Cross-category prediction of corrosion inhibitor performance based on molecular graph structures via a three-level message passing neural network model
Source abstract
Abstract
Current experimental verification, computational modeling, and machine learning methods for predicting corrosion inhibition efficiency (IE) are limited to specific inhibitor categories with high cost and poor generalization. In this study, a cross-category corrosion inhibitor dataset is constructed and a three-level direct message passing neural network (3 L–DMPNN) model using molecular structure information that integrates atomic-level, chemical bond-level, and molecular-level features to predict the IEs of compounds in a specific environment is established. This work demonstrates that the 3 L–DMPNN model can predict IEs of cross-category corrosion inhibitors from other independent literature and experimental dataset effectively and quickly.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Corrosion Behavior and Inhibition · Hydrogen embrittlement and corrosion behaviors in metals · Concrete Corrosion and Durability
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (70)
Show source-provided references
1. International measures of prevention, application, and economics of corrosion technologies study
2. The cost of corrosion in China
3. Handbook of science & engineering of green corrosion inhibitors: modern theory
4. Corrosion inhibition of carbon steel pipelines by some novel Schiff base compounds during acidizing treatment of oil wells studied by electrochemical and quantum chemical methods10.1016/j.molstruc.2016.10.078 ↗
5. A rapid screening multi-electrode method for the evaluation of corrosion inhibitors10.1016/j.electacta.2008.12.051 ↗
6. A combinatorial matrix of rare earth chloride mixtures as corrosion inhibitors of AA2024-T3: optimisation using potentiodynamic polarisation and EIS10.1016/j.electacta.2012.02.004 ↗
7. Density functional theory (DFT) as a powerful tool for designing new organic corrosion inhibitors. Part 1: An overview10.1016/j.corsci.2015.01.037 ↗
8. Cerium oxide nanoparticles influences on the binding and corrosion protection characteristics of a melamine-cured polyester resin on mild steel: An experimental, density functional theory and molecular dynamics simulation study10.1016/j.corsci.2017.01.021 ↗
9. Synthesis, characterization and the inhibition activity of 3-(4-cyanophenylazo)-2,4-pentanedione (L) on the corrosion of carbon steel, synergistic effect with other halide ions in 0.5 M H2SO410.1016/j.molstruc.2018.09.079 ↗
10. The use of quantum chemical methods in corrosion inhibitor studies10.1016/j.corsci.2008.08.043 ↗
11. Computational modeling: theoretical predictive tools for designing of potential organic corrosion inhibitors10.1016/j.molstruc.2021.130294 ↗
12. Density functional theory (DFT) as a powerful tool for designing new organic corrosion inhibitors. Part 1: An overview10.1016/j.corsci.2015.01.037 ↗
13. Theoretical evaluation of corrosion inhibition performance of some pyrazine derivatives10.1016/j.corsci.2014.03.008 ↗
14. A preliminary investigation of corrosion inhibition of mild steel in 0.5M H2SO4 by 2-amino-5-(n-pyridyl)-1,3,4-thiadiazole: Polarization, EIS and molecular dynamics simulations10.1016/j.corsci.2010.01.028 ↗
15. Quantitative structure–activity relationship model for amino acids as corrosion inhibitors based on the support vector machine and molecular design10.1016/j.corsci.2014.02.023 ↗
16. The discussion of descriptors for the QSAR model and molecular dynamics simulation of benzimidazole derivatives as corrosion inhibitors10.1016/j.corsci.2015.06.003 ↗
17. Computational design of inhibited prmers
18. Modeling corrosion inhibition efficacy of small organic molecules as non-toxic chromate alternatives using comparative molecular surface analysis (CoMSA)10.1016/j.chemosphere.2016.06.044 ↗
19. Development of QSAR-based (MLR/ANN) predictive models for effective design of pyridazine corrosion inhibitors
20. The effect of inhibitor structure on the corrosion of AA2024 and AA707510.1016/j.corsci.2011.02.040 ↗
21. Towards chromate-free corrosion inhibitors: structure–property models for organic alternatives10.1039/C3GC42540A ↗
22. Using high throughput experimental data and in silico models to discover alternatives to toxic chromate corrosion inhibitors10.1016/j.corsci.2016.02.008 ↗
23. Correlation between molecular features and electrochemical properties using an artificial neural network10.1016/j.matdes.2016.09.084 ↗
24. Elucidating Structure–Property Relationships in Aluminum Alloy Corrosion Inhibitors by Machine Learning10.1021/acs.jpcc.9b09538 ↗
25. Predicting the inhibition efficiencies of magnesium dissolution modulators using sparse machine learning models10.1038/s41524-021-00658-7 ↗
26. Data science based Mg corrosion engineering10.3389/fmats.2019.00053 ↗
27. Deep learning10.1038/nature14539 ↗
28. ASGN: An active semi-supervised graph neural network for molecular property prediction
29. Molecular property prediction: A multilevel quantum interactions modeling perspective
30. Graph neural network: A comprehensive review on non-euclidean space10.1109/ACCESS.2021.3071274 ↗
31. MoleculeNet: a benchmark for molecular machine learning10.1039/C7SC02664A ↗
32. Analyzing learned molecular representations for property prediction10.1021/acs.jcim.9b00237 ↗
33. Neural message passing for quantum chemistry
34. Molecule property prediction based on spatial graph embedding10.1021/acs.jcim.9b00410 ↗
35. B. Chen, G. Bécigneul, O.-E. Ganea, R. Barzilay, T. Jaakkola, Optimal transport graph neural networks, arXiv preprint arXiv: 2006.04804, (2020).
36. Deep learning based regression and multiclass models for acute oral toxicity prediction with automatic chemical feature extraction10.1021/acs.jcim.7b00244 ↗
37. Building attention and edge message passing neural networks for bioactivity and physical-chemical property prediction10.1186/s13321-019-0407-y ↗
38. K. Yang, K. Swanson, W. Jin, C. Coley, H. Gao, A. Guzman-Perez, T. Hopper, B.P. Kelley, A. Palmer, V. Settels, Are learned molecular representations ready for prime time? ChemRxiv (2019).10.26434/chemrxiv.7940594.v1 ↗
39. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules
40. Cristianini · 2000
41. Random forest in remote sensing: a review of applications and future directions10.1016/j.isprsjprs.2016.01.011 ↗
42. CORDATA: an open data management web application to select corrosion inhibitors, npj Mat
43. M. Aristarán, M. Tigas, J.B. Merrill, Tabula (Version 1.2.1), June 4, 2018. https://tabula.technology/.
44. ChemSchematicResolver: a toolkit to decode 2D chemical diagrams with labels and R-groups into annotated chemical named entities10.1021/acs.jcim.0c00042 ↗
45. Open babel: an open chemical toolbox10.1186/1758-2946-3-33 ↗
46. Optical structure recognition software to recover chemical information: OSRA, an open source solution10.1021/ci800067r ↗
47. E.J. Bjerrum, SMILES enumeration as data augmentation for neural network modeling of molecules, arXiv preprint arXiv: 1703.07076, (2017).
48. Convolutional networks on graphs for learning molecular fingerprints
49. A review of molecular representation in the age of machine learning10.1002/wcms.1603 ↗
50. Why is Tanimoto index an appropriate choice for fingerprint-based similarity calculations?10.1186/s13321-015-0069-3 ↗
51. Visualizing data using t-SNE
52. V. Nair, G.E. Hinton, Rectified linear units improve restricted boltzmann machines, in: Icml, 2010.
53. Taking the human out of the loop: A review of bayesian optimization10.1109/JPROC.2015.2494218 ↗
54. Simplistic correlations between molecular electronic properties and inhibition efficiencies: do they really exist?10.1016/j.corsci.2020.108856 ↗
55. Prediction of corrosion inhibition efficiency of pyridines and quinolines on an iron surface using machine learning-powered quantitative structure-property relationships10.1016/j.apsusc.2020.145612 ↗
56. Detailed DFT/MD simulation, surface analysis and electrochemical computer explorations of aldehyde derivatives for mild steel in 1.0 M HCl10.1016/j.colsurfa.2021.127822 ↗
57. Preventive behavior of phenol Schiff bases on mild steel corrosion in acidic medium part A: experimental and molecular modeling approach10.1016/j.cdc.2022.100864 ↗
58. Synthesis, structural characterization, Hirshfeld surface analysis and anti-corrosion on mild steel in 1M HCl of ethyl 2-(3-methyl-2-oxo-1,2-dihydroquinoxaline-1-yl)acetate10.1016/j.molstruc.2021.132047 ↗
59. Experimental and theoretical investigations of two quinolin-8-ol derivatives as inhibitors for carbon steel in 1 M HCl solution10.1016/j.jpcs.2022.110699 ↗
60. Touhami, Experimental and theoretical investigation of corrosion inhibition effect of two 8-hydroxyquinoline carbonitrile derivatives on mild steel in 1 M HCl solution10.1016/j.jpcs.2022.110866 ↗
61. New 3-(2-methoxyphenyl)-isoxazole-carvone: synthesis, spectroscopic characterization, and prevention of carbon steel corrosion in hydrochloric acid10.1016/j.molliq.2021.118311 ↗
62. Detailed experimental and computational explorations of pyran derivatives as corrosion inhibitors for mild steel in 1.0 M HCl: electrochemical/surface studies, DFT modeling, and MC simulation10.1016/j.molstruc.2022.132784 ↗
63. Experimental and theoretical examinations of two quinolin-8-ol-piperazine derivatives as organic corrosion inhibitors for C35E steel in hydrochloric acid10.1016/j.molliq.2022.118900 ↗
64. hajjaji, New epoxy resin as a corrosion inhibitor for the protection of carbon steel C38 in 1M HCl. experimental and theoretical studies (DFT, MC, and MD)10.1016/j.molstruc.2022.132425 ↗
65. Insight into the corrosion inhibition performance of two quinoline-3-carboxylate derivatives as highly efficient inhibitors for mild steel in acidic medium: Experimental and theoretical evaluations10.1016/j.molliq.2022.119470 ↗
66. Anti-corrosion behavior of 2-((3-(2-morpholino ethylamino) -N3-((pyridine-2-yl)methyl)propylimino)methyl)pyridine and its reduced form on carbon steel in hydrochloric acid solution: Experimental and theoretical studies10.1016/j.tsf.2021.139036 ↗
67. Experimental and theoretical investigation of corrosion inhibition effect of two piperazine-based ligands on carbon steel in acidic media10.1016/j.colsurfa.2022.128538 ↗
68. The study of surface activity and anti-corrosion of novel surfactants for carbon steel in 1 M HCl10.1016/j.molliq.2022.118747 ↗
69. Synthesis, characterizations, crystal structure, inhibition effects and theoretical study of novel Schiff base on the corrosion of carbon steel in 1 M HCl10.1016/j.molstruc.2022.132852 ↗
70. Electrochemical, density functional theory (DFT) and molecular dynamic (MD) simulations studies of synthesized three news Schiff bases as corrosion inhibitors on mild steel in the acidic environment10.1016/j.jics.2022.100492 ↗