Discovery of high energy and stable prismane derivatives by the high-throughput computation and machine learning combined strategy DOI Creative Commons
Shitai Guo, Jing Huang, Wen Qian

и другие.

FirePhysChem, Год журнала: 2023, Номер 4(1), С. 55 - 62

Опубликована: Июль 6, 2023

Motivated by the excellent detonation performance of octanitrocubane, prismane is another potential backbone with high strain energy in energetic molecule design. In this work, we aim to screen out candidates highly molecules from space derivatives. The high-throughput computation (HTC) performed based on 200 derived 1503 derivatives four substituents. Based calculated results, machine learning (ML) models density, velocity, pressure, heat formation and are established, thereby performances remaining 1303 samples predicted. It found that –NHNO2 group increases while both –NO2 –C(NO2)3 groups promote performances. velocity bond dissociation as criteria representing molecular stability, were screened good acceptable thermal stability. This work demonstrates efficiency HTC ML combined strategy for screening high-quality molecules.

Язык: Английский

Machine learning-based prediction and interpretation of decomposition temperatures of energetic materials DOI Creative Commons

Junnan Wu,

Siwei Song,

Xiaolan Tian

и другие.

Energetic Materials Frontiers, Год журнала: 2023, Номер 4(4), С. 254 - 261

Опубликована: Сен. 4, 2023

Exploring the application of machine learning (ML) in energetic materials (EMs) has been a hot research topic. Accordingly, prediction detonation properties EMs using ML methods attracted much attention. However, predictive models for thermal decomposition temperatures (Td) have scarcely reported. Furthermore, small datasets used these reports lead to weak generalization ability models. This study created dataset containing 1022 molecules with Td values 38–425 °C and determined an optimal model through training. The gradient boost regression (GBR) yielded coefficient determination (R2) 0.65 mean absolute error (MAE) 27.7 test set. further explored critical features, determining that accuracy was significantly influenced by descriptors representing molecular bond stability (i.e., BCUT metrics) atomic composition Molecular ID). Finally, analysis outlier structure indicated can be improved incorporating features related interactions. results this help gain deep understanding EM properties, particularly construction feature selection.

Язык: Английский

Процитировано

9
Machine Learning Techniques in Hydrogeological Research DOI
Song He, Xiaoping Zhou, Yuan Liu

и другие.

Springer hydrogeology, Год журнала: 2025, Номер unknown, С. 137 - 164

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
π-π2max: Bridging Molecular Characteristics to Crystal Packing in Nitro-Containing Two-Dimensional Energetic Materials DOI Creative Commons
Xiaokai He, Chao Chen, Zhixiang Zhang

и другие.

Defence Technology, Год журнала: 2025, Номер unknown

Опубликована: Фев. 1, 2025

Язык: Английский

Процитировано

0
Applications of Predictive Modeling for Energetic Materials DOI
Nasser Sheibani

Challenges and advances in computational chemistry and physics, Год журнала: 2025, Номер unknown, С. 339 - 364

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Harnessing machine learning enabled quickly predicting density of CHON molecules for discovering new energetic materials DOI Creative Commons

Ruoxu Zong,

Zi Li, Ziyu Hu

и другие.

AIP Advances, Год журнала: 2025, Номер 15(4)

Опубликована: Апрель 1, 2025

The application of machine learning in the research and development energetic materials is becoming increasingly widespread for performance prediction inverse design. Many advances have been achieved, especially discovery various new materials. However, main properties such as data acquisition, molecular characterization, limitations objects insufficient. Density, a critical factor influencing detonation materials, difficult to predict with high precision speed at large scale. In this study, techniques are employed density CHNO result explore simultaneously possessing stability. By screening dataset 16 548 candidate molecules, 175 potential high-performance molecules were identified. Among candidates, it noted that molecule velocity 7.328 Km/s pressure 24.48 GPa was which comparable TNT. study shows transformative accelerating novel vital diverse applications optimized expected accelerate next-generation

Язык: Английский

Процитировано

0
First principles calculations of electronic, vibrational, and thermodynamic properties of 3,6-dinitro-1,2,4,5-tetrazine biguanide DOI
Xuankai Dou

Journal of Molecular Modeling, Год журнала: 2025, Номер 31(5)

Опубликована: Апрель 21, 2025

Язык: Английский

Процитировано

0
Predicting the Melting Point of Energetic Molecules Using a Learnable Graph Neural Fingerprint Model DOI
Siwei Song, Yi Wang,

Xiaolan Tian

и другие.

The Journal of Physical Chemistry A, Год журнала: 2023, Номер 127(19), С. 4328 - 4337

Опубликована: Май 4, 2023

Melting point prediction for organic molecules has drawn widespread attention from both academic and industrial communities. In this work, a learnable graph neural fingerprint (GNF) was employed to develop melting model using dataset of over 90,000 molecules. The GNF exhibited significant advantage, with mean absolute error (MAE) 25.0 K, when compared other featurization methods. Furthermore, by integrating prior knowledge through customized descriptor set (i.e., CDS) into GNF, the accuracy resulting model, GNF_CDS, improved 24.7 surpassing performance previously reported models wide range structurally diverse compounds. Moreover, generalizability GNF_CDS significantly decreased MAE 17 K an independent containing melt-castable energetic This work clearly demonstrates that is still beneficial modeling molecular properties despite powerful learning capability networks, especially in specific fields where chemical data are lacking.

Язык: Английский

Процитировано

8
Promising Energetic Melt-Castable Material with Balanced Properties DOI
Fang Chen, Yi Wang, Siwei Song

и другие.

ACS Applied Materials & Interfaces, Год журнала: 2023, Номер 15(20), С. 24408 - 24415

Опубликована: Май 15, 2023

As one of the most widely used energetic materials to date, trinitrotoluene (TNT) suffers from several generally known drawbacks such as high toxicity, oil permeability, and poor mechanical properties, which are driving researchers explore new high-performance melt-castable for replacing TNT. However, it still remains a great challenge discover promising TNT alternative due multidimensional requirements practical applications. Herein, we reported molecule, 4-methoxy-1-methyl-3,5-dinitro-1H-pyrazole (named DMDNP). Besides reasonable melting point (Tm: 94.8 °C), good thermostability (Td: 293.2 excellent chemical compatibility, DMDNP exhibits some obvious advantages over including more environmentally friendly synthesis, yield, low volume shrinkage, electrostatic sensitivities, etc., demonstrating well-balanced properties promise replacement.

Язык: Английский

Процитировано

8
Molecular descriptor-enhanced graph neural network for energetic molecular property prediction DOI Open Access
Tianyu Gao, Yujin Ji, Cheng Liu

и другие.

Science China Materials, Год журнала: 2024, Номер 67(4), С. 1243 - 1252

Опубликована: Март 14, 2024

Язык: Английский

Процитировано

2
High-Throughput Screening of Promising Redox-Active Molecules with MolGAT DOI Creative Commons
Mesfin Diro Chaka, Chernet Amente Geffe, Álex Rodríguez

и другие.

ACS Omega, Год журнала: 2023, Номер 8(27), С. 24268 - 24278

Опубликована: Июнь 30, 2023

Redox flow batteries (RFBs) have emerged as a promising option for large-scale energy storage, owing to their high density, low cost, and environmental benefits. However, the identification of organic compounds with redox activity, aqueous solubility, stability, fast kinetics is crucial challenging step in developing an RFB technology. Density functional theory-based computational materials prediction screening time-consuming computationally expensive technique, yet it has success rate. To speed up discovery new desired properties, machine-learning-based models can be trained on large data sets. Graph neural networks (GNNs) are particularly well-suited non-Euclidean model complex relationships, making them ideal accelerating novel materials. In this study, GNN-based called MolGAT was developed predict potential molecules using molecular structures, atomic bond attributes. The set over 15,000 potentials ranging from -4.11 2.56. outperformed other GNN variants, such Attention Network, Convolution AttentiveFP models. used screen vast chemical comprising 581,014 molecules, namely OMDB, QM9, ZINC, CHEMBL, DELANEY, identified 23,467 redox-active use batteries. Of those, 20,716 were catholytes predicted 2.87 V, while 2,751 deemed anolytes -2.88 V. This work demonstrates capabilities graph condensed matter physics science species further electronic structure calculations experimental testing.

Язык: Английский

Процитировано

5