An analysis method and analysis system for analyzing the performance of a palladium complex in catalyzing copolymerization of ethylene and vinyl acetate

Abstract

The application provides an analysis method and analysis system for the performance of a palladium complex catalyst in copolymerization of ethylene and vinyl acetate. The analysis method comprises the following steps: obtaining the structural information of different palladium complex catalysts, wherein the palladium complex catalysts are composed of auxiliary ligands and palladium alkyl fragments; simplifying the structure of the palladium complex catalysts and calculating the reaction energy barrier of each palladium complex catalyst in the copolymerization reaction of ethylene and vinyl acetate; extracting the descriptors of the reaction center fragments of each palladium complex catalyst; and constructing a reaction energy barrier prediction model for the copolymerization of ethylene and vinyl acetate based on the calculated reaction energy barrier and the descriptors, which is used to analyze the performance of the palladium complex catalyst in the copolymerization of ethylene and vinyl acetate. The analysis method provided by the application can help to accelerate the screening of copolymerization catalysts, reduce the research and development cost, realize the rational design of copolymerization catalysts, and improve the success rate of copolymerization catalyst design.

Description

Technical Field

[0001] This invention relates to an analytical method and system for analyzing the performance of palladium complexes in the copolymerization of ethylene and vinyl acetate, belonging to the field of polyolefin technology. Background Technology

[0002] Palladium complexes play a crucial role in the catalytic copolymerization of ethylene and vinyl acetate (VAE). Palladium catalysts can adjust the ratio of ethylene to vinyl acetate units in the copolymer, thereby influencing the polymer's physical and chemical properties. By altering the chemical environment and catalyst structure, precise control over the copolymer's characteristics can be achieved. Therefore, accurate analysis of the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate is of great significance.

[0003] CN111048160B relates to a method for predicting catalyst performance. This method first requires constructing and optimizing a catalyst surface model and a molecular structure model involved in the reaction. Based on the molecular structure model involved in the reaction, the elementary reactions of the catalytic reaction are established. Then, based on the catalyst surface model and the molecular structure model involved in the reaction, an adsorption model is constructed and optimized to obtain stable adsorption models and adsorption energies of the molecules involved in the reaction at different adsorption sites on the catalyst surface model. Initial-state and final-state models of the elementary reactions are established, and a transition state model is searched and obtained. Based on the initial-state, transition-state, and final-state models, the reaction pathways and reaction energy barriers of the molecules involved in the catalytic reaction are simulated and determined. Combining the adsorption energy and reaction energy barriers of the molecules involved in the reaction, a reaction energy curve of the molecules involved in the catalytic reaction is plotted. Finally, electronic structure calculations and catalytic performance prediction are performed. This method predicts catalyst performance solely through quantum chemical calculations, which suffers from the drawbacks of high computational load and long processing time.

[0004] CN116741300A relates to a method for predicting the performance of metal-ligand catalysts in the trimerization or tetramerization of ethylene. This method first collects molecular data of several groups of metal-ligand catalysts as samples, including but not limited to catalyst performance. Then, descriptors are extracted from the collected sample data using computational chemistry methods. Next, the collected catalyst performance is used as an indicator, and the extracted descriptors are used as features. An optimal feature subset is selected using a feature selection method. Machine learning is then used to train the selected optimal feature subset. After successful training, a prediction model is obtained. Finally, the prediction model is used to predict the performance of new molecules to obtain the prediction results. This method is only applicable to the performance prediction of ethylene trimerization or tetramerization catalysts and cannot predict the performance of catalysts for the copolymerization of ethylene and vinyl acetate.

[0005] CN115424682A relates to a machine learning-based molecular design method for ethylene oligomerization catalysts. This method first extracts descriptors of the catalytic system as independent variables and experimental or computational oligomerization results as dependent variables, establishing a database of the catalytic system. Then, it uses the database to learn and establish a model relationship between the descriptors and the experimental or computational oligomerization results, obtaining a machine learning model. Based on the machine learning model, it predicts the target catalytic system and its ligands based on the target oligomerization results. This method is only applicable to predicting the performance of ethylene oligomerization catalysts and cannot predict the performance of ethylene-vinyl acetate copolymerization catalysts.

[0006] Therefore, developing an analytical method that can accurately predict the performance of palladium complexes in the copolymerization of ethylene and vinyl acetate is an urgent problem to be solved in this field. Summary of the Invention

[0007] To address the aforementioned technical problems, the present invention aims to provide an analytical method for the performance of palladium complexes in the copolymerization of ethylene and vinyl acetate. This method enables efficient screening of copolymerization catalysts and reduces research and development costs.

[0008] To achieve the above objectives, the present invention provides an analytical method for the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate, wherein the analytical method includes the following steps:

[0009] Step 1: Obtain the compositional and structural information of different palladium complex catalysts, wherein the palladium complex catalyst is composed of an auxiliary ligand and a palladium alkyl fragment:

[0010] Step 2: Simplify the structure of the palladium complex catalyst and calculate the reaction energy barrier of each palladium complex catalyst for the copolymerization of ethylene and vinyl acetate;

[0011] Step 3: Extract descriptors for the reaction center fragments of each palladium complex catalyst;

[0012] Step 4: Construct a reaction barrier prediction model for the copolymerization of ethylene and vinyl acetate based on the calculated reaction barrier and descriptor, and use it to analyze the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate.

[0013] In the above analysis method, preferably, step 2 includes:

[0014] Step 2.1: Construct models for ethylene monomer, vinyl acetate monomer, palladium complex catalyst, coordination complex of palladium complex catalyst and ethylene monomer, and coordination complex of palladium complex catalyst and vinyl acetate monomer;

[0015] Step 2.2: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of the catalyst and vinyl acetate monomer, respectively construct the transition state model of ethylene monomer insertion into palladium complex catalyst and the transition state model of vinyl acetate insertion into palladium complex catalyst.

[0016] Step 2.3: Based on the ethylene monomer model, vinyl acetate monomer model, palladium complex catalyst model, palladium complex catalyst and ethylene monomer coordination complex model, palladium complex catalyst and vinyl acetate monomer coordination complex model, ethylene monomer insertion into palladium complex catalyst transition state model, and vinyl acetate monomer insertion into palladium complex catalyst transition state model, calculate the reaction energy barrier of each palladium complex catalyst catalyzing the copolymerization of ethylene and vinyl acetate.

[0017] In the above analysis method, preferably, step 2.1 includes:

[0018] Step 2.1.1: Preliminary construction of 3D molecular structure models of ethylene monomer, vinyl acetate monomer, different palladium complex catalysts, coordination complexes of each palladium complex catalyst with ethylene monomer, and coordination complexes of each palladium complex catalyst with vinyl acetate monomer.

[0019] Step 2.1.2: Using density functional theory, geometric optimization and frequency analysis were performed on each of the initially constructed 3D molecular structure models to obtain the 3D molecular structure models of thermodynamically stable ethylene monomers, thermodynamically stable vinyl acetate monomers, thermodynamically stable palladium complex catalysts, thermodynamically stable complexes of palladium complex catalysts with ethylene monomers, and thermodynamically stable complexes of palladium complex catalysts with vinyl acetate monomers.

[0020] In the above analysis method, preferably, in step 2.1.1, the model construction process is performed in GaussView and the model file is exported in gjf format.

[0021] In the above analytical method, preferably, in step 2.1.2, the geometric optimization and frequency analysis calculations are performed in Gaussian 16 software. In the gas phase, a B3LYP functional with D3 dispersion correction is used for geometric optimization and frequency calculations. For C, H, O, N, P, S, F, and Cl atoms, a 6-31G(d) basis set is used, and for metal atoms, a LANL2DZ pseudopotential basis set is used. There are no symmetry restrictions during the optimization process, and the results are output in log file format. Applying the above method to the calculation of the energy barrier of the reaction between ethylene and vinyl acetate helps improve the accuracy of the calculation results, and the calculated values ​​agree better with the experimental values.

[0022] In the above analysis method, preferably, step 2.2 includes:

[0023] Step 2.2.1: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of the catalyst and vinyl acetate monomer, the transition state configuration of ethylene monomer inserted into each palladium complex catalyst and the transition state configuration of vinyl acetate monomer inserted into each palladium complex catalyst are preliminarily determined.

[0024] Step 2.2.2: Using density functional theory, the geometric structure optimization and frequency analysis of each initially determined transition state configuration are performed to obtain the transition state configuration of thermodynamically stable ethylene monomer inserted into palladium complex catalyst and the transition state configuration of thermodynamically stable vinyl acetate monomer inserted into palladium complex catalyst.

[0025] In the above analysis method, preferably, step 2.2.1 includes:

[0026] Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, as well as the coordination complex models of palladium complex catalyst and vinyl acetate monomer, the opt=TS method in GaussView software was used to preliminarily determine the transition state configurations of ethylene monomers inserted into the palladium complex catalyst and vinyl acetate monomers inserted into the palladium complex catalyst for each palladium complex catalyst, and the files were exported in gjf format.

[0027] In the above analysis method, preferably, in step 2.2.2:

[0028] The geometric optimization and frequency analysis of the initial transition state configuration were performed in Gaussian16 software. In the gas phase, the B3LYP functional with D3 dispersion correction was used for geometric optimization and frequency calculation. The 6-31G(d) basis set was used for C, H, O, N, P, S, F, and Cl atoms, and the LANL2DZ pseudopotential basis set was used for metal atoms. There were no symmetry restrictions during the optimization process, and the results were output in log file format.

[0029] In the process of using the above software, functions, etc., in order to achieve the evaluation / prediction of catalysts, the present invention made the following adjustments: (1) The established palladium metal catalyst structure model was simplified, and all substituents of auxiliary ligands that may bring complex conformations were replaced with methyl groups, thereby reducing the calculation error caused by conformation problems and improving the model prediction accuracy; (2) For the asymmetric structure of auxiliary ligand connecting atoms, cis and trans isomers were constructed, and the reaction energy barrier was calculated.

[0030] In the above analysis method, preferably, step 2.3 includes:

[0031] Step 2.3.1: Based on the ethylene monomer model, vinyl acetate monomer model, palladium complex catalyst model, palladium complex catalyst and ethylene monomer coordination complex model, palladium complex catalyst and vinyl acetate monomer coordination complex model, ethylene monomer insertion into palladium complex catalyst transition state model, and vinyl acetate monomer insertion into palladium complex catalyst transition state model, determine the Gibbs free energy of ethylene monomer, vinyl acetate monomer, each palladium complex catalyst, each palladium complex catalyst and ethylene monomer coordination complex, each palladium complex catalyst and vinyl acetate monomer coordination complex, the transition state of ethylene monomer insertion into each palladium complex catalyst, and the transition state of vinyl acetate monomer insertion into each palladium complex catalyst.

[0032] Step 2.3.2: For different palladium complex catalysts, using the sum of the Gibbs free energies of the ethylene monomer and the palladium complex catalyst as zero, determine the energy difference (first energy difference) of the Gibbs free energy of the trans coordination complex of the palladium complex catalyst and the ethylene monomer relative to zero; the energy difference (second energy difference) of the Gibbs free energy of the cis coordination complex of the palladium complex catalyst and the ethylene monomer relative to zero; the energy difference (third energy difference) of the Gibbs free energy of the ethylene monomer intercalating into the transition state of the palladium complex catalyst relative to zero; and the energy difference (fourth energy difference) of the Gibbs free energy of the hydroxyl coordination complex of the palladium complex catalyst and the vinyl acetate monomer relative to zero. The energy differences are as follows: the fifth energy difference is the Gibbs free energy of the palladium complex catalyst and the coordination complex of the vinyl acetate monomer relative to zero; the sixth energy difference is the Gibbs free energy of the transition state of the vinyl acetate monomer inserted into the palladium complex catalyst relative to zero; the seventh energy difference is the Gibbs free energy of the palladium complex catalyst after the vinyl acetate monomer insertion relative to zero; the eighth energy difference is the Gibbs free energy of the palladium complex catalyst after the vinyl acetate insertion relative to zero; and the ninth energy difference is the Gibbs free energy of the transition state of the palladium complex catalyst after the vinyl acetate insertion relative to zero.

[0033] Step 2.3.3: For different palladium complex catalysts, using the first, second, third, fourth, fifth, sixth, seventh, eighth, and ninth energy differences of the palladium complex catalyst, determine the isomerization energy between trans-coordinated and cis-coordinated palladium complexes, the activation energy of the first ethylene molecule inserted into the palladium complex, the activation energy of vinyl acetate inserted into the palladium complex, the reaction energy of vinyl acetate with the palladium complex, and the activation energy of the second ethylene molecule inserted into the palladium complex.

[0034] In the above analysis method, preferably, in step 2.3.1:

[0035] The Gibbs free energy of the ethylene monomer, the Gibbs free energy of the vinyl acetate monomer, the Gibbs free energy of each palladium complex catalyst, the Gibbs free energy of the coordination complex of each palladium complex catalyst and the ethylene monomer, the Gibbs free energy of the coordination complex of each palladium complex catalyst and the vinyl acetate monomer, the Gibbs free energy of the transition state of the ethylene monomer inserted into each palladium complex catalyst, and the Gibbs free energy of the transition state of the vinyl acetate monomer inserted into each palladium complex catalyst are read from the result files of steps 2.1 and 2.2.

[0036] In the above analytical method, preferably, in step 3, the descriptors include geometric descriptors, electron occupancy descriptors, and atomic charge descriptors. This invention extracts the geometric and electronic descriptors of the palladium alkyl fragment and correlates them with the reaction energy barrier to construct a predictive model, thereby achieving accurate prediction of the energy barrier of the copolymerization reaction between ethylene and vinyl acetate. This invention identifies the geometric and electronic descriptors of the palladium alkyl fragment as key descriptors affecting the copolymerization reaction energy barrier, and using them in model construction helps improve the model's prediction accuracy. Furthermore, it reduces the amount of data required for model building: since the aforementioned descriptors are key descriptors affecting the reaction energy barrier and have a strong correlation with it, the relationship between the two is quite obvious. Therefore, only a small amount of data is needed to build an accurate predictive model, reducing the amount of data required for modeling and lowering modeling costs.

[0037] In the above analysis method, preferably, step 4 includes:

[0038] Step 4.1: Perform data cleaning and preprocessing on the reaction energy barrier from Step 2 and the descriptor from Step 3;

[0039] Step 4.2: Construct energy barrier prediction models for the copolymerization reaction of ethylene and vinyl acetate using different machine learning algorithms, and select the optimal prediction model;

[0040] Step 4.3: Analyze the performance of palladium complex catalysts in the copolymerization of ethylene and vinyl acetate based on the optimal prediction model.

[0041] In the above analysis method, preferably, step 4.1 includes:

[0042] Step 4.1.1: Handle missing values ​​in the calculated reaction barrier data and descriptor data, fill in missing data, and delete data that cannot be filled.

[0043] Step 4.1.2: Perform outlier detection on the calculated reaction barrier data and descriptor data, and correct or delete the identified outliers;

[0044] Step 4.1.3: Screen key descriptor data that affect the copolymerization reaction energy barrier.

[0045] In the above analysis method, preferably, step 4.2 includes:

[0046] Step 4.2.1: Divide the reaction barrier calculation data and descriptor data into training set and test set;

[0047] Step 4.2.2: For different machine learning algorithms, use the training set divided in Step 4.2.1 to select the optimal hyperparameters suitable for each machine learning algorithm;

[0048] Step 4.2.3: Under the optimal hyperparameter settings, construct a prediction model for the energy barrier of the copolymerization reaction of ethylene and vinyl acetate using the training set partitioned in Step 4.2.1;

[0049] Step 4.2.4: Use the test set divided in Step 4.2.1 to test the prediction model constructed in Step 4.2.3 and select the optimal prediction model.

[0050] In the above analysis method, preferably, step 4.3 includes:

[0051] Step 4.3.1: Obtain the descriptor of the palladium complex catalyst to be tested;

[0052] Step 4.3.2: Input the descriptor obtained in step 4.3.1 into the prediction model constructed in step 4.2 to obtain the prediction results of the performance of the palladium complex catalyst in the copolymerization of ethylene and vinyl acetate.

[0053] To accurately predict the performance of the ethylene-vinyl acetate copolymerization catalyst, the present invention has made the following improvements: (1) The structure of the palladium metal catalyst is simplified, and the reaction energy barrier is calculated based on the simplified catalyst; specifically, all substituents of auxiliary ligands that may bring complex conformations are replaced with methyl groups, thereby reducing the calculation error caused by structure-activity relationship problems, improving the calculation accuracy, and laying the foundation for subsequent model establishment; (2) Descriptors are extracted from palladium alkyl fragments and correlated with the reaction energy barrier to construct a prediction model; since the auxiliary ligands of different palladium metal catalysts are quite different, it is difficult to determine the type of descriptor to be extracted uniformly. Therefore, descriptors are extracted from palladium alkyl fragments and correlated with the reaction energy barrier. Key descriptors affecting catalyst performance are designed and screened; (3) When the atoms connected to palladium metal in the auxiliary ligand are asymmetric, the connecting atoms may exchange positions and generate two connecting atoms that can exchange their coordination positions, and generate trans / cis isomer reaction energy barriers. Therefore, trans / cis isomers are used as training samples respectively, and the trans / cis isomer reaction energy barrier is used as one of the model outputs. When the auxiliary ligand is symmetric, the reaction energy barrier value is set to 0.

[0054] This invention also provides an analytical system for evaluating the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate, wherein the analytical system comprises:

[0055] Acquisition module: used to acquire the compositional and structural information of different palladium complex catalysts, wherein the palladium complex catalyst is composed of auxiliary ligands and palladium alkyl fragments.

[0056] Calculation module: used to simplify the structure of palladium complex catalysts and calculate the reaction energy barrier of each palladium complex catalyst for the copolymerization of ethylene and vinyl acetate;

[0057] Extraction module: Used to extract descriptors for the reaction center fragments of each palladium complex catalyst;

[0058] Analysis module: Used to construct a predictive model of the reaction energy barrier for the copolymerization of ethylene and vinyl acetate based on calculated reaction energy barriers and descriptors, and used to analyze the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate.

[0059] The method provided by this invention can achieve the following technical effects:

[0060] The method of this invention helps to accelerate the screening of copolymerization catalysts and reduce research and development costs. This method constructs a predictive model of the energy barrier for the copolymerization reaction of ethylene and vinyl acetate by combining quantum chemical calculations and machine learning. This allows for the rapid acquisition of copolymerization performance data for different copolymerization catalysts, thereby saving significant human and material resources and accelerating the research and development process of copolymerization catalysts.

[0061] The method of this invention facilitates the rational design of copolymerization catalysts and improves the success rate of copolymerization catalyst design. During the modeling process, this invention screens descriptors that affect the catalytic performance of copolymerization catalysts, which helps to clarify the microscopic factors affecting copolymerization performance based on the chemical meaning of the descriptors, thus facilitating the rational design of new catalyst molecules. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the structure of a palladium complex catalyst.

[0063] Figure 2 This is a schematic diagram illustrating the general mechanism of the copolymerization of ethylene and vinyl acetate catalyzed by palladium complex catalysts.

[0064] Figure 3 The graph shows the reaction energy of the copolymerization of ethylene and vinyl acetate.

[0065] Figure 4 A simplified schematic diagram of the auxiliary ligand.

[0066] Figure 5 This is a schematic diagram of the trans and cis isomers.

[0067] Figure 6 The test results of the copolymerization prediction model constructed for Example 1 are shown in the figure. Detailed Implementation

[0068] In order to provide a clearer understanding of the technical features, objectives and beneficial effects of the present invention, the technical solution of the present invention will now be described in detail below, but it should not be construed as limiting the scope of implementation of the present invention.

[0069] Example 1

[0070] This embodiment provides an analytical method for the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate, specifically including the following steps:

[0071] Step 1: Obtain the compositional and structural information of different palladium complex catalysts, wherein the palladium complex catalysts are composed of auxiliary ligands and allyl palladium fragments, such as... Figure 1 As shown; this embodiment collected three-dimensional structural information of 331 different palladium complex catalysts; the general mechanism of palladium complex catalysts catalyzing the copolymerization of ethylene and vinyl acetate is as follows. Figure 2 As shown;

[0072] Step 2: Simplify the structure of the palladium complex (e.g.) Figure 4 (As shown), calculate the energy barrier values ​​for the copolymerization reaction of ethylene and vinyl acetate catalyzed by different palladium complex catalysts, specifically including:

[0073] Step 2.1: Use GaussView to construct models of ethylene monomer, vinyl acetate monomer, palladium complex catalyst, coordination complex of palladium complex catalyst and ethylene monomer, and coordination complex of palladium complex catalyst and vinyl acetate monomer, and export the model files in gjf format.

[0074] Then, using Gaussian 16 software, the geometric structure optimization and frequency analysis of the initially constructed 3D molecular structure models were performed to obtain the 3D molecular structure models of thermodynamically stable ethylene monomers (i.e., the ethylene monomer model), thermodynamically stable vinyl acetate monomers (i.e., the vinyl acetate monomer model), thermodynamically stable palladium complex catalysts (i.e., the palladium complex catalyst model), thermodynamically stable coordination complexes of palladium complex catalysts and ethylene monomers (i.e., the coordination complexes of palladium complex catalysts and vinyl acetate monomers), and thermodynamically stable coordination complexes of palladium complex catalysts and vinyl acetate monomers (i.e., the coordination complexes of palladium complex catalysts and vinyl acetate monomers), which were output in log file format.

[0075] During the geometric structure optimization and frequency analysis, the theoretical calculation method used was the B3LYP functional with D3 dispersion correction; the 6-31G(d) basis set was used for C, H, O, N, P, S, F, and Cl atoms, and the LANL2DZ pseudopotential basis set was used for metal atoms (palladium and nickel). There were no symmetry restrictions during the optimization process, and the molecular structure model was determined to be thermodynamically stable when all frequency values ​​were positive.

[0076] Step 2.2: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of the catalyst and vinyl acetate monomer, the opt=TS method in GaussView software is used to preliminarily determine the transition state model of ethylene monomer inserted into palladium complex catalyst and the transition state model of vinyl acetate inserted into palladium complex catalyst corresponding to each palladium complex catalyst, and export the file in gjf format.

[0077] The initial transition state configuration geometry was optimized and frequency analyzed using Gaussian16 software. The theoretical calculation method used in the geometry optimization and frequency analysis was the B3LYP functional with D3 dispersion correction. The 6-31G(d) basis set was used for C, H, O, N, P, S, F, and Cl atoms, and the LANL2DZ pseudopotential basis set was used for metal atoms (palladium and nickel). No symmetry restrictions were imposed during the optimization process. The transition state configurations of thermodynamically stable ethylene monomers inserted into palladium complex catalysts (i.e., the transition state models of ethylene monomers inserted into palladium complex catalysts) and thermodynamically stable vinyl acetate monomers inserted into palladium complex catalysts (i.e., the transition state models of vinyl acetate monomers inserted into palladium complex catalysts) were obtained, and the results were output in log file format.

[0078] Step 2.3: Determine the Gibbs free energy of ethylene monomer, vinyl acetate monomer, each palladium complex catalyst, the Gibbs free energy of the coordination complex of each palladium complex catalyst and ethylene monomer, the Gibbs free energy of the coordination complex of each palladium complex catalyst and vinyl acetate monomer, the Gibbs free energy of the transition state of ethylene monomer inserted into each palladium complex catalyst, and the Gibbs free energy of the transition state of vinyl acetate monomer inserted into each palladium complex catalyst from the log files output in Steps 2.1 and 2.2.

[0079] For different palladium complex catalysts, the energy difference between the Gibbs free energy of the ethylene monomer and the palladium complex catalyst and the zero point is used to determine the energy difference between the trans coordination complex of the palladium complex catalyst and the ethylene monomer relative to the zero point, i.e., the first energy difference; the energy difference between the Gibbs free energy of the cis coordination complex of the palladium complex catalyst and the ethylene monomer relative to the zero point, i.e., the second energy difference; the energy difference between the Gibbs free energy of the ethylene monomer inserted into the transition state of the palladium complex catalyst and the zero point, i.e., the third energy difference; the energy difference between the Gibbs free energy of the hydroxyl coordination complex of the palladium complex catalyst and the vinyl acetate monomer relative to the zero point, i.e., the fourth energy difference; the energy difference between the Gibbs free energy of the coordination complex of the palladium complex catalyst and the vinyl acetate monomer relative to the zero point, i.e., the fifth energy difference; and the energy difference between the hydroxyl coordination complex of the palladium complex catalyst and the vinyl acetate monomer relative to the zero point. The energy differences relative to zero of the Gibbs free energy of the transition state of the palladium complex catalyst (i.e., the sixth energy difference), the energy differences relative to zero of the Gibbs free energy of the palladium complex catalyst after vinyl acetate monomer insertion (i.e., the seventh energy difference), the energy differences relative to zero of the Gibbs free energy of the palladium complex catalyst after vinyl acetate insertion and the coordination complex of the palladium complex catalyst with the ethylene monomer (i.e., the eighth energy difference), and the energy differences relative to zero of the transition state of the palladium complex catalyst after vinyl acetate insertion (i.e., the ninth energy difference) are used to determine the isomerization energy between the trans-coordinated and cis-coordinated palladium complexes (see the first, second, third, fourth, fifth, sixth, seventh, eighth, and ninth energy differences). Figure 5 The activation energies are: the activation energy of the first ethylene molecule inserted into the palladium complex, the activation energy of the vinyl acetate ester inserted into the palladium complex, the reaction energy of vinyl acetate with the palladium complex, and the activation energy of the second ethylene molecule inserted into the palladium complex. Table 1 below shows the calculated reaction energy barriers for some of the 331 catalysts, and the reaction energy of the ethylene-vinyl acetate copolymerization unit is as follows: Figure 3 As shown;

[0080] Table 1 Partial Reaction Energy Barriers

[0081]

[0082] Step 3: Extract descriptors for the reaction center fragments of different palladium complex catalysts; determine the descriptor values ​​based on each palladium complex catalyst model, specifically including (descriptor values ​​of some of the 331 collected catalysts are shown in Table 2):

[0083] (1) Geometric structural parameters: the distance from the C atom in the alkyl group (R) directly connected to the central metal in the catalyst molecule to the central metal, the distance between adjacent C atoms in the alkyl group R, and the angle between the central metal and adjacent C atoms;

[0084] (2) Electronic structure parameters: Mulliken charge of each atom in the reaction center.

[0085] Table 2 Partial Catalyst Descriptor Values

[0086]

[0087]

[0088] Step 4: Construct a reaction barrier prediction model for the copolymerization of ethylene and vinyl acetate based on the calculated reaction barrier and extracted descriptors, and use it to analyze the performance of palladium complex catalysts in the copolymerization of ethylene and vinyl acetate.

[0089] This embodiment uses the kernel ridge regression machine learning algorithm to construct a predictive model for the energy barrier of the copolymerization reaction of ethylene and vinyl acetate.

[0090] The model's predictive performance was evaluated using 10-fold cross-validation, and the results are as follows: Figure 6 As shown, the model's mean absolute error on the validation set is 0.77 kcal / mol.

[0091] Furthermore, this embodiment used eight experimentally reported catalyst structures (as shown in Table 3) to perform external validation of the model.

[0092] Table 3 External validation catalyst auxiliary ligand structure

[0093]

[0094] Through testing, the model's average absolute error on the test data was only 1.02 kcal / mol, which can accurately predict the catalytic performance of the catalyst.

Claims

1. An analytical method for the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate, wherein, The analytical method includes the following steps: Step 1: Obtain the compositional and structural information of different palladium complex catalysts, wherein the palladium complex catalyst is composed of an auxiliary ligand and a palladium alkyl fragment: Step 2: Simplify the structure of the palladium complex catalyst and calculate the reaction energy barrier of each palladium complex catalyst for the copolymerization of ethylene and vinyl acetate; Step 3: Extract descriptors for the reaction center fragments of each palladium complex catalyst; Step 4: Construct a reaction barrier prediction model for the copolymerization of ethylene and vinyl acetate based on the calculated reaction barrier and descriptor, and use it to analyze the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate.

2. The analytical method according to claim 1, wherein, Step 2 includes: Step 2.1: Construct models for ethylene monomer, vinyl acetate monomer, palladium complex catalyst, coordination complex of palladium complex catalyst and ethylene monomer, and coordination complex of palladium complex catalyst and vinyl acetate monomer; Step 2.2: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of palladium complex catalyst and vinyl acetate monomer, construct the transition state models of ethylene monomer insertion into palladium complex catalyst and vinyl acetate insertion into palladium complex catalyst, respectively. Step 2.3: Based on the ethylene monomer model, vinyl acetate monomer model, palladium complex catalyst model, palladium complex catalyst and ethylene monomer coordination complex model, palladium complex catalyst and vinyl acetate monomer coordination complex model, ethylene monomer insertion into palladium complex catalyst transition state model, and vinyl acetate monomer insertion into palladium complex catalyst transition state model, calculate the reaction energy barrier of each palladium complex catalyst catalyzing the copolymerization of ethylene and vinyl acetate.

3. The analytical method according to claim 2, wherein, Step 2.1 includes: Step 2.1.1: Preliminary construction of 3D molecular structure models of ethylene monomer, vinyl acetate monomer, palladium complex catalysts, coordination complexes of palladium complex catalysts and ethylene monomer, and coordination complexes of palladium complex catalysts and vinyl acetate monomer. Step 2.1.2: Using density functional theory, geometric optimization and frequency analysis were performed on each of the initially constructed 3D molecular structure models to obtain the 3D molecular structure models of thermodynamically stable ethylene monomers, thermodynamically stable vinyl acetate monomers, thermodynamically stable palladium complex catalysts, thermodynamically stable complexes of palladium complex catalysts with ethylene monomers, and thermodynamically stable complexes of palladium complex catalysts with vinyl acetate monomers.

4. The analytical method according to claim 3, wherein, In step 2.1.1, the model building process is performed in GaussView, and the model file is exported in gjf format.

5. The analytical method according to claim 3, wherein, In step 2.1.2, the geometry optimization and frequency analysis calculations are performed in Gaussian16 software. In the gas phase, the B3LYP functional with D3 dispersion correction is used for geometry optimization and frequency calculation. The 6-31G(d) basis set is used for C, H, O, N, P, S, F, and Cl atoms, and the LANL2DZ pseudopotential basis set is used for metal atoms. There are no symmetry restrictions during the optimization process, and the results are output in log file format.

6. The analytical method according to claim 1, wherein, Step 2.2 includes: Step 2.2.1: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of palladium complex catalyst and vinyl acetate monomer, the transition state configuration of ethylene monomer inserted into each palladium complex catalyst and the transition state configuration of vinyl acetate monomer inserted into each palladium complex catalyst are preliminarily determined. Step 2.2.2: Using density functional theory, the geometric structure optimization and frequency analysis of each initially determined transition state configuration are performed to obtain the transition state configuration of thermodynamically stable ethylene monomer inserted into palladium complex catalyst and the transition state configuration of thermodynamically stable vinyl acetate monomer inserted into palladium complex catalyst.

7. The analytical method according to claim 6, wherein, Step 2.2.1 includes: Based on the coordination complex models of each palladium complex catalyst and ethylene monomer, and the coordination complex model of palladium complex catalyst and vinyl acetate monomer, the opt=TS method in GaussView software was used to preliminarily determine the transition state configuration of ethylene monomer inserted into the palladium complex catalyst and the transition state configuration of vinyl acetate monomer inserted into the palladium complex catalyst for each palladium complex catalyst, and the files were exported in gjf format.

8. The analytical method according to claim 6, wherein, In step 2.2.2: The geometric optimization and frequency analysis of the initial transition state configuration were performed in Gaussian16 software. In the gas phase, the B3LYP functional with D3 dispersion correction was used for geometric optimization and frequency calculation. The 6-31G(d) basis set was used for C, H, O, N, P, S, F, and Cl atoms, and the LANL2DZ pseudopotential basis set was used for metal atoms. There were no symmetry restrictions during the optimization process, and the results were output in log file format.

9. The analytical method according to claim 1, wherein, Step 2.3 includes: Step 2.3.1: Based on the ethylene monomer model, vinyl acetate monomer model, palladium complex catalyst model, palladium complex catalyst and ethylene monomer coordination complex model, palladium complex catalyst and vinyl acetate monomer coordination complex model, ethylene monomer insertion into palladium complex catalyst transition state model, and vinyl acetate monomer insertion into palladium complex catalyst transition state model, determine the Gibbs free energy of ethylene monomer, vinyl acetate monomer, each palladium complex catalyst, each palladium complex catalyst and ethylene monomer coordination complex, each palladium complex catalyst and vinyl acetate monomer coordination complex, the transition state of ethylene monomer insertion into each palladium complex catalyst, and the transition state of vinyl acetate monomer insertion into each palladium complex catalyst. Step 2.3.2: For different palladium complex catalysts, using the sum of the Gibbs free energies of the ethylene monomer and the palladium complex catalyst as zero, determine the energy difference (first energy difference) of the Gibbs free energy of the trans coordination complex of the palladium complex catalyst and the ethylene monomer relative to zero; the energy difference (second energy difference) of the Gibbs free energy of the cis coordination complex of the palladium complex catalyst and the ethylene monomer relative to zero; the energy difference (third energy difference) of the Gibbs free energy of the ethylene monomer intercalating into the transition state of the palladium complex catalyst relative to zero; and the energy difference (fourth energy difference) of the Gibbs free energy of the hydroxyl coordination complex of the palladium complex catalyst and the vinyl acetate monomer relative to zero. The energy differences are as follows: the fifth energy difference is the Gibbs free energy of the palladium complex catalyst and the coordination complex of the vinyl acetate monomer relative to zero; the sixth energy difference is the Gibbs free energy of the transition state of the vinyl acetate monomer inserted into the palladium complex catalyst relative to zero; the seventh energy difference is the Gibbs free energy of the palladium complex catalyst after the vinyl acetate monomer insertion relative to zero; the eighth energy difference is the Gibbs free energy of the palladium complex catalyst after the vinyl acetate insertion relative to zero; and the ninth energy difference is the Gibbs free energy of the transition state of the palladium complex catalyst after the vinyl acetate insertion relative to zero. Step 2.3.3: For different palladium complex catalysts, using the first, second, third, fourth, fifth, sixth, seventh, eighth, and ninth energy differences of the palladium complex catalyst, determine the isomerization energy between trans-coordinated and cis-coordinated palladium complexes, the activation energy of the first ethylene molecule inserted into the palladium complex, the activation energy of vinyl acetate inserted into the palladium complex, the reaction energy of vinyl acetate with the palladium complex, and the activation energy of the second ethylene molecule inserted into the palladium complex.

10. The analytical method according to claim 9, wherein, In step 2.3.1: The Gibbs free energy of the ethylene monomer, the Gibbs free energy of the vinyl acetate monomer, the Gibbs free energy of each palladium complex catalyst, the Gibbs free energy of the coordination complex of each palladium complex catalyst and the ethylene monomer, the Gibbs free energy of the coordination complex of each palladium complex catalyst and the vinyl acetate monomer, the Gibbs free energy of the transition state of the ethylene monomer inserted into each palladium complex catalyst, and the Gibbs free energy of the transition state of the vinyl acetate monomer inserted into each palladium complex catalyst are read from the result files of steps 2.1 and 2.

2.

11. The analytical method according to claim 1, wherein, In step 3, the descriptor includes a geometric descriptor, an electron occupancy descriptor, and an atomic charge descriptor.

12. The analytical method according to claim 1, wherein, Step 4 includes: Step 4.1: Perform data cleaning and preprocessing on the reaction energy barrier from Step 2 and the descriptor from Step 3; Step 4.2: Construct energy barrier prediction models for the copolymerization reaction of ethylene and vinyl acetate using different machine learning algorithms, and select the optimal prediction model; Step 4.3: Analyze the performance of palladium complexes in the copolymerization of ethylene and vinyl acetate based on the optimal prediction model.

13. The analytical method according to claim 12, wherein, Step 4.1 includes: Step 4.1.1: Handle missing values ​​in the calculated reaction barrier data and descriptor data, fill in missing data, and delete data that cannot be filled. Step 4.1.2: Perform outlier detection on the calculated reaction barrier data and descriptor data, and correct or delete the identified outliers; Step 4.1.3: Screen key descriptor data that affect the copolymerization reaction energy barrier.

14. The analytical method according to claim 12, wherein, Step 4.2 includes: Step 4.2.1: Divide the reaction barrier calculation data and descriptor data into training set and test set; Step 4.2.2: For different machine learning algorithms, use the training set divided in Step 4.2.1 to select the optimal hyperparameters suitable for each machine learning algorithm; Step 4.2.3: Under the optimal hyperparameter settings, construct a prediction model for the energy barrier of the copolymerization reaction of ethylene and vinyl acetate using the training set partitioned in Step 4.2.1; Step 4.2.4: Use the test set divided in Step 4.2.1 to test the prediction model constructed in Step 4.2.3 and select the optimal prediction model.

15. The analytical method according to claim 12 or 14, wherein, Step 4.3 includes: Step 4.3.1: Obtain the descriptor of the palladium complex catalyst to be tested; Step 4.3.2: Input the descriptor obtained in step 4.3.1 into the prediction model constructed in step 4.2 to obtain the prediction results of the performance of the palladium complex catalyst in the copolymerization of ethylene and vinyl acetate.

16. An analytical system for the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate, wherein, The analysis system includes: Acquisition module: used to acquire the compositional and structural information of different palladium complex catalysts, wherein the palladium complex catalyst is composed of auxiliary ligands and palladium alkyl fragments. Calculation module: used to simplify the structure of palladium complex catalysts and calculate the reaction energy barrier of each palladium complex catalyst for the copolymerization of ethylene and vinyl acetate; Extraction module: Used to extract descriptors for the reaction center fragments of each palladium complex catalyst; Analysis module: Used to construct a predictive model of the reaction energy barrier for the copolymerization of ethylene and vinyl acetate based on calculated reaction energy barriers and descriptors, and used to analyze the performance of palladium complexes in catalyzing the copolymerization of ethylene and vinyl acetate.

Images