Method for identifying layered stacking mode of planar molecular crystal
By calculating parameters such as the dipole moment and electrostatic potential of planar molecules, a predictive model for the interaction energy of π-π dimers is established, which solves the problems of randomness in the design and long identification time of layered energetic materials in the prior art, and realizes efficient identification of layered stacking patterns.
Patent Information
- Application Number
- CN202410990413.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-23
AI Technical Summary
Existing design methods for layered energetic materials are prone to randomness, resulting in a limited variety of compounds. Furthermore, existing methods for identifying layered crystal stacking patterns are time-consuming and lack interpretability, making it difficult to achieve high-throughput design.
By calculating the dipole moment of planar molecules, the minimum electrostatic potential of molecular surfaces, and the number of heavy atoms, a predictive model of π-π dimer interaction energy was established to identify layered stacking modes. Data analysis and model fitting were performed using Gaussian16, Multiwfn, and Crystalexployer17 software.
It has accelerated the creation of novel layered energetic materials, simplified the identification process of crystal layered stacking modes, and improved the efficiency and interpretability of identification.
Smart Images

Figure FDA0004958711290000012 
Figure HDA0004958711310000011 
Figure HDA0004958711310000012
Abstract
Description
Technical Field
[0001] The invention relates to a theoretical method for identifying a planar molecular crystal layered stacking mode, which is suitable for predicting the crystal stacking mode of planar monocyclic and bicyclic compounds. Background Art
[0002] Layered energetic materials are a type of crystalline energetic materials with layered stacking characteristics formed by planar organic energetic molecules through intermolecular interactions such as hydrogen bonding and π-π stacking. They are low-sensitivity high-energy materials with great potential. The molecular structure and stacking pattern of the typical layered energetic material TATB can be seen in Figure 1 In addition, layered energetic crystals usually have a higher stacking coefficient, which can increase the crystal density, reduce the free volume, and help improve the detonation performance and impact sensitivity. However, there are still very few existing layered energetic materials, mainly because the existing design methods are too accidental, and one design strategy often only designs one compound.
[0003] Therefore, to achieve the goal of having more types of layered energetic materials, it is necessary to innovate existing design ideas and adopt high-throughput design methods. The two most important links in the high-throughput design method are molecular generation and rapid locking of energetic compounds with target crystal layered stacking patterns from massive chemical space. However, existing methods for recognizing crystal layered stacking patterns have the problems of difficulty in statistical analysis, long time consumption, and poor interpretability. Therefore, it is extremely important to establish a model for predicting planar molecular stacking patterns based on the physical characteristics of the molecules themselves, which can accelerate the creation of new layered energetic materials. Summary of the invention
[0004] In view of the defects or shortcomings of the prior art, the present invention provides a method for constructing a planar molecular crystal layered stacking pattern recognition model.
[0005] To this end, the method for constructing a planar molecular crystal layered stacking pattern recognition model provided by the present invention comprises:
[0006] (1) Obtain the dipole moment Dipole_moment of multiple planar molecules and the minimum value ESP of the molecular surface electrostatic potential min , the number of rings RingNum and the number of heavy atoms HeavyAtomNum and the central molecule and the distance from the central molecule in the corresponding planar molecular crystal The strongest dimer interaction energy is π-π 2max ;
[0007] (2) Statistical step (1) π-π of planar molecules with the same number of rings 2max , and confirmed the π-π 2max Threshold value;
[0008] (3) π-π 2max As the dependent variable, the dipole moment, the minimum value of the molecular surface electrostatic potential and the number of heavy atoms are used as independent variables. The data obtained in step (1) are fitted to obtain the planar molecule π-π 2ma x prediction models;
[0009] The π-π of the planar molecules with different numbers of rings 2max Threshold and planar π-π molecules 2ma x The prediction model constitutes a planar molecular crystal layer stacking pattern recognition model.
[0010] An alternative solution is that the Dipole_moment and ESP min The calculation method is as follows: use M06-2X-D3 and B3LYP / 6-311G(d,p) in Gaussian16 software to optimize the molecular structure, optimize the molecular structure to the local energy minimum on the potential energy surface, and obtain the wave function file of the optimal molecular structure; use Multiwfn to analyze the wave function file to obtain the molecular dipole moment Dipole_moment and the minimum value ESP of the molecular surface electrostatic potential min .
[0011] The alternative is π-π 2max The calculation method is: using the B3LYP / 6-311G(d,p) method of Crystalexployer17 software.
[0012] An alternative solution is to use the RDKIT package to obtain the number of rings RingNum and the number of heavy atoms HeavyAtomNum of the planar molecule.
[0013] An optional solution is that when RingNum=1 is obtained in step (2), the π-π 2max The threshold is 24 kJ / mol; if RingNum = 2, the π-π 2max The threshold is 44 kJ / mol;
[0014] π-π obtained in step (3) 2max The prediction model is:
[0015]
[0016] Where Pred(π-π 2max ) represents π-π 2max The predicted value, x1 represents Dipole_moment, x2 represents ESP min, x3 represents HeavyAtomCount.
[0017] Optionally, the planar molecule is a nitro-containing planar molecule.
[0018] The present invention also provides a method for identifying a planar molecular crystal layered stacking pattern. The provided method for identifying a planar molecular crystal layered stacking pattern includes:
[0019] (1) Obtain the dipole moment Dipole_moment of the planar molecule to be identified and the minimum value of the molecular surface electrostatic potential ESP min , the number of rings RingNum and the number of heavy atoms HeavyAtomNum, as well as the central molecule of the planar molecular crystal to be identified and the distance from the central molecule The strongest dimer interaction energy is π-π 2max ;
[0020] (2) Substitute the dipole moment, the minimum value of the molecular surface electrostatic potential and the number of heavy atoms obtained in step (1) into the π-π 2max Prediction model, get the π-π of the planar molecule to be identified 2max Predicted value;
[0021] (3) Select the corresponding π-π according to the number of rings in the model to be identified 2max Threshold, if the value obtained in step (2) is π-π 2max Prediction value ≥ selected π-π 2max threshold, the crystal stacking pattern of the planar molecule to be identified is lamellar; otherwise, the crystal stacking pattern of the planar molecule to be identified is non-laminar.
[0022] The present invention calculates the minimum value of the molecular surface electrostatic potential, dipole moment and number of heavy atoms of the planar molecule itself; and classifies the planar molecules according to the number of rings. It can predict whether the stacking pattern of the planar molecule has layered stacking characteristics through simple molecular property calculations. It avoids the time-consuming crystal structure prediction process based on first-principles calculations. At the same time, the prediction model is based on the physical meaning of the molecule and has strong explanatory power, which accelerates the discovery of new layered energetic materials. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 The molecular structure and crystal stacking pattern of the typical layered energetic material TATB;
[0024] Figure 2The molecular structures and crystal packing diagrams of ASAXIC (2-Hydroxy-3-nitrobenzoyl azide), COGGAI ((E)-1,2-bis(3-Nitro-1H-1,2,4-triazol-1-yl)diazene), FELVAW (bis(3-nitrophenyl)peroxyanhydride) and DNITPY (3,5-Dinitropyridine). DETAILED DESCRIPTION
[0025] Unless otherwise specified, the scientific and technical terms used herein are understood according to the knowledge of ordinary technicians in the relevant fields.
[0026] The present invention calculates the minimum value ESP of the molecular surface electrostatic potential of all known layered energetic material molecules (i.e., planar molecules containing nitro groups) by calculating the ring number RingNum, the molecular dipole moment Dipole_moment, and the molecular surface electrostatic potential. min and the number of heavy atoms in the moleculeHeavyAtomNum and the π-π of the corresponding molecule 2max Then, we established π-π through multiple regression. 2max The prediction model is used to determine the π-π of molecules with different numbers of rings based on the obtained data. 2max Threshold. Further based on the obtained π-π 2max The prediction model and the π-π of molecules with different numbers of ring gaps, 2max The threshold value identifies the crystal layer stacking pattern of the molecule to be identified. In a specific embodiment, the relevant parameters involved in the scheme of the present invention can be obtained using existing software.
[0027] In order to further explain the present invention, the present invention is further described below in conjunction with specific examples, wherein ASAXIC, COGGAI, FELVAW and DNITPY are the crystal structures corresponding to Example 2. However, the present invention is not limited to the following examples. The methods are conventional methods unless otherwise specified.
[0028] Embodiment 1:
[0029] This embodiment constructs a planar molecular crystal layered stacking pattern recognition model based on several layered energetic materials with 1 and 2 ring numbers disclosed in the prior art, and the method includes:
[0030] Draw each molecular structure in Gaussian view, and use M06-2X-D3 and B3LYP / 6-311G(d,p) in Gaussian16 to optimize the molecular structure; analyze the wave function of each molecular structure in Multiwfn software to obtain the molecular dipole moment Dipole_moment and the minimum value ESP of the molecular surface electrostatic potentialmin ;
[0031] Use the RDKIT package to obtain the number of rings (RingNum) and the number of heavy atoms (HeavyAtomNum) of each molecule;
[0032] The π-π of each molecule was calculated using the B3LYP / 6-311G(d,p) method of Crystalexployer17 software. 2max , including the calculation of the corresponding planar molecular crystal center molecule and the distance to it. The interaction energy between molecules within is then analyzed through energy framework to identify the dimer interaction energy with the strongest π-π interaction with the central molecule.
[0033] Multiple regression establishes the π-π 2max The prediction model is as follows:
[0034]
[0035] In the formula, x1 represents Dipole_moment, x2 represents ESP min , x3 represents HeavyAtomCount.
[0036] At the same time, statistics and confirmation are made that if RingNum = 1, the crystal stacking mode of the planar molecule is layered and must satisfy π-π 2max ≥24kJ / mol (i.e. π-π 2max The minimum threshold is 24 kJ / mol); if RingNum = 2, the crystal stacking mode of the planar molecule is layered and must satisfy π-π 2max ≥44kJ / mol (i.e. π-π 2max The minimum threshold is 44 kJ / mol).
[0037] Further based on the disclosure of the present invention, those skilled in the art can obtain the corresponding π-π when RingNum≥3 2max Prediction model and related π-π 2max Threshold value is set to identify molecules with more ring numbers.
[0038] Embodiment 2:
[0039] This embodiment selects four planar molecular energetic crystals with reported crystal stacking patterns, and the Refcodes (the unique identification number of the crystal in the CSD database) are ASAXIC, COGGAI, FELVAW and DNITPY respectively. The real stacking patterns of the four molecules are layered, non-layered, layered and non-layered, respectively. The following is a prediction using the model of the present invention:
[0040] The dipole_moment of ASAXIC, COGGAI, FELVAW and DNITPY obtained by the method described in Example 1 is 6.27 Debye, 0 Debye, 0 Debye and 1.88 Debye respectively, and ESP min They are -35.52 kJ / mol, -25.82 kJ / mol, -28.98 kJ / mol, and -22.88 kJ / mol respectively;
[0041] The SMILES of the four planar molecules were used as input of RDKIT to obtain the ring number (RingNum) and heavy atom number (HeavyAtomNum) corresponding to each molecule. The calculated RingNum of ASAXIC, COGGAI, FELVAW and DNITPY were 1, 2, 2, 1, and the HeavyAtomNum were 15, 18, 24, 12 respectively.
[0042] The calculated RingNum, Dipole_moment, ESP min Substitute HeavyAtomNum into the formula (1-1) obtained in Example 1; obtain the π-π of ASAXIC, COGGAI, FELVAW and DNITPY 2max The predicted results are 36.78 kJ / mol, 38.04 kJ / mol, 84.81 kJ / mol and 16.52 kJ / mol respectively.
[0043] According to the π-π of the corresponding ring number molecule in Example 1 of the present invention 2max Threshold conditions, it can be concluded that the predicted stacking modes of ASAXIC, COGGAI, FELVAW and DNITPY are lamellar, non-lamellar, lamellar and non-lamellar, respectively, which is consistent with the actual crystal stacking mode. The prediction results of the four planar molecules are all correct. The molecular and crystal stacking diagrams involved in this embodiment are shown in Figure 2 .
Claims
1. A method for constructing a planar molecular crystal layer stacking pattern recognition model, Characterized in that the method include: (1) Obtain the dipole moment Dipole_moment of multiple planar molecules and the minimum value ESP of the molecular surface electrostatic potential min , the number of rings RingNum and the number of heavy atoms HeavyAtomNum and the central molecule and the distance from the central molecule in the corresponding planar molecular crystal The strongest dimer interaction energy is π-π 2max ; (2) Statistical step (1) π-π of planar molecules with the same number of rings 2max , and confirmed the π-π 2max Threshold value; (3) π-π 2max As the dependent variable, the dipole moment, the minimum value of the molecular surface electrostatic potential and the number of heavy atoms are used as independent variables. The data obtained in step (1) are fitted to obtain the planar molecule π-π 2max prediction models; The π-π of the planar molecules with different numbers of rings 2max Threshold and planar π-π molecules 2max The prediction model constitutes a planar molecular crystal layer stacking pattern recognition model.
2. The method for constructing a planar molecular crystal layer stacking pattern recognition model according to claim 1, It is characterized in that The Dipole_moment and ESP min The calculation method is as follows: use M06-2X-D3 and B3LYP / 6-311G(d,p) in Gaussian16 software to optimize the molecular structure, optimize the molecular structure to the local energy minimum on the potential energy surface, and obtain the wave function file of the optimal molecular structure; use Multiwfn to analyze the wave function file to obtain the molecular dipole moment Dipole_moment and the minimum value ESP of the molecular surface electrostatic potential min .
3. The method for constructing a planar molecular crystal layer stacking pattern recognition model according to claim 1, It is characterized in that π-π 2max The calculation method is: using the B3LYP / 6-311G(d,p) method of Crystalexployer17 software.
4. The method for constructing a planar molecular crystal layer stacking pattern recognition model according to claim 1, It is characterized in that Use the RDKIT package to obtain the number of rings RingNum and the number of heavy atoms HeavyAtomNum of the planar molecule.
5. The method for constructing a planar molecular crystal layer stacking pattern recognition model according to claim 1, It is characterized in that When RingNum = 1 in step (2), the π-π 2max The threshold is 24 kJ / mol; if RingNum = 2, the π-π 2max The threshold is 44 kJ / mol; π-π obtained in step (3) 2max The prediction model is: Where Pred(π-π 2max ) represents π-π 2max The predicted value, x 1 Represents Dipole_moment, x 2 ESP min 、x 3 Represents HeavyAtomCount.
6. The method for constructing a planar molecular crystal layer stacking pattern recognition model according to claim 1, It is characterized in that The planar molecule is a nitro-containing planar molecule.
7. A method for recognizing the layered stacking pattern of planar molecular crystals, Characterized in that the method include: (1) Obtain the dipole moment Dipole_moment of the planar molecule to be identified and the minimum value of the molecular surface electrostatic potential ESP min , the number of rings RingNum and the number of heavy atoms HeavyAtomNum, as well as the central molecule of the planar molecular crystal to be identified and the distance from the central molecule The strongest dimer interaction energy is π-π 2max ; (2) Substituting the dipole moment, the minimum value of the molecular surface electrostatic potential and the number of heavy atoms obtained in step (1) into the π-π obtained in any of claims 1 to 5 2max Prediction model, get the π-π of the planar molecule to be identified 2max Predicted value; (3) Select the corresponding π-π according to the number of rings in the model to be identified 2max Threshold, if the value obtained in step (2) is π-π 2max Prediction value ≥ selected π-π 2max threshold, the crystal stacking pattern of the planar molecule to be identified is lamellar; otherwise, the crystal stacking pattern of the planar molecule to be identified is non-laminar.
Citation Information
Patent Citations
Novel method for regulating and controlling supermolecule polymerization
CN110358107A
Sulfur hexafluoride buffer gas selection method based on electronic localization function
CN113707229A
Method for identifying molecular layered stacking mode
CN115240780A
Self-assembled peptide structure prediction method based on coarse graining molecular dynamics simulation
CN118072851A
Apparatus for estimating packing structure of pi-conjugated molecule, method for estimating packing structure and utilization thereof
JP2010168354A
