Prediction method for thermal stability of coordination polymerization organic metal catalyst

Abstract

The invention provides a prediction method for thermal stability of a coordination polymerization organic metal catalyst. The method comprises the following steps: constructing a model of a catalyst active species, an ethylene monomer, a catalyst ligand and a bimolecular ethylene coordinated complex, carrying out geometric structure optimization and frequency analysis to obtain a thermodynamically stable 3D structure, and collecting and sorting corresponding stable molecular structures and thermodynamic correction values of enthalpy values and Gibbs free energy of the stable molecular structures; obtaining corresponding single-point energy, enthalpy value and Gibbs free energy of each molecular structure, bond dissociation enthalpy and bond dissociation energy; quantum chemical descriptors are calculated; combining the quantum chemical descriptors, the key dissociation enthalpy and the key dissociation energy to construct a database; and adopting multiple linear stepwise regression analysis to remove multiple collinearity of the descriptors, and constructing a prediction model. According to the prediction method provided by the invention, the thermal stability of the coordination polymerization organic metal catalyst can be accurately predicted, so that the workload is reduced, and efficient screening of the polymerization catalyst is realized.

Description

Technical Field

[0001] This invention relates to a method for predicting the thermal stability of coordination polymerization organometallic catalysts, belonging to the field of polymer synthesis technology. Background Technology

[0002] The thermal stability of a catalyst is of great significance for maintaining its catalytic activity, durability, and safety. Therefore, accurately determining the thermal stability of a catalyst is crucial when designing and selecting it.

[0003] Wenhong Yang et al. (Wenhong Yang, Zhifeng Ma, Jun Yi, Wen-Hua Sun; Quantitative Structure-Thermostability Relationship of Late Transition Metal Catalysts in Ethylene Oligo / Polymerization; Catalysts 2017, 7, 120) investigated the relationship between the quantitative structure and thermal stability of catalysts. Specifically, based on experimentally measured catalytic activity data of 15 diimine pyridine complexes and 5 alpa-diimine nickel complexes for olefin polymerization, the thermal stability of the catalysts was represented by calculating the optimal temperature and the rate of change of catalytic activity at room temperature. Three molecular descriptors for the corresponding catalysts were calculated using quantum chemical methods. Twenty samples were divided into four groups of five samples each to construct a small dataset. Multiple linear regression analysis was used to construct predictive models for the thermal stability of the catalysts. This technology uses only 5 samples when constructing each catalyst thermal stability prediction model, which is too few samples. Therefore, the prediction model has poor universality and very limited application scope. Secondly, the prediction model depends on experimental data, and data from different sources will also lead to poor performance of the prediction model.

[0004] Narges Zohari et al. (Narges Zohari, Fatemeh Abrishami, Vida Zeynali; Prediction of decomposition temperature of azole-based energetic compounds in order to assess their thermal stability; Journal of Thermal Analysis and Calorimetry 2020, 141, 1453–1463) used experimentally measured thermal decomposition temperatures as a standard for assessing thermal stability. They calculated molecular descriptors for 115 azole-based energetic compounds, constructed a training set, and built a thermal stability prediction model (R²) through multiple linear regression analysis. 2 =0.936, RMSE=16.25K), tested using internal and external validation (Q). 2 LMO =0.937,Q 2 Ext =0.92), indicating that the model has good predictive ability. The samples involved in this technique are azole energetic compounds, rather than organometallic complexes; in addition, the construction of this predictive model also depends on experimental data, and data from different sources can lead to poor performance of the predictive model.

[0005] CN102980972A (application number 201910269693.7) discloses a method for predicting the thermal stability of organic compounds. Specifically, it involves identifying organic compounds in a mixture sample using a detection and identification device, separating the organic compound components in the mixture sample by gas chromatography, identifying each component by Raman spectroscopy, and predicting various thermal stability parameters using a thermal stability parameter prediction module. This method requires the synthesis of organic compounds and the use of instruments to test thermal stability parameters, making it entirely dependent on analytical testing methods.

[0006] As can be seen from the above, there is currently a lack of methods to accurately predict the thermal stability of coordination polymerization organometallic catalysts. Summary of the Invention

[0007] To address the aforementioned technical problems, the present invention aims to provide a method for predicting the thermal stability of coordination polymerization organometallic catalysts, which, with the aid of quantum chemistry, enables accurate prediction of the thermal stability of coordination polymerization organometallic catalysts.

[0008] To achieve the above objectives, the present invention provides a method for predicting the thermal stability of coordination polymerization organometallic catalysts, comprising the following steps:

[0009] Step 1: Construct models of catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene-coordinated complex D. Perform geometric structure optimization and frequency analysis on the models to obtain their respective thermodynamically stable 3D structures. Collect and organize the corresponding thermodynamically stable 3D structures and their enthalpy values ​​and thermodynamic correction values ​​of Gibbs free energy.

[0010] Step 2: Based on the thermodynamically stable 3D structure, obtain the corresponding single-point energy;

[0011] Step 3: Based on the thermodynamically stable 3D structure, enthalpy, thermodynamic correction value of Gibbs free energy, and single-point energy, obtain the enthalpy and Gibbs free energy of each thermodynamically stable 3D structure, i.e., the enthalpy H of molecules A, B, C, and D. A H B H C H D Gibbs free energy G A G B G C G D ;

[0012] Step 4: Based on the enthalpy and Gibbs free energy of each thermodynamically stable 3D structure, obtain the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG);

[0013] Step 5: Based on the molecular structure of the catalyst active species A optimized in Step 1, calculate the corresponding quantum chemical descriptor;

[0014] Step 6: Combine the quantum chemical descriptors, bond dissociation enthalpy, and bond dissociation energy to construct a database;

[0015] Step 7: Based on the database, use multiple linear stepwise regression analysis to remove multicollinearity of descriptors and construct a prediction model;

[0016] Step 8: Using the prediction model constructed in Step 7, perform leave-one-out validation on the data in Step 6, record the root mean square error (RMSE) of the leave-one-out method on the training set and the test set, and then evaluate its prediction performance.

[0017] In the above prediction method, preferably, in step 1, the geometric structure optimization and frequency analysis of the above models are performed by density functional theory at calculation level 1 to obtain their respective thermodynamically stable 3D structures, and the corresponding thermodynamically stable 3D structures and their enthalpy values ​​and thermodynamic correction values ​​of Gibbs free energy are collected and organized.

[0018] In the above prediction method, preferably, the catalyst is a catalyst having an [N,N] skeleton and an aromatic ring.

[0019] In the above prediction method, preferably, step 1 includes:

[0020] Step 1.1: Use nickel or palladium as the central metal of the active species A of the catalyst;

[0021] Step 1.2: By changing the substituents R on the [N,N] backbone and the aromatic ring Ph in the catalyst molecule structure (e.g., Figure 2 The OMe shown t Bu, F, CF3 i Models were constructed using Pr, CHPh2, Et, etc., to represent different active catalyst species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene-coordinated complex D.

[0022] Step 1.3: Use GaussView software to create all the three-dimensional molecular structure models from Step 1.2, and export the model files in gjf format;

[0023] Step 1.4: Use Gaussian 16 to perform geometric optimization and frequency analysis on all the three-dimensional molecular structure models constructed in Step 1.3, and output a log file; the optimization tends to the lowest energy; there are no symmetry restrictions during the optimization process; if all frequency values ​​are positive, the molecule is determined to be a stable molecular conformation, that is, a thermodynamically stable 3D structure.

[0024] Step 1.5: Read the enthalpy and Gibbs free energy correction values ​​of different catalyst active species A, catalyst ligand C, and bimolecular ethylene-coordinated complex D from the output log file.

[0025] In the above prediction method, preferably, step 2 includes:

[0026] Step 2.1: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at a high level 2, collect the single-point energy of the corresponding thermodynamically stable 3D structure, and output a log file;

[0027] Step 2.2: Read the single-point energy from the log file output in Step 2.1.

[0028] In the above prediction method, preferably, in step 5, the quantum chemical descriptor includes: dipole moment, fourth-order moment, E... LUMO E HOMO %V bur q M d M-C d M-N ;

[0029] in:

[0030] Dipole moment: D=(X 2 +Y 2 +Z 2 ) 1 / 2 ;

[0031] Fourth-order moment: Q = (X) 2 +Y 2 +Z 2 ) 1 / 2 ;

[0032] %V bur The embedding volume;

[0033] E LUMO E HOMO For orbital energy levels;

[0034] q M The Mulliken charge at the metal center;

[0035] d M-C denoted as , where is the distance between the center of metal M and the nearest C atom;

[0036] d M -N is the minimum distance between the center of the metal M and the two N atoms attached to it.

[0037] In the above prediction method, preferably, the embedding volume of the catalyst active species A is calculated using SambVca2.1 software.

[0038] In the above prediction method, preferably, in step 6, combining the quantum chemical descriptor, bond dissociation enthalpy, and bond dissociation energy means: using the quantum chemical descriptor as the independent variable, and the bond dissociation enthalpy and bond dissociation energy as two dependent variables, and putting the three together to form a database, wherein the quantum chemical descriptor establishes a relationship with the bond dissociation enthalpy and bond dissociation energy respectively.

[0039] In the above prediction method, preferably, in step 7, when removing multicollinearity of descriptors, only descriptors with p-value < 0.05 are retained.

[0040] In the above prediction method, preferably, in step 8, the prediction performance is evaluated in the following manner: a prediction model is obtained based on the data of the training set, and the prediction model is used to predict on the test set: the catalyst structure on the test set is optimized, and the descriptor value is calculated. The value is substituted into the prediction model relation to obtain the predicted bond dissociation enthalpy and bond dissociation energy.

[0041] The correlation coefficient R between the predicted bond dissociation enthalpy and bond dissociation energy and the calculated bond dissociation enthalpy and bond dissociation energy was calculated. 2 The root mean square error (RMSE) is used to evaluate the predictive performance of the prediction model.

[0042] In the above prediction method, preferably, the prediction model is considered to be able to accurately predict the thermal stability of the catalyst when the following condition is met: the correlation coefficient R between the predicted bond dissociation enthalpy and bond dissociation energy and the calculated bond dissociation enthalpy and bond dissociation energy. 2 Greater than 0.85, and the root square error is less than 2.0 kcal / mol (DFT calculation error).

[0043] In the above prediction method, preferably, the thermal stability of the catalyst is predicted in the following manner based on the ability to accurately predict the thermal stability of the catalyst:

[0044] After optimizing the structure of the catalyst under test, the descriptor value is calculated. The value is then substituted into the prediction model to obtain the bond dissociation enthalpy and bond dissociation energy. The larger the values ​​of the bond dissociation enthalpy and bond dissociation energy, the better its thermal stability, and vice versa. Based on this, the performance of the catalyst under test is evaluated.

[0045] According to a specific embodiment of the present invention, preferably, the above prediction method includes the following specific steps:

[0046] Step 1.1: Use nickel or palladium as the central metal of the active species A of the catalyst;

[0047] Step 1.2: By changing the substituents R on the [N,N] backbone and the aromatic ring Ph in the catalyst molecule structure (e.g., Figure 2 The OMe shown t Bu, F, CF3 i Models were constructed for different catalysts, including active species A, catalyst ligand C, and bimolecular ethylene-coordinated complex D, using Pr, CHPh2, Et, etc.

[0048] Step 1.3: Use GaussView software to create all the three-dimensional molecular structure models from Step 1.2, and export the model files in gjf format;

[0049] Step 1.4: Use Gaussian 16 to perform geometric optimization and frequency analysis on all the three-dimensional molecular structure models constructed in Step 1.3, and output a log file; the optimization tends to the lowest energy; there are no symmetry restrictions during the optimization process; if all frequency values ​​are positive, the molecule is determined to be a stable molecular conformation, that is, a thermodynamically stable 3D structure.

[0050] Step 1.5: Read the enthalpy and Gibbs free energy correction values ​​of different catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene coordinated complex D from the output log file and save them in an Excel spreadsheet.

[0051] Step 2.1: Based on the stable molecular structure (thermodynamically stable 3D structure) obtained in Step 1, perform single-point calculations at a high level 2, collect the corresponding single-point energies of the molecules, and output the log file.

[0052] Step 2.2: Read the single-point energy from the log file output in Step 2.1 and save it in an Excel spreadsheet;

[0053] Step 3: Based on the statistical data from steps 1.5 and 2.2, calculate the enthalpy (H) of the molecules (A, B, C, D) respectively. A H B H C H D Gibbs free energy (G) A G B G C G D );

[0054] Step 4: Based on the enthalpy and Gibbs free energy of each molecule obtained in Step 3, calculate the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG);

[0055] Step 5: Based on the optimized molecular structure of catalyst active species A in Step 1.4, calculate the corresponding quantum chemical descriptors. The quantum chemical descriptors of catalyst active species A are: dipole moment, fourth-order moment, and E. LUMO E HOMO ,%V bur q M d M-C d M-N Wherein, the dipole moment: D=(X 2 +Y 2 +Z 2 ) 1 / 2 Fourth-order moment: Q = (X 2 +Y 2 +Z 2 ) 1 / 2 %V bur E represents the embedding volume. LUMO E HOMO For orbital energy levels; q M Mulliken charge at the metal center; d M-C The distance between the center of metal M and the nearest C atom (e.g.) Figure 3 C); d M-N It is the minimum distance between the center of metal M and the two N atoms it is connected to;

[0056] The embedding volume (V) of catalyst active species A bur The calculations were performed using SambVca 2.1 software;

[0057] Step 6: Combine the quantum chemical descriptor calculated in Step 5.1 with the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG) calculated in Step 4 to construct a database;

[0058] Step 7: Using the database constructed in Step 6, perform multiple linear stepwise regression analysis to remove multicollinearity of descriptors (only retaining descriptors with p-value < 0.05) and construct a model to obtain the prediction model.

[0059] Step 8: Using the prediction model trained in Step 7, perform leave-one-out validation on the data in Step 6, record the root mean square error (RMSE) of the leave-one-out method on the training set and the test set, and then evaluate its prediction performance.

[0060] The method for predicting the thermal stability of coordination polymerization organometallic catalysts provided by this invention can accurately predict the thermal stability of coordination polymerization organometallic catalysts, thereby reducing the workload of theoretical and experimental research, shortening the research and development cycle, and achieving efficient screening of polymerization catalysts.

[0061] The technical solution of the present invention can bring the following beneficial effects:

[0062] Beneficial Effect 1: The technical solution of this invention allows for the creation of different catalyst models through simple computer experiments by altering the catalyst skeleton and substituents on the aromatic ring. A database containing comprehensive information is constructed using quantum chemical calculations. Multiple linear regression analysis is then used to screen key descriptors, facilitating the rapid identification of microscopic factors affecting thermal stability and promoting the scientific design of catalyst molecules in subsequent steps. Furthermore, this invention features a large sample size, extremely high correlation coefficient R², and low root mean square error, demonstrating good universality.

[0063] Beneficial Effect 2: The technical solution of this invention only requires simple optimization of the catalyst structure through quantum chemistry and calculation of the necessary descriptors to predict the thermal stability of the new catalyst, thereby achieving preliminary screening of the catalyst. This helps to accelerate the screening of catalysts, reduce the workload of artificial synthesis and testing, shorten the research and development cycle, reduce research and development costs, and save a lot of human and material resources. Attached Figure Description

[0064] Figure 1 A flowchart for constructing a prediction model for the thermal stability of polymerization catalysts based on quantum chemical calculations and multiple linear regression analysis.

[0065] Figure 2 The diagram shows the calculated reaction enthalpy and Gibbs free energy of Example 1, as well as the design diagram of the catalyst active species molecule.

[0066] Figure 3This is a schematic diagram of the molecular structure of the optimized catalyst active species A in Example 1.

[0067] Figure 4 This is an example diagram illustrating the three-dimensional structure of the nickel catalyst active species A, ethylene monomer B, catalyst ligand C, and their bimolecular ethylene-coordinated complex D in Example 1.

[0068] Figure 5 The predicted model and correlation curves for the thermal stability of the nickel catalyst in Example 1 are shown.

[0069] Figure 6 This is an example diagram illustrating the three-dimensional structure of the palladium catalyst active species A, ethylene monomer B, catalyst ligand C, and its bimolecular ethylene-coordinated complex D in Example 2.

[0070] Figure 7 The predicted model and correlation curves for the thermal stability of the palladium catalyst in Example 2 are shown. Detailed Implementation

[0071] 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.

[0072] Example 1

[0073] This embodiment provides a method for predicting the thermal stability of nickel catalysts, the process of which is as follows: Figure 1 This includes the following steps:

[0074] Step 1: Construct models for catalyst active species A, ethylene monomer B, catalyst ligand C, and its bimolecular ethylene-coordinated complex D. Optimize the geometric structure and perform frequency analysis on these models using density functional theory at level 1 to obtain their respective thermodynamically stable 3D structures. Collect and organize the corresponding thermodynamically stable 3D structures and their enthalpy values ​​and thermodynamic correction values ​​for Gibbs free energy. This includes the following steps:

[0075] Step 1.1: The central metal of the catalyst is nickel;

[0076] Step 1.2: By changing the substituents R on the [N,N] backbone and the aromatic ring Ph in the catalyst molecule structure (e.g., Figure 2 As shown), models are provided for constructing different catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene-coordinated complexes D using ligands with different substituents (e.g., Figure 2 As shown, where M is the central nickel catalyst metal, Et is ethyl, Ph is an aromatic ring, and L is an alpa-diimine ligand;

[0077] Step 1.3: Use GaussView software to create all the three-dimensional molecular structure models from Step 1.2 (e.g., Figure 4 (as shown), and export the model file in gjf format;

[0078] Step 1.4: Use the Gaussian16 program to perform geometric optimization and frequency analysis on all the three-dimensional molecular structure models constructed in Step 1.3;

[0079] The theoretical calculation method is B3LYP / 6-31G* (non-metallic atom) / Lanl2dz (metallic atom); optimization tends towards the lowest energy; there are no symmetry restrictions during the optimization process; if all frequency values ​​are positive, the molecule is determined to be a stable molecular conformation;

[0080] Step 1.5: Read the enthalpy and Gibbs free energy correction values ​​of different catalyst active species A, catalyst ligand C, and bimolecular ethylene-coordinated complex D from the output log file and save them in an Excel spreadsheet (as shown in Table 1).

[0081] Step 2: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at a high level (level 2) to collect the corresponding single-point energies of the thermodynamically stable 3D structure (as shown in Table 1). This includes the following steps:

[0082] Step 2.1: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at a high level 2. The theoretical calculation method is (SMD,solvent=toluene)B3LYP / 6-311G**(non-metal atom) / SDD(metal atom), and collect the single-point energy of the corresponding thermodynamically stable 3D structure.

[0083] Step 2.2: Read the single-point energy from the log file output in Step 2.1 and save it in an Excel spreadsheet, as shown in Table 1.

[0084] Step 3: Based on the statistical data from steps 1.5 and 2.2, calculate the enthalpy (H) of the molecules (A, B, C, D) respectively. A H B H C H D Gibbs free energy (G) A G B G C G D As shown in Table 1;

[0085] Step 4: Based on the enthalpy and Gibbs free energy of each molecule obtained in Step 3, calculate the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG), as shown in Table 1.

[0086] Step 5: Based on the optimized molecular structure of catalyst active species A in Step 1.4 (e.g., Figure 3 (As shown), the corresponding quantum chemical descriptor is calculated, specifically including the following steps:

[0087] Based on the molecular structure of the catalyst active species A optimized in step 1.4, the corresponding quantum chemical descriptor was calculated (as shown in Table 2).

[0088] The quantum chemical descriptors for the active species A of the catalyst are: dipole moment, fourth-order moment, and E. LUMO E HOMO ,%V bur q M d M-C d M-N Where, the dipole moment: D=(X 2 +Y 2 +Z 2 ) 1 / 2 Fourth-order moment: Q = (X 2 +Y 2 +Z 2 ) 1 / 2 %V bur : Embedded volume; E LUMO E HOMO For orbital energy levels; q M Mulliken charge at the metal center; d M-C d: The distance between the center of metal M and the red C atom; M-N The minimum distance step between the center of metal M and the two attached N atoms; the embedding volume (V) of the active species A in the catalyst. bur The calculations were performed using SambVca 2.1 software;

[0089] Step 6: Combine the quantum chemical descriptor calculated in Step 5 with the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG) calculated in Step 4 to construct a database;

[0090] Step 7: Using the database obtained in Step 6, perform multiple linear stepwise regression analysis to remove multicollinearity of descriptors and construct a model, retaining only descriptors with p-value < 0.05 to obtain the prediction model (e.g., ...). Figure 5 As shown in the following formula:

[0091] ΔH pred =1940.99E LUMO -158.53q M +515.36R 2 =0.91

[0092] ΔG pred =1649.91ELUMO -299.95q M +589.57R 2 =0.90

[0093] Among them, E LUMO q represents the lowest unoccupied molecular orbital energy of the active species A in the catalyst. M This represents the Mulliken charge of the central metal M of the active species A in the catalyst;

[0094] Step 8: Using the prediction model trained in Step 7, perform leave-one-out verification on the data in Step 6. If the correlation coefficient is greater than 0.85 and the error is less than 2.0 kcal / mol (DFT calculation error), it can be considered a relatively accurate prediction.

[0095] The tests revealed that the root mean square errors (RMS) for the dissociation enthalpy (ΔH) on the training and test sets were 2.37 kJ / mol and 2.46 kJ / mol, respectively; and the RMS errors for the dissociation energy (ΔG) on the training and test sets were 2.32 kJ / mol and 2.41 kJ / mol, respectively. This demonstrates that the prediction model trained in step 7 can accurately predict the thermal stability of nickel catalysts.

[0096] Table 1. Corrected values ​​of enthalpy, Gibbs free energy, and single-point energy for some nickel catalyst active species A, ethylene monomer B, catalyst ligand C, and its bimolecular ethylene-coordinated complex D.

[0097] Table 2. Quantum chemical descriptors of active species A in some nickel catalysts

[0098]

[0099] Based on the prediction model obtained in Example 1, the descriptor value can be calculated after optimizing the structure of a specific catalyst. Substituting the value into the prediction model, the dissociation enthalpy / energy can be obtained. The larger the value of the dissociation enthalpy / energy, the better its thermal stability, and vice versa. Thus, the performance of the catalyst can be evaluated.

[0100] Example 2

[0101] This embodiment provides a method for predicting the thermal stability of palladium catalysts, including the following steps:

[0102] Step 1: Construct models for catalyst active species A, ethylene monomer B, catalyst ligand C, and its bimolecular ethylene-coordinated complex D. Optimize the geometric structure and perform frequency analysis on these models at computational level 1 using density functional theory to obtain their respective thermodynamically stable 3D structures. Collect and organize the corresponding thermodynamically stable 3D structures and their enthalpy values ​​and thermodynamic correction values ​​for Gibbs free energy. This includes the following steps:

[0103] Step 1.1: The central metal of the catalyst is palladium;

[0104] Step 1.2: By changing the substituents R on the [N,N] backbone and the aromatic ring Ph in the catalyst molecule structure (e.g., Figure 2 As shown), models are provided for constructing different catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene-coordinated complexes D using ligands with different substituents (e.g., Figure 2 As shown, where M is the central metal palladium of the catalyst, Et is ethyl, Ph is an aromatic ring, and L is the alpa-diimide ligand;

[0105] Step 1.3: Use GaussView software to create all the three-dimensional molecular structure models from Step 1.2 (e.g., ...). Figure 6 (as shown), and export the model file in gjf format;

[0106] Step 1.4: Use the Gaussian16 program to perform geometric optimization and frequency analysis on all the three-dimensional molecular structure models constructed in Step 1.3; the theoretical calculation method is B3LYP / 6-31G* (non-metal atoms) / Lanl2dz (metal atoms); the optimization tends to the lowest energy; there are no symmetry restrictions during the optimization process; if all frequency values ​​are positive, the molecule is determined to be a stable molecular conformation;

[0107] Step 1.5: Read the enthalpy and Gibbs free energy correction values ​​of different catalyst active species A, catalyst ligand C, and bimolecular ethylene-coordinated complex D from the output log file and save them in an Excel spreadsheet (as shown in Table 3).

[0108] Step 2: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at a high level (level 2) to collect the corresponding single-point energies of the thermodynamically stable 3D structure (as shown in Table 3). This includes the following steps:

[0109] Step 2.1: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at the high-level level 2. The theoretical calculation method is (SMD,solvent=toluene)B3LYP / 6-311G**(non-metal atom) / SDD(metal atom). Collect the single-point energies of the corresponding stable molecular structures, as shown in Table 3.

[0110] Step 2.2: Read the single-point energy from the log file output in Step 2.1 and save it in an Excel spreadsheet, as shown in Table 3.

[0111] Step 3: Based on the statistical data from steps 1.5 and 2.2, calculate the enthalpy (H) of the molecules (A, B, C, D) respectively. A H B H C H D Gibbs free energy (G) A G B G C G D As shown in Table 3;

[0112] Step 4: Based on the enthalpy and Gibbs free energy of each molecule obtained in Step 3.1, calculate the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG), as shown in Table 3;

[0113] Step 5: Based on the molecular structure of the catalyst active species (A) optimized in Step 1 (e.g., Figure 3 (As shown), the corresponding quantum chemical descriptor is calculated, specifically including the following steps:

[0114] Based on the optimized molecular structure of catalyst active species A in step 1.4, the corresponding quantum chemical descriptors were calculated (as shown in Table 4); the quantum chemical descriptors of catalyst active species A are: dipole moment, fourth-order moment, E... LUMO E HOMO ,%V bur q M d M-C d M-N Wherein, the dipole moment: D=(X 2 +Y 2 +Z 2 ) 1 / 2 Fourth-order moment: Q = (X 2 +Y 2 +Z 2 ) 1 / 2 %V bur : Embedded volume; E LUMO E HOMO For orbital energy levels; q M Mulliken charge at the metal center; d M-Cd: The distance between the center of metal M and the red C atom; M-N The minimum distance step between the center of metal M and the two attached N atoms; the embedding volume (V) of the active species A in the catalyst. bur The calculations were performed using SambVca 2.1 software;

[0115] Step 6: Combine the quantum chemical descriptor calculated in Step 5 with the bond dissociation enthalpy (ΔH) and bond dissociation energy (ΔG) calculated in Step 4 to construct a database;

[0116] Step 7: Using the database obtained in Step 6, perform multiple linear stepwise regression analysis to remove multicollinearity of descriptors and construct a model, retaining descriptors with p-value < 0.05 to obtain the prediction model (e.g., ...). Figure 7 As shown in the following formula:

[0117] ΔH pred =1776.66E LUMO -365.24q M +602.29R 2 =0.90

[0118] ΔG pred =1881.21E LUMO -129.26q M +531.85R 2 =0.90

[0119] Among them, E LUMO q represents the lowest unoccupied molecular orbital energy of the active species (A) in the catalyst. M This represents the Mulliken charge of the central metal M of the active species A in the catalyst;

[0120] Step 8: Using the prediction model trained in Step 7, the data from Step 6 were validated using the leave-one-out method. The tests revealed that the root mean square errors (RMS) for the dissociation enthalpy (ΔH) on the training and test sets were 2.2 kJ / mol and 2.27 kJ / mol, respectively; the RMS errors for the dissociation energy (ΔG) on the training and test sets were 2.25 kJ / mol and 2.33 kJ / mol, respectively. This demonstrates that the prediction model trained in Step 7 can accurately predict the thermal stability of the palladium catalyst.

[0121] Table 3. Corrected values ​​of enthalpy, Gibbs free energy, and single-point energy for some palladium catalyst active species A, ethylene monomer B, catalyst ligand C, and its bimolecular ethylene-coordinated complex D.

[0122]

[0123] Table 4. Quantum chemical descriptors of active species A in some palladium catalysts

[0124]

[0125] Based on the prediction model obtained in Example 2, the descriptor value can be calculated after optimizing the structure of a specific catalyst. Substituting the value into the prediction model, the dissociation enthalpy / energy can be obtained. The larger the value of the dissociation enthalpy / energy, the better its thermal stability, and vice versa. Thus, the performance of the catalyst can be evaluated.

Claims

1. A method for predicting the thermal stability of coordination polymerization organometallic catalysts, comprising the following steps: Step 1: Construct models of catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene-coordinated complex D. Perform geometric structure optimization and frequency analysis on the models to obtain their respective thermodynamically stable 3D structures. Collect and organize the corresponding thermodynamically stable 3D structures and their enthalpy values ​​and thermodynamic correction values ​​of Gibbs free energy. Step 2: Based on the thermodynamically stable 3D structure, obtain the corresponding single-point energy; Step 3: Based on the thermodynamically stable 3D structure, enthalpy, thermodynamic correction value of Gibbs free energy, and single-point energy, obtain the enthalpy and Gibbs free energy of each of the thermodynamically stable 3D structures. Step 4: Based on the enthalpy and Gibbs free energy of each thermodynamically stable 3D structure, obtain the bond dissociation enthalpy and bond dissociation energy; Step 5: Based on the molecular structure of the catalyst active species A optimized in Step 1, calculate the corresponding quantum chemical descriptor; Step 6: Combine the quantum chemical descriptors, bond dissociation enthalpy, and bond dissociation energy to construct a database; Step 7: Based on the database, use multiple linear stepwise regression analysis to remove multicollinearity of descriptors and construct a prediction model; Step 8: Using the prediction model trained in Step 7, perform leave-one-out validation on the data in Step 6, record the root mean square error of the leave-one-out method on the training and test sets, and then evaluate its prediction performance.

2. The prediction method according to claim 1, wherein, In step 1, the geometric structure of the above models is optimized and frequency analysis is performed at computational level 1 using density functional theory to obtain their respective thermodynamically stable 3D structures. The corresponding stable molecular structures and their enthalpy values ​​and thermodynamic correction values ​​of Gibbs free energy are collected and organized.

3. The prediction method according to claim 1 or 2, wherein, Step 1 includes: Step 1.1: Use nickel or palladium as the central metal of the active species A of the catalyst; Step 1.2: By changing the substituents on the backbone and aromatic ring in the catalyst molecule structure, models of different catalyst active species A, ethylene monomer B, catalyst ligand C, and bimolecular ethylene coordinated complex D are constructed. Step 1.3: Use GaussView software to create all the three-dimensional molecular structure models from Step 1.2, and export the model files in gjf format; Step 1.4: Use Gaussian 16 to perform geometric optimization and frequency analysis on all the three-dimensional molecular structure models constructed in Step 1.3, and output a log file; the optimization tends to the lowest energy; there are no symmetry restrictions during the optimization process; if all frequency values ​​are positive, the molecule is determined to be a stable molecular conformation, that is, a thermodynamically stable 3D structure. Step 1.5: Read the enthalpy and Gibbs free energy correction values ​​of different catalyst active species A, catalyst ligand C, and bimolecular ethylene-coordinated complex D from the output log file.

4. The prediction method according to claim 1, wherein, Step 2 includes: Step 2.1: Based on the thermodynamically stable 3D structure obtained in Step 1, perform single-point calculations at a high level 2, collect the single-point energy of the corresponding thermodynamically stable 3D structure, and output a log file; Step 2.2: Read the single-point energy from the log file output in Step 2.

1.

5. The prediction method according to claim 1, wherein, In step 5, the quantum chemical descriptors include: dipole moment, fourth-order moment, and E. LUMO E HOMO %V bur q M d M-C d M-N ; in: Dipole moment: D=(X 2 +Y 2 +Z 2 ) 1 / 2 ; Fourth-order moment: Q = (X) 2 +Y 2 +Z 2 ) 1 / 2 ; %V bur The embedding volume; E LUMO E HOMO For orbital energy levels; q M The Mulliken charge at the metal center; d M-C denoted as , where is the distance between the center of metal M and the nearest C atom; d M-N It is the minimum distance between the center of metal M and the two N atoms attached to it.

6. The prediction method according to claim 5, wherein, The embedding volume of the catalyst active species A was calculated using SambVca 2.1 software.

7. The prediction method according to claim 1, wherein, In step 6, combining the quantum chemical descriptor, bond dissociation enthalpy, and bond dissociation energy means: using the quantum chemical descriptor as the independent variable, and the bond dissociation enthalpy and bond dissociation energy as two dependent variables, putting the three together to form a database, and establishing a relationship between the quantum chemical descriptor and the bond dissociation enthalpy and bond dissociation energy respectively.

8. The prediction method according to claim 1, wherein, In step 8, when removing multicollinearity of descriptors, only descriptors with p-value < 0.05 are retained.

9. The prediction method according to claim 1, wherein, In step 8, the predictive performance is evaluated as follows: the catalyst structure on the test set is optimized and the descriptor values ​​are calculated. The values ​​are then substituted into the predictive model relation to obtain the predicted bond dissociation enthalpy and bond dissociation energy. The correlation coefficient R between the predicted bond dissociation enthalpy and bond dissociation energy and the calculated bond dissociation enthalpy and bond dissociation energy was calculated. 2 The root mean square error (RMSE) is used to evaluate the predictive performance of the prediction model.

10. The prediction method according to claim 9, wherein, A prediction model is considered to accurately predict the thermal stability of a catalyst when the following condition is met: the correlation coefficient R... 2 Greater than 0.85, root mean square error less than 2.0 kcal / mol.

11. The prediction method according to claim 10, wherein, Based on the ability to accurately predict the thermal stability of a catalyst, the thermal stability of the catalyst is predicted as follows: After optimizing the structure of the catalyst under test, the descriptor value is calculated. The value is then substituted into the prediction model to obtain the bond dissociation enthalpy and bond dissociation energy. The larger the values ​​of the bond dissociation enthalpy and bond dissociation energy, the better its thermal stability, and vice versa. Based on this, the performance of the catalyst under test is evaluated.

Images