Which molecular properties determine the impact sensitivity of an explosive? A machine learning quantitative investigation of nitroaromatic explosives

Duarte, Júlio César; Rocha, Romulo Dias da; Borges, Itamar

doi:10.1039/d2cp05339j

Cited by 16 publications

(7 citation statements)

References 101 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From the latter, we want to build the data set for modeling efforts. Current modeling efforts − are based on the available data, mostly restricted to early developed energetic materials, and often incorporate data from several sources. A single benchmarked data set with compounds representing different chemical classes is urgently needed.…”

Section: Resultsmentioning

confidence: 99%

Impact and Friction Sensitivity of Reactive Chemicals: From Reproducibility Study to Benchmark Data Set for Modeling

Muravyev,

Meerov,

Monogarov

et al. 2024

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

Mechanical stress is an important trigger of reactions in chemicals. Historically, the standard testing protocols for impact and friction sensitivity have been developed mainly for energetic materials and explosives. As a result, the structure− mechanical safety data is available for common explosives and is constantly reported for newly synthesized energetic compounds. The present work is motivated by the widely held among practitioners idea of high variability of mechanical sensitivity data, the advancements of new heterocyclic and high-nitrogen chemistry, and clear need in benchmark reference data set for QSPR modeling. We started from literature analysis and have already noted that many chemical papers lack the details required to replicate their findings regarding mechanical sensitivity. Next, we prepared over 100 species that have been previously synthesized and whose sensitivity had been reported by other researchers. The scatter within the literature and present study's results is illustrated and analyzed. Finally, we proposed a data set of 83 chemicals, which have the most reliable mechanical sensitivity data. This benchmark data set is recommended to be used for modeling of mechanical hazards of reactive chemicals. The logics of how this data set can be expanded in future is given; it might involve the collaborative efforts by different groups.

show abstract

Section: Resultsmentioning

confidence: 99%

Impact and Friction Sensitivity of Reactive Chemicals: From Reproducibility Study to Benchmark Data Set for Modeling

Muravyev,

Meerov,

Monogarov

et al. 2024

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

show abstract

“…Recently, such molecular properties and chemical descriptors determining impact sensitivity were searched using cutting-edge machine learning methods. [12,13] The QSPR studies are usually aimed to predict impact sensitivity of different nitro derivatives (nitro-and nitratoaromatics and aliphatics, nitramines, azides), [12][13][14][15] whereas the main trend of the last decade is the synthesis of nitrogen-rich energetic materials (both molecular and salt-like), often not including a nitro group. [14][15][16][17] All the nitro derivatives have a similar mode of decomposition, which assumes the XÀ NO 2 bond breaking (the trigger bond).…”

Section: Introductionmentioning

confidence: 99%

A General Model of Impact Sensitivity for Nitrogen‐Rich Energetic Materials: A Combined Incremental Theory and Genetic Function Approximation Study

Zhang,

Ma,

Chen

et al. 2024

ChemPhysChem

View full text Add to dashboard Cite

In this paper, we report the first example of impact sensitivity prediction based on the genetic function approximation (GFA) as a regression method. The prediction is applicable for a wide variety of chemical families, which include nitro compounds, peroxides, nitrogen‐rich salts, heterocycles, etc. Within this work, we have obtained 7 empirical models (with 27‐32 basis functions), which all provide 0.80 ≤ R2 ≤ 0.83 and 7.2 J ≤ RMSE ≤ 7.8 J (for 450 training set compounds) and 0.64 ≤ R2 ≤ 0.70 and 11.2 J ≤ RMSE ≤ 12.4 J (for 170 test set compounds). The models were developed using Friedman Lack‐of‐Fit function, which allows avoiding an overfitting. All the models have simple descriptors as basis functions and include linear splines. Furthermore, the applied descriptors do not require expensive calculation procedures, namely, non‐empirical quantum‐chemical calculations, complex iterative procedures, real space electron density analysis, etc. Most descriptors are based on structural and topological analysis and a part of them require very cheap semi‐empirical PM6 calculations. The prediction takes a few minutes as an average, and most of the time is for the structure preparation and manual calculation of the descriptor “Increment”, which is based on our recent incremental theory

show abstract

“…However, for reaching a good level of prediction accuracy, a sufficiently large dataset with size, variety, and homogeneity is usually important, although for smaller datasets, accurate data (e.g., from quantum chemical calculations) and a careful selection input of features is crucial for obtaining good results, as has been recently shown . In this vein, we recently used a ML approach combined with carefully selected quantum chemical input features to successfully investigate molecular properties affecting the sensitivity, thus the safety, of explosives …”

Section: Introductionmentioning

confidence: 99%

Machine Learning Determination of New Hammett’s Constants for meta- and para-Substituted Benzoic Acid Derivatives Employing Quantum Chemical Atomic Charge Methods

2023

Self Cite

View full text Add to dashboard Cite

Hammett's constants σ quantify the electron donor or electron acceptor power of a chemical group bonded to an aromatic ring. Their experimental values have been successfully used in many applications, but some are inconsistent or not measured. Therefore, developing an accurate and consistent set of Hammett's values is paramount. In this work,we employed different types of machine learning (ML) algorithms combined with quantum chemical calculations of atomic charges to predict theoretically new Hammett's constants σ m , σ p , σ m 0 , σ p 0 , σ p + , σ p − , σ R , and σ I for 90 chemical donor or acceptor groups. New σ values (219), including previously unknown 92, are proposed. The substituent groups were bonded to benzene and meta-and para-substituted benzoic acid derivatives. Among the charge methods (Mulliken, Loẅdin, Hirshfeld, and ChelpG), Hirshfeld showed the best agreement for most kinds of σ values. For each type of Hammett constant, linear expressions depending on carbon charges were obtained. The ML approach overall showed very close predictions to the original experimental values, with meta-and para-substituted benzoic acid derivative values showing the most accurate values. A new consistent set of Hammett's constants is presented, as well as simple equations for predicting new values for groups not included in the original set of 90.

show abstract

Which molecular properties determine the impact sensitivity of an explosive? A machine learning quantitative investigation of nitroaromatic explosives

Abstract: We decomposed density functional theory charge densitiesatom-centered of 53 nitroaromatic molecules intoelectric multipoles using the distributed multipole analysis that provides a detailed picture of the molecular electronic structure. Three electric...

Cited by 16 publications

References 101 publications

Impact and Friction Sensitivity of Reactive Chemicals: From Reproducibility Study to Benchmark Data Set for Modeling

Impact and Friction Sensitivity of Reactive Chemicals: From Reproducibility Study to Benchmark Data Set for Modeling

A General Model of Impact Sensitivity for Nitrogen‐Rich Energetic Materials: A Combined Incremental Theory and Genetic Function Approximation Study

Machine Learning Determination of New Hammett’s Constants for meta- and para-Substituted Benzoic Acid Derivatives Employing Quantum Chemical Atomic Charge Methods

Contact Info

Product

Resources

About