms in the heterojunction upper structure and the heterojunction lower structure is selected to generate the two-dimensional electrocatalyst structure model; The calculating of the overpotential according to the electrochemical reaction mechanism includes: For the nitrogen reduction reaction, perform the following steps: Preliminary screening of catalyst performance based on N2 molecule adsorption energy; The catalytic performance was evaluated based on the free energy changes of the 1st, 2nd, and 6th hydrogenation reactions of adsorbed N2 molecules. If the free energy change of hydrogenation in the above three steps is less than 0.50 eV, the performance is judged to be excellent, and all hydrogenation steps are supplemented; For the oxygen reduction reaction, the following steps are performed: Select catalyst model and reaction type; Calculate the overpotential based on the adsorption free energy of the intermediate products OOH, O, and OH; The adsorption free energy of the intermediate product OH is correlated with the overpotential to generate a two-dimensional volcano plot; For sulfur oxidation / reduction reactions, the following steps are performed: Select catalyst model and reaction type; According to the free energy change of the reduction process, a lithium-sulfur long chain adsorption structure model was constructed; Determine the reaction process with maximum energy change; Calculate the dissociation energy barrier of Li2S clusters.

2. The data-driven high-throughput screening method for low-dimensional energy materials according to claim 1, characterized in that: The obtaining of substrate structure information includes: Establish user interface; Read the substrate structure information through the user operation interface.

3. The data-driven high-throughput screening method for low-dimensional energy materials according to claim 2, characterized in that: The reading of the substrate structure information includes: Extract substrate structure information from the Materials Project database.

4. The data-driven high-throughput screening method for low-dimensional energy materials according to claim 1, characterized in that: The constructing of the two-dimensional electrocatalyst structure model comprises: Reading the substrate structure information through a Python program; generating several types of defect information based on the substrate structure information; Heterogeneous atom information is introduced into the substrate structure information to construct the metal atom coordination environment information and obtain the two-dimensional electrocatalyst structure model.

5. The data-driven high-throughput screening method for low-dimensional energy materials according to any one of claims 1 to 4, characterized in that: The simulating the electrochemical reaction mechanism according to the substrate structure information and the two-dimensional electrocatalyst structure model includes: Select catalyst model and reaction type; Determine the adsorption site based on the structural characteristics of the substrate; Implement the adsorption model of intermediate products.

6. The data-driven high-throughput screening method for low-dimensional energy materials according to any one of claims 1 to 4, characterized in that: The machine learning model is a regression function model, a decision tree model, a random forest model or a neural network model.

7. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory is used to store at least one program, and the processor is used to load the at least one program to execute the data-driven low-dimensional energy material high-throughput screening method according to any one of claims 1 to 6.

8. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to execute the data-driven low-dimensional energy material high-throughput screening method described in any one of claims 1 to 6 when executed by the processor.

Images