A method and device for predicting the selectivity of catalytic conversion of alkanes, electronic equipment and storage medium

Abstract

This invention discloses a method, apparatus, electronic device, and storage medium for predicting the catalytic conversion selectivity of alkanes, belonging to the field of computational chemistry. The method includes: performing steady-state simulation calculations based on a first-precision basis set of parameters to generate the steady-state geometry of reactants, reactant free energies, and the steady-state geometry of each product; identifying transition states based on a second-precision basis set of parameters to generate transition state geometry data; calculating the free energy of each transition state geometry data based on the first-precision basis set of parameters to generate the transition state free energy; calculating the difference between the transition state free energy and the reactant free energy to determine the activation energy barrier value of each reaction path; selecting the reaction path with the smallest activation energy barrier value as the main reaction path; and determining the predicted result for the catalytic conversion selectivity of the alkane molecule to be predicted. By implementing this invention, the problem of unreliable prediction of the catalytic conversion selectivity of alkanes in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of computational chemistry, specifically to a method, apparatus, electronic device, and storage medium for predicting the catalytic conversion selectivity of alkanes. Background Technology

[0002] With the large-scale decommissioning of high-voltage cables, the green and high-value recycling of cross-linked polyethylene (XLPE), the core insulating material, has become a pressing issue for the industry. Catalytic cracking technology is a key pathway for converting waste XLPE into high-value-added fuels or chemicals. The core challenge of catalytic cracking lies in the precise control of microscopic reaction pathways. Using short-chain alkanes as model compounds and exploring the competitive mechanism between carbon-carbon bond breaking and carbon-hydrogen bond breaking under catalysis is the theoretical basis for achieving targeted conversion. Therefore, accurately predicting the catalytic conversion selectivity of alkanes on specific catalyst surfaces at the atomic scale is of significant industrial importance for guiding the rational design of efficient catalysts, reducing experimental trial-and-error costs, and improving the yield of high-value liquid products.

[0003] However, existing methods for predicting the selectivity of alkane catalytic conversion have limitations in their computational strategies. The competition between alkane dehydrogenation and cracking reactions depends on extremely small differences in activation barriers, and accurately distinguishing between dehydrogenation and cracking reactions requires high-precision electronic structure calculations. In actual simulations, the identification of transition state geometry consumes significant computational resources and is difficult to converge. Existing techniques typically employ a single-precision calculation strategy. If high-precision basis set parameters are used throughout the calculation, the identification of transition state geometry is prone to non-convergence or excessively long calculation times due to the complexity of the potential energy surface. If low-precision basis set parameters are used throughout the calculation, although the transition state geometry can be identified, the calculated activation barrier values ​​contain errors and cannot accurately distinguish the main reaction pathway. The contradiction between computational efficiency and computational accuracy makes it difficult for existing techniques to balance the success rate of transition state identification with the accuracy of barrier calculation, thus failing to provide reliable predictions of the catalytic conversion selectivity of alkane molecules. Summary of the Invention

[0004] The present invention provides a method, apparatus, electronic device and storage medium for predicting the catalytic conversion selectivity of alkanes, which can solve the problem of unreliable prediction of the catalytic conversion selectivity of alkanes in the prior art.

[0005] An embodiment of the present invention provides a method for predicting the catalytic conversion selectivity of alkanes, comprising: Obtain the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model; Based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, steady-state simulation calculations are performed based on the preset first-precision basis set parameters to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path. Based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction pathway, transition states are identified using preset second-precision basis set parameters, generating transition state geometry data for each reaction pathway; wherein, the calculation precision of the preset second-precision basis set parameters is lower than that of the preset first-precision basis set parameters. Based on the transition state geometry data of each reaction path, and using the preset first precision basis set parameters, the free energy is calculated to generate the transition state free energy for each reaction path. Calculate the difference between the transition state free energy and the reactant free energy for each reaction pathway to generate the activation energy barrier value for each reaction pathway. The reaction path with the smallest activation energy barrier is taken as the main reaction path; based on the main reaction path, the predicted result of the catalytic conversion selectivity of the alkane molecule to be predicted is determined.

[0006] Furthermore, the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model are obtained, including: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

[0007] Furthermore, the reaction pathway includes a dehydrogenation reaction pathway and a cracking reaction pathway; Based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, steady-state simulation calculations are performed based on preset first-precision basis set parameters to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products for each reaction pathway, including: Based on the preset reaction adsorption site information, the relative spatial position of the initial three-dimensional coordinate data of the alkane molecule to be predicted in the niobium-based catalyst cluster model is adjusted to generate the positioning coordinate data of the alkane molecule to be predicted and the positioning coordinate data of the niobium-based catalyst cluster model. The initial geometric configuration of the reactant system is constructed by combining the location coordinate data of the alkane molecule to be predicted and the location coordinate data of the niobium-based catalyst cluster model. The reactant system includes the atoms that make up the alkane molecule to be predicted and the atoms that make up the niobium-based catalyst cluster model. Based on the preset first precision basis set parameters, the initial geometric configuration of the reactant system is geometrically optimized to generate the steady-state geometric configuration of the reactants; The vibrational frequencies of the steady-state geometry of the reactants are calculated to generate the free energy of the reactants. Based on the pre-defined structure of the dehydrogenation reaction products and the steady-state geometry of the reactants, the initial geometry of the dehydrogenation reaction pathway is constructed. Based on the pre-defined structure of the pyrolysis reaction products and the steady-state geometry of the reactants, the initial geometry of the pyrolysis reaction pathway is constructed. The initial geometry of the dehydrogenation reaction pathway is geometrically optimized to generate the steady-state geometry of the dehydrogenation reaction pathway. The initial geometry of the pyrolysis reaction pathway is geometrically optimized to generate the steady-state geometry of the pyrolysis reaction pathway.

[0008] Furthermore, based on the steady-state geometry of the reactants and the steady-state geometry of the products along each reaction pathway, and using preset second-precision basis set parameters, transition states are identified to generate transition state geometry data for each reaction pathway, including: Based on the steady-state geometry of the reactants and the steady-state geometry of the products in each reaction pathway, the key atom pairs corresponding to the chemical bonds that are broken or formed in each reaction pathway are identified. For each reaction path, the transition state search operation is repeated until the positive and negative structural matching deviation data of each reaction path are both less than the preset similarity threshold, and the transition state geometric configuration data under each reaction path is generated. The transition state search operation includes: Based on the preset second precision basis set parameters, a transition state search calculation is performed on the current initial transition state structure to generate the current candidate transition state geometry of the current reaction path; wherein, the initial transition state structure is a geometry constructed based on the steady-state geometry of the reactants and the steady-state geometry of the products of the current reaction path; Based on the preset second-precision basis set parameters, the reaction path tracking calculation is performed on the current candidate transition state geometry to generate the current reaction path trajectory data; Extract the last geometric configuration that evolves along the positive direction of the reaction coordinates from the current reaction path trajectory data, and generate the current positive evolution endpoint structure of the current reaction path; Extract the last geometric configuration that evolves along the negative direction of the reaction coordinate from the current reaction path trajectory data, and generate the current negative evolution endpoint structure of the current reaction path; Spatially align the current forward evolution endpoint structure and the steady-state geometry of the product of the current reaction path to generate the current forward-aligned evolution structure and the current forward-aligned product structure; Spatially align the current negative evolution endpoint structure and the reactant steady-state geometry to generate the current negative alignment evolution structure and the current negative alignment reactant structure; Calculate the structural deviation between the current forward alignment evolution structure and the current forward alignment product structure, and generate the current forward structure matching deviation data; Calculate the structural deviation between the current negative alignment evolution structure and the current negative alignment reactant structure, and generate the current negative structure matching deviation data; If both the current positive structure matching deviation data and the current negative structure matching deviation data are less than the preset similarity threshold, then the current candidate transition state geometry of the current reaction path is determined as the transition state geometry data of the current reaction path. If not, obtain the current inter-atomic distance data of the key atom pairs in the current initial transition state structure; adjust the current inter-atomic distance data based on the preset adjustment step size to generate corrected inter-atomic distance data; generate a corrected initial transition state structure based on the corrected inter-atomic distance data; update the corrected initial transition state structure to the current initial transition state structure.

[0009] Furthermore, based on the transition state geometry data of each reaction pathway and the preset first-precision basis set parameters, free energy calculations are performed to generate the transition state free energy for each reaction pathway, including: Based on the preset first precision basis set parameters, geometric optimization calculations are performed on the transition state geometric configuration data of each reaction path to generate a high precision transition state geometric configuration for each reaction path. Based on the preset first-precision basis set parameters, single-point energy calculations are performed on the high-precision transition state geometry of each reaction path to generate transition state electron energy data for each reaction path. Based on the preset first precision basis set parameters, the vibrational frequency of the high-precision transition state geometry of each reaction path is calculated, and transition state thermodynamic correction data of each reaction path is generated. Based on the transition state electron energy data and transition state thermodynamic correction data of each reaction path, the transition state free energy of each reaction path is calculated.

[0010] Furthermore, based on the main reaction pathway, the predicted results of the catalytic conversion selectivity of the alkane molecule to be predicted are determined, including: Obtain the type of chemical bond change corresponding to the main reaction path; When the chemical bond change type is carbon-hydrogen bond change, a first prediction result is generated to characterize the alkane molecule to be predicted as having a dehydrogenation reaction advantage. When the chemical bond change type is carbon-carbon bond change, a second prediction result is generated to characterize the alkane molecule to be predicted as having a cracking reaction advantage. The first or second prediction result is determined as the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted.

[0011] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0012] An embodiment of the present invention provides a device for predicting the catalytic conversion selectivity of alkanes, comprising: a data acquisition module, a steady-state simulation module, a transition state identification module, and a selectivity prediction module; The data acquisition module is used to acquire the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model. The steady-state simulation module is used to perform steady-state simulation calculations based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, and based on the preset first precision basis set parameters, to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path. The transition state identification module is used to identify the transition state based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction path, and to generate transition state geometry data for each reaction path based on a preset second precision basis set parameter; wherein the calculation precision of the preset second precision basis set parameter is lower than that of the preset first precision basis set parameter. The selective prediction module is used to calculate the free energy based on the transition state geometry data of each reaction path and a preset first precision basis set parameter, thereby generating the transition state free energy for each reaction path; calculate the difference between the transition state free energy and the reactant free energy for each reaction path, thereby generating the activation energy barrier value for each reaction path; select the reaction path with the smallest activation energy barrier value as the main reaction path; and determine the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted based on the main reaction path.

[0013] Furthermore, the data acquisition module acquires the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, including: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

[0014] Based on the above method embodiments, the present invention provides corresponding electronic device embodiments.

[0015] An embodiment of the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the method for predicting the catalytic conversion selectivity of alkanes as described in any of the above-described method embodiments.

[0016] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0017] An embodiment of the present invention provides a storage medium storing a computer program thereon, wherein, when the computer program is executed, it controls the device where the storage medium is located to perform the method for predicting the catalytic conversion selectivity of alkanes as described in any of the above-described method embodiments.

[0018] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a method, apparatus, electronic device, and storage medium for predicting the catalytic conversion selectivity of alkanes. The method acquires initial three-dimensional coordinate data of alkane molecules and niobium-based catalyst cluster models; performs steady-state simulation calculations based on first-precision basis set parameters to generate the steady-state geometry of reactants, reactant free energies, and the steady-state geometry of products corresponding to each reaction path; identifies transition states using second-precision basis set parameters to generate transition state geometry data for each reaction path; calculates the transition state free energy based on the first-precision basis set parameters to generate the transition state free energy for each reaction path; determines the activation energy barrier of each reaction path by calculating the difference between the transition state free energy and the reactant free energy, and selects the reaction path with the smallest activation energy barrier as the main reaction path to determine the predicted catalytic conversion selectivity of the alkane molecule.

[0019] This invention effectively avoids the risks of non-convergence or excessively long processing times in the transition state identification process caused by high-precision calculations throughout the entire process, as mentioned in the background art, by employing second-precision basis set parameters with lower computational accuracy in the transition state identification step, thus ensuring a high success rate. Simultaneously, this invention combines first-precision basis set parameters with higher computational accuracy to calculate the free energy of the identified transition state, overcoming the shortcomings of the background art, such as large numerical errors in the activation barrier and the inability to accurately distinguish the main reaction pathway, resulting from low-precision calculations. This achieves reliable prediction of the catalytic conversion selectivity of the alkane molecule to be predicted. Attached Figure Description

[0020] Figure 1 This is a schematic flowchart of a method for predicting the catalytic conversion selectivity of alkanes according to an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of a device for predicting the selectivity of catalytic conversion of alkane according to an embodiment of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] like Figure 1 As shown, to address the problem of unreliable prediction of catalytic conversion selectivity of alkanes in existing technologies, an embodiment of the present invention provides a method for predicting the catalytic conversion selectivity of alkanes, comprising at least the following steps: Step S1: Obtain the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model.

[0024] In a preferred embodiment, obtaining initial three-dimensional coordinate data of the alkane molecule to be predicted, and initial three-dimensional coordinate data of the niobium-based catalyst cluster model, includes: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

[0025] Specifically, the initial three-dimensional coordinate data of the alkane molecule to be predicted, as well as the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, are obtained through the following process: First, the model construction and coordinate extraction of the alkane molecule to be predicted are performed. This process begins with obtaining the carbon chain structure information of the alkane molecule to be predicted. The carbon chain structure information defines the connection sequence between carbon atoms, the branching distribution, and the saturated connection state between carbon and hydrogen atoms in the alkane molecule. Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule is constructed using molecular construction tools. The molecular structure model of the alkane molecule to be predicted is a digital representation of the alkane molecule in a computer simulation environment, and it includes all the atoms constituting the alkane molecule and their chemical bond relationships. Subsequently, the molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom constituting the molecular structure model. The positional information of each atom is specifically represented by the X-axis coordinates, Y-axis coordinates, and Z-axis coordinates of each carbon atom and each hydrogen atom in a preset three-dimensional Cartesian coordinate system. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is extracted and determined as the initial three-dimensional coordinate data of the alkane molecule to be predicted, thereby completing the geometric initialization of the reactant molecule.

[0026] Secondly, the construction and coordinate extraction of the niobium-based catalyst cluster model are performed. This process requires obtaining the structural information of the silica cluster support and the niobium-containing active centers. The structural information of the silica cluster support describes the local atomic arrangement used to simulate the surface of the catalyst support; typically, representative silica ring or cage-like structures are selected as the support model. The structural information of the niobium-containing active centers describes the chemical composition and coordination environment of the catalytic active sites. Based on the structural information of the silica cluster support and the niobium-containing active centers, the niobium-based catalyst cluster model is constructed.

[0027] In the constructed niobium-based catalyst cluster model, the niobium-based catalyst cluster model consists of a silica cluster support and niobium-containing active centers supported on the silica cluster support. The niobium-containing active centers are anchored to the surface of the silica cluster support through a specific chemical bonding mechanism. Specifically, niobium atoms in the niobium-containing active centers are connected to silicon atoms in the silica cluster support through oxygen atoms, forming a niobium-oxygen-silicon linkage structure. This linkage structure simulates the grafting state of the active metal component on the support surface. At the same time, the niobium-containing active centers also include hydrogen atoms coordinated with niobium atoms. These hydrogen atoms form niobium-hydrogen coordination bonds with niobium atoms, constituting surface hydride species with specific catalytic activity. This specific cluster model ((=SiO-)2Nb-H3) was chosen because it can accurately reproduce the local electronic environment of the silica-supported niobium catalyst observed in experiments with minimal computational cost. By bridging oxygen atoms to simulate the rigid support effect of the support surface, while preserving the coordination unsaturation characteristics of niobium atoms, this model can reflect both the shape-selective mechanism of the catalyst and meet the requirements of model simplicity for high-throughput computation. The constructed niobium-based catalyst cluster model is analyzed by traversing each atom in the model to extract the positional information of each atom (including niobium, oxygen, silicon, and hydrogen atoms). The positional information of each atom is also represented as three-dimensional spatial coordinates. This positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the model.

[0028] Through the above steps, microscopic chemical substances are transformed into computer-recognizable geometric matrix parameters, establishing the initial spatial topology of each component in the catalytic reaction system. This provides an accurate geometric configuration basis for subsequent high-precision quantum chemical steady-state simulation calculations, ensuring that the simulated reaction path is consistent with the physicochemical characteristics of the actual catalytic system.

[0029] Step S2: Based on the initial three-dimensional coordinate data of the alkane molecule to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, and based on the preset first precision basis set parameters, perform steady-state simulation calculations to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path.

[0030] In a preferred embodiment, the reaction pathway includes a dehydrogenation reaction pathway and a cracking reaction pathway; Based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, steady-state simulation calculations are performed based on preset first-precision basis set parameters to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products for each reaction pathway, including: Based on the preset reaction adsorption site information, the relative spatial position of the initial three-dimensional coordinate data of the alkane molecule to be predicted in the niobium-based catalyst cluster model is adjusted to generate the positioning coordinate data of the alkane molecule to be predicted and the positioning coordinate data of the niobium-based catalyst cluster model. The initial geometric configuration of the reactant system is constructed by combining the location coordinate data of the alkane molecule to be predicted and the location coordinate data of the niobium-based catalyst cluster model. The reactant system includes the atoms that make up the alkane molecule to be predicted and the atoms that make up the niobium-based catalyst cluster model. Based on the preset first precision basis set parameters, the initial geometric configuration of the reactant system is geometrically optimized to generate the steady-state geometric configuration of the reactants; The vibrational frequencies of the steady-state geometry of the reactants are calculated to generate the free energy of the reactants. Based on the pre-defined structure of the dehydrogenation reaction products and the steady-state geometry of the reactants, the initial geometry of the dehydrogenation reaction pathway is constructed. Based on the pre-defined structure of the pyrolysis reaction products and the steady-state geometry of the reactants, the initial geometry of the pyrolysis reaction pathway is constructed. The initial geometry of the dehydrogenation reaction pathway is geometrically optimized to generate the steady-state geometry of the dehydrogenation reaction pathway. The initial geometry of the pyrolysis reaction pathway is geometrically optimized to generate the steady-state geometry of the pyrolysis reaction pathway.

[0031] Specifically, the reaction pathways include dehydrogenation and cracking reaction pathways. The steady-state simulation calculation process for these steps includes three stages: reactant system construction, reactant steady-state calculation, and product system construction and calculation.

[0032] First, the reaction system is constructed and initially positioned. This process is based on preset reaction adsorption site information. Reaction adsorption site information characterizes the active regions where the predicted alkane molecule interacts with the niobium-based catalyst cluster model, typically corresponding to the spatial location of niobium-containing active centers. This preset reaction adsorption site information can be obtained in two ways: first, by parsing the user interface's input command containing a specified atomic number for prior adsorption site settings; second, by using a molecular docking algorithm or a conformational global search algorithm to calculate the van der Waals forces and electrostatic potentials of the predicted alkane molecule at different positions on the surface of the niobium-based catalyst cluster model, automatically identifying the region with the lowest binding energy as the reaction adsorption site information. Based on the reaction adsorption site information, the initial three-dimensional coordinates of the predicted alkane molecule are adjusted to their relative spatial positions within the niobium-based catalyst cluster model. This simulates the process of the predicted alkane molecule approaching the surface of the niobium-based catalyst cluster model and undergoing physical adsorption, thereby generating the positioning coordinates of the predicted alkane molecule and the niobium-based catalyst cluster model. Subsequently, the location coordinates of the alkane molecules to be predicted are combined with the location coordinates of the niobium-based catalyst cluster model to construct the initial geometric configuration of the reactant system. The reactant system is a comprehensive computational model encompassing the atoms constituting the alkane molecules to be predicted and the atoms constituting the niobium-based catalyst cluster model. The dataset of the initial geometric configuration of the reactant system, containing the location coordinates of the alkane molecules to be predicted and the niobium-based catalyst cluster model, provides a reasonable geometric starting point for subsequent quantum chemical calculations.

[0033] Next, the steady-state geometry and free energy of the reactants are calculated. Using preset first-precision basis set parameters, the initial geometry of the reactant system is geometrically optimized. High-precision basis sets (e.g., all-electron basis sets or high-precision pseudopotential basis sets) are typically selected to ensure the accuracy of the electronic structure description. The geometry optimization process iteratively adjusts the positions of atoms in the reactant system, searching for local minima on the potential energy surface until the forces acting on the reactant system converge to a preset standard, thus generating the steady-state geometry of the reactants. The steady-state geometry represents the microstructure of alkane molecules when stably adsorbed on the catalyst surface. Based on this, vibrational frequencies are calculated for the steady-state geometry. These vibrational frequency calculations confirm whether the steady-state geometry represents a true potential energy surface minimum (i.e., no imaginary frequencies), and the free energy of the reactants is generated based on statistical thermodynamic properties derived from the vibrational frequency data. Reactant free energy is a thermodynamic state function characterizing a reactant system at a specific temperature and pressure. It is obtained by summing the electronic energy, zero-point energy correction, and thermal Gibbs free energy correction of the reactant steady-state geometry.

[0034] Finally, the steady-state geometry of the products in each reaction pathway is constructed and calculated. Based on the pre-defined reaction mechanism, the reaction pathways are divided into dehydrogenation and cracking pathways. For the dehydrogenation pathway, the initial geometry is constructed based on the pre-defined dehydrogenation product structure (involving olefins or alkyl species after carbon-hydrogen bond breakage) and the steady-state geometry of the reactants. For the cracking pathway, the initial geometry is constructed based on the pre-defined cracking product structure (involving low-carbon chain species after carbon-carbon bond breakage) and the steady-state geometry of the reactants.

[0035] It should be noted that the preset dehydrogenation reaction product structures and preset pyrolysis reaction product structures can be automatically obtained through the following programmed rules: using a preset chemical reaction template library, based on the steady-state geometry of the reactants as the topological framework, for dehydrogenation reactions, the program automatically breaks the specified carbon-hydrogen bonds and adsorbs the dissociated hydrogen atoms onto the niobium-based active center; for pyrolysis reactions, the program automatically breaks the specified carbon-carbon skeletal bonds. Alternatively, the preset reaction product structures can also be obtained by receiving customized bond-breaking and bond-forming operation commands from expert users in a graphical molecular editor based on the steady-state geometry of the reactants.

[0036] Subsequently, based on the preset first precision basis set parameters, the initial geometric configurations of the dehydrogenation reaction pathway and the cracking reaction pathway were geometrically optimized. The geometric optimization process aimed to eliminate the structural tension introduced by artificial construction, so that the product structure reaches the lowest energy state on the catalyst surface, thereby generating the steady-state geometric configurations of the dehydrogenation reaction pathway and the cracking reaction pathway, respectively.

[0037] Through the above steps, the precise geometric structures and thermodynamic energy scales of the reaction initiation point (reactants) and reaction endpoint (products of each pathway) can be determined at a high level of theoretical accuracy, laying a reliable data foundation for the subsequent accurate identification of the transition state structure connecting the initiation point and the endpoint, as well as the calculation of the activation energy barrier.

[0038] Step S3: Based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction path, and using the preset second-precision basis set parameters, perform transition state identification to generate transition state geometry data for each reaction path; wherein, the calculation precision of the preset second-precision basis set parameters is lower than that of the preset first-precision basis set parameters.

[0039] In a preferred embodiment, based on the steady-state geometry of the reactants and the steady-state geometry of the products along each reaction pathway, and using preset second-precision basis set parameters, transition state identification is performed to generate transition state geometry data for each reaction pathway, including: Based on the steady-state geometry of the reactants and the steady-state geometry of the products in each reaction pathway, the key atom pairs corresponding to the chemical bonds that are broken or formed in each reaction pathway are identified. For each reaction path, the transition state search operation is repeated until the positive and negative structural matching deviation data of each reaction path are both less than the preset similarity threshold, and the transition state geometric configuration data under each reaction path is generated. The transition state search operation includes: Based on the preset second precision basis set parameters, a transition state search calculation is performed on the current initial transition state structure to generate the current candidate transition state geometry of the current reaction path; wherein, the initial transition state structure is a geometry constructed based on the steady-state geometry of the reactants and the steady-state geometry of the products of the current reaction path; Based on the preset second-precision basis set parameters, the reaction path tracking calculation is performed on the current candidate transition state geometry to generate the current reaction path trajectory data; Extract the last geometric configuration that evolves along the positive direction of the reaction coordinates from the current reaction path trajectory data, and generate the current positive evolution endpoint structure of the current reaction path; Extract the last geometric configuration that evolves along the negative direction of the reaction coordinate from the current reaction path trajectory data, and generate the current negative evolution endpoint structure of the current reaction path; Spatially align the current forward evolution endpoint structure and the steady-state geometry of the product of the current reaction path to generate the current forward-aligned evolution structure and the current forward-aligned product structure; Spatially align the current negative evolution endpoint structure and the reactant steady-state geometry to generate the current negative alignment evolution structure and the current negative alignment reactant structure; Calculate the structural deviation between the current forward alignment evolution structure and the current forward alignment product structure, and generate the current forward structure matching deviation data; Calculate the structural deviation between the current negative alignment evolution structure and the current negative alignment reactant structure, and generate the current negative structure matching deviation data; If both the current positive structure matching deviation data and the current negative structure matching deviation data are less than the preset similarity threshold, then the current candidate transition state geometry of the current reaction path is determined as the transition state geometry data of the current reaction path. If not, obtain the current inter-atomic distance data of the key atom pairs in the current initial transition state structure; adjust the current inter-atomic distance data based on the preset adjustment step size to generate corrected inter-atomic distance data; generate a corrected initial transition state structure based on the corrected inter-atomic distance data; update the corrected initial transition state structure to the current initial transition state structure.

[0040] Specifically, the transition state identification process is a crucial step connecting the reaction initiation and termination points, aiming to find a first-order saddle point with a single imaginary frequency on the potential energy surface. To balance computational efficiency and search success rate, this step employs pre-defined second-precision basis set parameters. These pre-defined second-precision basis set parameters typically use a smaller basis set size or a simplified pseudopotential strategy. Compared to the pre-defined first-precision basis set parameters used for subsequent high-precision energy calculations, these parameters significantly reduce the computational time of single-step geometry optimization, thus allowing for multiple iterative searches within limited computational resources.

[0041] In a specific embodiment, based on the steady-state geometry of the reactants and the steady-state geometry of the products under each reaction pathway, and using preset second-precision basis set parameters, transition state identification is performed to generate transition state geometry data for each reaction pathway. The detailed execution process is as follows: First, key atom pairs are identified and an initial structure is constructed. Based on the steady-state geometry of the reactants and the steady-state geometry of the products in the current reaction path, key atom pairs corresponding to the broken or formed chemical bonds in each reaction path are identified by comparing the changes in atomic connections between the two. Key atom pairs typically involve carbon-hydrogen bonds undergoing dehydrogenation or carbon-carbon bonds undergoing cleavage in the alkane molecule to be predicted, as well as active site atoms participating in the reaction in the niobium-based catalyst cluster model. The initial transition state structure is usually constructed based on a linear synchronous transition or a quadratic synchronous transition algorithm, according to the steady-state geometry of the reactants and the steady-state geometry of the products in the current reaction path, to ensure that the initial guessed structure is located on the geometric path between the reactants and products. Specifically, the computer program can calculate the distance matrix between each atom in the steady-state geometry of the reactants and the steady-state geometry of the products. By comparing the two distance matrices, atom pairs whose rate of change in inter-atomic distance exceeds a preset threshold (e.g., a rate of change exceeding 20% ​​or a distance change exceeding 0.5 Å) are selected and marked as key atom pairs. For example, in the alkane dehydrogenation pathway, the distance between the carbon and hydrogen atoms involved in the reaction is the bonding distance (about 1.09 Å) in the reactants, but the non-bonding distance (greater than 2.0 Å) in the products. This significant change will be captured and flagged by the program.

[0042] Secondly, for each reaction path, the transition state search operation and verification loop are repeated. Within the loop, based on preset second-precision basis set parameters, a transition state search calculation is performed on the current initial transition state structure. The transition state search calculation calculates the Hessian matrix and searches for negative eigenvalues ​​to locate saddle points on the potential energy surface, generating the current candidate transition state geometry for the current reaction path. The current candidate transition state geometry must satisfy the physical condition of having only a unique imaginary frequency.

[0043] Subsequently, the correctness of the current candidate transition state geometry is verified. This process is achieved through reaction path tracing calculations, specifically by calculating intrinsic reaction coordinates. Based on preset second-precision basis set parameters, starting from the current candidate transition state geometry, the steepest descent path evolution is performed along both the positive (product direction) and negative (reactant direction) directions of the reaction coordinates, generating the current reaction path trajectory data. The current reaction path trajectory data records a series of intermediate geometry configurations as the transition state slides towards the potential valleys on both sides.

[0044] Next, spatial alignment and deviation verification of the evolution endpoint structures are performed. From the current reaction path trajectory data, the last geometric configuration evolving along the positive direction of the reaction coordinates is extracted to generate the current positive evolution endpoint structure of the current reaction path; simultaneously, the last geometric configuration evolving along the negative direction of the reaction coordinates is extracted to generate the current negative evolution endpoint structure of the current reaction path. Since molecules may rotate or translate during the calculation process, directly comparing coordinate values ​​may introduce errors, so spatial alignment is required. The current positive evolution endpoint structure is spatially aligned with the steady-state geometric configuration of the product of the current reaction path, and the current positive aligned evolution structure and the current positive aligned product structure are generated by minimizing the root mean square deviation algorithm; similarly, the current negative evolution endpoint structure is spatially aligned with the steady-state geometric configuration of the reactants to generate the current negative aligned evolution structure and the current negative aligned reactant structure. Based on this, the structural deviation between the current positively aligned evolutionary structure and the current positively aligned product structure is calculated to generate the current positive structure matching deviation data; and the structural deviation between the current negatively aligned evolutionary structure and the current negatively aligned reactant structure is calculated to generate the current negative structure matching deviation data.

[0045] Finally, a loop termination judgment and structural correction are performed. It is determined whether both the current positive and negative structural matching deviation data are less than a preset similarity threshold. The preset similarity threshold is used to determine the degree of geometrical overlap. In a specific embodiment, when the root mean square deviation (RMSD) algorithm is used to calculate the structural deviation, the preset similarity threshold is preferably between 0.01 Å and 0.5 Å, more preferably between 0.05 Å and 0.1 Å. When the structural deviation is less than this range, within the error range allowed by computational chemistry accuracy, the two geometric configurations can be considered spatially equivalent, meaning the transition state correctly connects the target reactant or product. If the judgment result is yes, it indicates that the current candidate transition state geometric configuration correctly connects the preset reactant and product. Therefore, the current candidate transition state geometric configuration of the current reaction path is determined as the transition state geometric configuration data of the current reaction path, and the search for the current path ends. If the result is negative, it indicates that the searched transition state is connected to the wrong reactant or product. In this case, it is necessary to obtain the current interatomic distance data of the key atom pairs in the current initial transition state structure, and adjust the current interatomic distance data based on a preset adjustment step size (e.g., increasing or decreasing the key bond length) to generate corrected interatomic distance data. The geometry is then constrained or deformed based on the corrected interatomic distance data to generate a corrected initial transition state structure. This corrected initial transition state structure is then updated to the current initial transition state structure, and the transition state search calculation step is returned until a transition state that meets the conditions is found.

[0046] It should be noted that the preset adjustment step size should be selected as a value that allows for fine-tuning of the structure without causing structural collapse. In a preferred embodiment, the preset adjustment step size ranges from 0.01 Å to 0.2 Å. For example, if the currently searched transition state is too far from the reactants, the distance between the key bonding atoms is shortened by 0.05 Å; if the distance is too close, the distance is extended by 0.05 Å. This fine-tuning strategy can guide the transition state search algorithm (such as the Berny algorithm or the QST2 / 3 algorithm) to escape erroneous local minima and reconverge to the correct saddle point region.

[0047] By employing the above steps and utilizing a double-precision strategy and a closed-loop verification mechanism, the problem of transition state search in complex catalytic systems being difficult to converge or converging to an incorrect structure can be effectively solved. Under the premise of ensuring controllable computational resources, the uniqueness and correctness of the identified transition state geometry in the chemical reaction logic can be guaranteed.

[0048] Step S4: Based on the transition state geometry data of each reaction path and the preset first precision basis set parameters, calculate the free energy to generate the transition state free energy under each reaction path.

[0049] In a preferred embodiment, based on the transition state geometry data of each reaction path and a preset first precision basis set parameter, the free energy is calculated to generate the transition state free energy for each reaction path, including: Based on the preset first precision basis set parameters, geometric optimization calculations are performed on the transition state geometric configuration data of each reaction path to generate a high precision transition state geometric configuration for each reaction path. Based on the preset first-precision basis set parameters, single-point energy calculations are performed on the high-precision transition state geometry of each reaction path to generate transition state electron energy data for each reaction path. Based on the preset first precision basis set parameters, the vibrational frequency of the high-precision transition state geometry of each reaction path is calculated, and transition state thermodynamic correction data of each reaction path is generated. Based on the transition state electron energy data and transition state thermodynamic correction data of each reaction path, the transition state free energy of each reaction path is calculated.

[0050] In a specific embodiment, based on the transition state geometry data of each reaction path and a preset first-precision basis set parameter, the free energy is calculated to generate the transition state free energy for each reaction path. The specific execution process is as follows: First, a high-precision geometric re-optimization operation is performed. Since the transition state geometric configuration data for each reaction path generated in the previous steps are obtained based on preset second-precision basis set parameters with relatively low computational accuracy, fine-tuning is required to eliminate basis set overlap errors and obtain geometric structures that more closely resemble the actual physical state. Based on preset first-precision basis set parameters, geometric optimization calculations are performed on the transition state geometric configuration data for each reaction path. The preset first-precision basis set parameters typically employ an all-electron basis set or a high-precision pseudopotential basis set. For example, the LANL2TZ pseudopotential basis set is used in calculations for niobium atoms, and the 6-311++G(d,p) basis set is used in calculations for carbon, hydrogen, oxygen, and silicon atoms. Through geometric optimization calculations, atoms are re-found on a higher-precision potential energy surface to find their force equilibrium points, thereby generating high-precision transition state geometric configurations for each reaction path.

[0051] It should be noted that the above basis set is only a specific example of the first-precision basis set parameters, and not a limitation. In practical applications, the first-precision basis set parameters are intended to provide a high-precision description of the electronic correlation energy. For main group elements (such as C, H, O, Si), the first-precision basis set parameters can also be selected from 6-311G, def2-TZVP, cc-pVTZ, or other basis sets with triplet valence or higher precision; for transition metal elements (such as Nb), the first-precision basis set parameters can also be selected from SDD, def2-TZVP, or other pseudopotential basis sets that include relativistic effect corrections. Correspondingly, the preset second-precision basis set parameters (low precision) mentioned in step S3 above must have a significantly lower computational cost than the first-precision basis set parameters. For example, 3-21G, 6-31G(d), LANL2MB, or STO-3G split valence basis sets or minimal basis sets can be selected.

[0052] Secondly, precise calculations of electron energies are performed. Based on preset first-precision basis set parameters, single-point energy calculations are conducted on the high-precision transition state geometry of each reaction path. The single-point energy calculation aims to solve the Schrödinger equation without altering the geometry to obtain the electron wavefunction and corresponding potential energy value of the system at absolute zero, thereby generating transition state electron energy data for each reaction path. The transition state electron energy data for each reaction path represents the pure electron energy of the transition state system under the Born-Oppenheimer approximation.

[0053] Subsequently, thermodynamic property correction calculations are performed. Based on preset first-precision basis set parameters, vibrational frequencies are calculated for the high-precision transition state geometry of each reaction path. The vibrational frequency calculations obtain the normal vibrational modes of the system by solving the Hessian matrix. Based on these normal vibrational modes, and combined with preset temperature parameters (e.g., 298.15 K) and pressure parameters (e.g., 1 atmosphere), statistical thermodynamic methods are used to calculate the zero-point vibrational energy correction, enthalpy correction, and entropy correction. These corrections are then integrated to generate the transition state thermodynamic correction data for each reaction path.

[0054] Finally, the Gibbs free energy synthesis calculation is performed. Based on the transition state electron energy data and transition state thermodynamic correction data for each reaction pathway, the transition state free energy for each reaction pathway is calculated. Specifically, the transition state electron energy data and the transition state thermodynamic correction data for each reaction pathway are summed. By introducing thermodynamic correction, the microscopic quantum chemical calculation results are transformed into macroscopic thermodynamic state functions, thus obtaining the transition state free energy for each reaction pathway that includes contributions from enthalpy and entropy changes.

[0055] By combining the high-precision basis set for accurate description of electronic correlation energy with thermodynamic correction for temperature effects, the energy error caused by low-precision search is eliminated, ensuring that the activation barrier value obtained in subsequent calculations has high chemical accuracy, thereby enabling precise differentiation of the competitive advantages between different reaction pathways.

[0056] Step S5: Calculate the difference between the transition state free energy and the reactant free energy for each reaction path to generate the activation barrier value for each reaction path.

[0057] In one specific embodiment, the difference between the transition state free energy and the reactant free energy of each reaction pathway is calculated to generate the activation energy barrier value for each reaction pathway. This aims to quantify, from a kinetic perspective, the ease with which the alkane molecule to be predicted undergoes chemical transformation along different reaction pathways. The activation energy barrier value not only reflects the minimum energy required for the reactant molecule to cross the saddle point of the potential energy surface, but is also the core basis for determining the dominant pathway in a competing reaction.

[0058] Specifically, the reactant free energies generated in step S2 and the transition state free energies of each reaction pathway generated in step S4 are retrieved. For each independent reaction pathway, an activation barrier calculation is performed. The activation barrier calculation is based on a preset barrier calculation formula, which is as follows: In the formula for calculating the energy barrier, This indicates the activation barrier value of the currently calculated reaction path. The unit of activation barrier value is usually kilocalories per mole (kcal / mol) or electron volts (eV). This represents the transition state free energy of the reaction path currently being calculated. The transition state free energy is the Gibbs free energy calculated using a high-precision basis set and after thermodynamic correction. This represents the reactant free energy, which is also the Gibbs free energy obtained from high-precision basis set calculations. Through the above calculations, the activation energy barriers for the dehydrogenation reaction pathway and the cracking reaction pathway are obtained, respectively. A higher activation energy barrier value indicates a slower reaction rate, requiring higher temperatures or more stringent conditions; a lower activation energy barrier value indicates a faster reaction rate, possessing a kinetic advantage in competing reactions.

[0059] Through the above steps, complex microscopic electronic structure data is transformed into an intuitive energy difference index, providing a direct and quantitative evaluation criterion for subsequent screening of main reaction pathways based on the principle of minimum energy.

[0060] Step S6: Select the reaction path with the smallest activation energy barrier as the main reaction path; based on the main reaction path, determine the predicted result of the catalytic conversion selectivity of the alkane molecule to be predicted.

[0061] In a preferred embodiment, the predicted result of the catalytic conversion selectivity of the alkane molecule to be predicted is determined based on the main reaction pathway, including: Obtain the type of chemical bond change corresponding to the main reaction path; When the chemical bond change type is carbon-hydrogen bond change, a first prediction result is generated to characterize the alkane molecule to be predicted as having a dehydrogenation reaction advantage. When the chemical bond change type is carbon-carbon bond change, a second prediction result is generated to characterize the alkane molecule to be predicted as having a cracking reaction advantage. The first or second prediction result is determined as the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted.

[0062] In a specific embodiment, step S6 aims to perform a final qualitative and quantitative assessment of the complex competing reaction network based on the principle of kinetic advantage. In the catalytic reaction system, the reaction rate is negatively correlated with the activation energy barrier value; the lower the activation energy barrier value, the higher the probability of the reaction occurring and the faster the reaction rate. Therefore, the activation energy barrier values ​​of each reaction path generated in step S5 are sorted and compared, and the reaction path corresponding to the lowest activation energy barrier value is selected. The reaction path corresponding to the lowest activation energy barrier value is determined as the main reaction path. The main reaction path represents the chemical transformation pathway that the predicted alkane molecule is most likely to undergo under the current niobium-based catalyst cluster model.

[0063] Subsequently, a microscopic mechanism analysis based on the main reaction pathway is performed to determine the predicted catalytic conversion selectivity of the alkane molecule. This process specifically includes: obtaining the types of chemical bond changes corresponding to the main reaction pathway; and identifying the key chemical bonds that break in the main reaction pathway by comparing the atomic connections between the steady-state geometry of the reactants and the transition state geometry (or the steady-state geometry of the products) in the main reaction pathway. The types of chemical bond changes are mainly classified into carbon-hydrogen bond changes and carbon-carbon bond changes.

[0064] When the chemical bond change type is a carbon-hydrogen bond change, it indicates that the microscopic process corresponding to the main reaction pathway involves the breaking of the chemical bond between the carbon and hydrogen atoms in the alkane molecule, while the carbon skeleton remains intact. This situation corresponds to a dehydrogenation reaction mechanism, meaning that the alkane molecule to be predicted tends to lose hydrogen atoms under the action of a catalyst to generate the corresponding olefin product. In this case, a first prediction result is generated to characterize the alkane molecule's predominance in dehydrogenation reactions. Specifically, the first prediction result indicates that the catalytic conversion selectivity of the alkane molecule to be predicted tends to be highly selective in generating the dehydrogenation product.

[0065] When the chemical bond change type is a carbon-carbon bond change, it indicates that the microscopic process corresponding to the main reaction pathway involves the breaking of the skeletal bonds between carbon atoms in the alkane molecule. This corresponds to a cracking reaction mechanism, meaning that the alkane molecule to be predicted tends to undergo carbon chain breakage under the action of a catalyst, generating lower-carbon alkane or lower-carbon olefin fragments with fewer carbon atoms. In this case, a second prediction result is generated to characterize the alkane molecule's predominance in cracking reactions. Specifically, the second prediction result indicates that the catalytic conversion selectivity of the alkane molecule to be predicted tends towards deep cracking side reactions.

[0066] Finally, the generated first or second prediction result is determined as the prediction result for the catalytic conversion selectivity of the alkane molecule to be predicted. Through the above steps, this method can directly map the energy barrier difference at the atomic level to the macroscopic reaction selectivity, thereby predicting the catalytic tendency of niobium-based catalysts for alkanes of specific chain lengths before experimental synthesis, and realizing the theoretical prediction of catalyst shape selectivity and product distribution.

[0067] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0068] like Figure 2 As shown, an embodiment of the present invention provides a device for predicting the catalytic conversion selectivity of alkanes, comprising: a data acquisition module, a steady-state simulation module, a transition state identification module, and a selectivity prediction module; The data acquisition module is used to acquire the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model. The steady-state simulation module is used to perform steady-state simulation calculations based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, and based on the preset first precision basis set parameters, to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path. The transition state identification module is used to identify the transition state based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction path, and to generate transition state geometry data for each reaction path based on a preset second precision basis set parameter; wherein the calculation precision of the preset second precision basis set parameter is lower than that of the preset first precision basis set parameter. The selective prediction module is used to calculate the free energy based on the transition state geometry data of each reaction path and a preset first precision basis set parameter, thereby generating the transition state free energy for each reaction path; calculate the difference between the transition state free energy and the reactant free energy for each reaction path, thereby generating the activation energy barrier value for each reaction path; select the reaction path with the smallest activation energy barrier value as the main reaction path; and determine the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted based on the main reaction path.

[0069] The data acquisition module acquires the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, including: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

[0070] It should be noted that the embodiments of the apparatus described above correspond to the embodiments of the present invention described above, and are capable of realizing the method for predicting the catalytic conversion selectivity of alkanes as described in any one of the above embodiments of the present invention. Furthermore, the embodiments of the apparatus described above are merely illustrative. The modules described as separation components may or may not be physically separated, and the components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Additionally, in the accompanying drawings of the apparatus embodiments provided by the present invention, the connection relationship between modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort.

[0071] Based on the above-described method embodiments of the present invention, a corresponding embodiment of an electronic device is provided.

[0072] An embodiment of the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the method for predicting the catalytic conversion selectivity of alkanes according to any one of the present invention, or, when the processor executes the computer program, it implements the functions of each module in the above-described device embodiments.

[0073] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the terminal device.

[0074] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0075] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0076] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function, etc.; the data storage area may store data created based on the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0077] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments; Another embodiment of the present invention provides a storage medium comprising a stored computer program, wherein, when the computer program is executed, the device in which the storage medium is located is controlled to perform the method for predicting the catalytic conversion selectivity of any one of the alkanes described above.

[0078] The aforementioned storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0079] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0080] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for predicting the selectivity of a catalytic conversion of an alkane, characterized in that, include: Obtain the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model; Based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, steady-state simulation calculations are performed based on the preset first-precision basis set parameters to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path. Based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction pathway, transition states are identified using preset second-precision basis set parameters, generating transition state geometry data for each reaction pathway; wherein, the calculation precision of the preset second-precision basis set parameters is lower than that of the preset first-precision basis set parameters. Based on the transition state geometry data of each reaction path, and using the preset first precision basis set parameters, the free energy is calculated to generate the transition state free energy for each reaction path. Calculate the difference between the transition state free energy and the reactant free energy for each reaction pathway to generate the activation energy barrier value for each reaction pathway. The reaction path with the smallest activation energy barrier is taken as the main reaction path; based on the main reaction path, the predicted result of the catalytic conversion selectivity of the alkane molecule to be predicted is determined.

2. The method of predicting the selectivity of the catalytic conversion of an alkane according to claim 1, wherein, Obtain the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, including: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

3. The method of predicting the selectivity of the catalytic conversion of an alkane according to claim 2, wherein, The reaction pathway includes a dehydrogenation reaction pathway and a cracking reaction pathway; Based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, steady-state simulation calculations are performed based on preset first-precision basis set parameters to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products for each reaction pathway, including: Based on the preset reaction adsorption site information, the relative spatial position of the initial three-dimensional coordinate data of the alkane molecule to be predicted in the niobium-based catalyst cluster model is adjusted to generate the positioning coordinate data of the alkane molecule to be predicted and the positioning coordinate data of the niobium-based catalyst cluster model. The initial geometric configuration of the reactant system is constructed by combining the location coordinate data of the alkane molecule to be predicted and the location coordinate data of the niobium-based catalyst cluster model. The reactant system includes the atoms that make up the alkane molecule to be predicted and the atoms that make up the niobium-based catalyst cluster model. Based on the preset first precision basis set parameters, the initial geometric configuration of the reactant system is geometrically optimized to generate the steady-state geometric configuration of the reactants; The vibrational frequencies of the steady-state geometry of the reactants are calculated to generate the free energy of the reactants. Based on the pre-defined structure of the dehydrogenation reaction products and the steady-state geometry of the reactants, the initial geometry of the dehydrogenation reaction pathway is constructed. Based on the pre-defined structure of the pyrolysis reaction products and the steady-state geometry of the reactants, the initial geometry of the pyrolysis reaction pathway is constructed. The initial geometry of the dehydrogenation reaction pathway is geometrically optimized to generate the steady-state geometry of the dehydrogenation reaction pathway. The initial geometry of the pyrolysis reaction pathway is geometrically optimized to generate the steady-state geometry of the pyrolysis reaction pathway.

4. The method of predicting the selectivity of the catalytic conversion of an alkane according to claim 3, wherein, Based on the steady-state geometry of the reactants and the steady-state geometry of the products along each reaction pathway, and using preset second-precision basis set parameters, transition states are identified to generate transition state geometry data for each reaction pathway, including: Based on the steady-state geometry of the reactants and the steady-state geometry of the products in each reaction pathway, the key atom pairs corresponding to the chemical bonds that are broken or formed in each reaction pathway are identified. For each reaction path, the transition state search operation is repeated until the positive and negative structural matching deviation data of each reaction path are both less than the preset similarity threshold, and the transition state geometric configuration data under each reaction path is generated. The transition state search operation includes: Based on the preset second precision basis set parameters, a transition state search calculation is performed on the current initial transition state structure to generate the current candidate transition state geometry of the current reaction path; wherein, the initial transition state structure is a geometry constructed based on the steady-state geometry of the reactants and the steady-state geometry of the products of the current reaction path; Based on the preset second-precision basis set parameters, the reaction path tracking calculation is performed on the current candidate transition state geometry to generate the current reaction path trajectory data; Extract the last geometric configuration that evolves along the positive direction of the reaction coordinates from the current reaction path trajectory data, and generate the current positive evolution endpoint structure of the current reaction path; Extract the last geometric configuration that evolves along the negative direction of the reaction coordinate from the current reaction path trajectory data, and generate the current negative evolution endpoint structure of the current reaction path; Spatially align the current forward evolution endpoint structure and the steady-state geometry of the product of the current reaction path to generate the current forward-aligned evolution structure and the current forward-aligned product structure; Spatially align the current negative evolution endpoint structure and the reactant steady-state geometry to generate the current negative alignment evolution structure and the current negative alignment reactant structure; Calculate the structural deviation between the current forward alignment evolution structure and the current forward alignment product structure, and generate the current forward structure matching deviation data; Calculate the structural deviation between the current negative alignment evolution structure and the current negative alignment reactant structure, and generate the current negative structure matching deviation data; If both the current positive structure matching deviation data and the current negative structure matching deviation data are less than the preset similarity threshold, then the current candidate transition state geometry of the current reaction path is determined as the transition state geometry data of the current reaction path. If not, obtain the current inter-atomic distance data of the key atom pairs in the current initial transition state structure; adjust the current inter-atomic distance data based on the preset adjustment step size to generate corrected inter-atomic distance data; generate a corrected initial transition state structure based on the corrected inter-atomic distance data; update the corrected initial transition state structure to the current initial transition state structure.

5. The method of predicting the selectivity of the catalytic conversion of an alkane according to claim 4, wherein, Based on the transition state geometry data of each reaction pathway, and using preset first-precision basis set parameters, free energy calculations are performed to generate the transition state free energy for each reaction pathway, including: Based on the preset first precision basis set parameters, geometric optimization calculations are performed on the transition state geometric configuration data of each reaction path to generate a high precision transition state geometric configuration for each reaction path. Based on the preset first-precision basis set parameters, single-point energy calculations are performed on the high-precision transition state geometry of each reaction path to generate transition state electron energy data for each reaction path. Based on the preset first precision basis set parameters, the vibrational frequency of the high-precision transition state geometry of each reaction path is calculated, and transition state thermodynamic correction data of each reaction path is generated. Based on the transition state electron energy data and transition state thermodynamic correction data of each reaction path, the transition state free energy of each reaction path is calculated.

6. The method of predicting the selectivity of the catalytic conversion of an alkane according to claim 5, wherein, Based on the main reaction pathway, the predicted results for the catalytic conversion selectivity of the alkane molecules to be predicted are determined, including: Obtain the type of chemical bond change corresponding to the main reaction path; When the chemical bond change type is carbon-hydrogen bond change, a first prediction result is generated to characterize the alkane molecule to be predicted as having a dehydrogenation reaction advantage. When the chemical bond change type is carbon-carbon bond change, a second prediction result is generated to characterize the alkane molecule to be predicted as having a cracking reaction advantage. The first or second prediction result is determined as the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted.

7. An apparatus for predicting the selectivity of catalytic conversion of an alkane, characterized by include: The module includes a data acquisition module, a steady-state simulation module, a transient state identification module, and a selective prediction module. The data acquisition module is used to acquire the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model. The steady-state simulation module is used to perform steady-state simulation calculations based on the initial three-dimensional coordinate data of the alkane molecules to be predicted and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, and based on the preset first precision basis set parameters, to generate the steady-state geometry of the reactants, the free energy of the reactants, and the steady-state geometry of the products of each reaction path. The transition state identification module is used to identify the transition state based on the steady-state geometry of the reactants and the steady-state geometry of the products of each reaction path, and to generate transition state geometry data for each reaction path based on a preset second precision basis set parameter; wherein the calculation precision of the preset second precision basis set parameter is lower than that of the preset first precision basis set parameter. The selective prediction module is used to calculate the free energy based on the transition state geometry configuration data of each reaction path and a preset first precision basis set parameter, and generate the transition state free energy under each reaction path. The difference between the transition state free energy and the reactant free energy of each reaction pathway is calculated to generate the activation energy barrier value of each reaction pathway; the reaction pathway with the smallest activation energy barrier value is taken as the main reaction pathway; based on the main reaction pathway, the prediction result of the catalytic conversion selectivity of the alkane molecule to be predicted is determined.

8. The device for predicting the catalytic conversion selectivity of alkanes as described in claim 7, characterized in that, The data acquisition module acquires the initial three-dimensional coordinate data of the alkane molecule to be predicted, and the initial three-dimensional coordinate data of the niobium-based catalyst cluster model, including: Obtain the carbon chain structure information of the alkane molecule to be predicted, the structural information of the silica cluster support, and the structural information of the niobium-containing active center; Based on the carbon chain structure information of the alkane molecule to be predicted, a molecular structure model of the alkane molecule to be predicted is constructed. The molecular structure model of the alkane molecule to be predicted is analyzed to extract the positional information of each atom that constitutes the molecular structure model of the alkane molecule to be predicted. The positional information of each atom in the molecular structure model that constitutes the alkane molecule to be predicted is used as the initial three-dimensional coordinate data of the alkane molecule to be predicted. Based on the structural information of the silica cluster support and the niobium-containing active center, a niobium-based catalyst cluster model is constructed. The niobium-based catalyst cluster model includes a silica cluster support and a niobium-containing active center supported on the silica cluster support. The niobium-containing active center includes niobium atoms and hydrogen atoms coordinated to the niobium atoms. The niobium atoms are connected to the silica cluster support through oxygen atoms. The cluster model of niobium-based catalyst was analyzed to extract the positional information of each atom constituting the cluster model of niobium-based catalyst; The positional information of each atom constituting the niobium-based catalyst cluster model is used as the initial three-dimensional coordinate data for the niobium-based catalyst cluster model.

9. An electronic device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the method for predicting the catalytic conversion selectivity of alkanes as described in any one of claims 1 to 6.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the storage medium to perform the method for predicting the catalytic conversion selectivity of alkanes as described in any one of claims 1 to 6.

Images