Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method
Abstract Benzene separation from hydrocarbon mixtures is a challenge in the refining and petrochemical industries. The application of liquid–liquid extraction process using ionic liquids (I.Ls) is an option for this separation. The selection of the most appropriate I.L. for this application is a cha...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-12-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-79639-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846101204431536128 |
|---|---|
| author | Mahdieh Amereh Ali Ebrahimpoor Gorji Mohammad Amin Sobati |
| author_facet | Mahdieh Amereh Ali Ebrahimpoor Gorji Mohammad Amin Sobati |
| author_sort | Mahdieh Amereh |
| collection | DOAJ |
| description | Abstract Benzene separation from hydrocarbon mixtures is a challenge in the refining and petrochemical industries. The application of liquid–liquid extraction process using ionic liquids (I.Ls) is an option for this separation. The selection of the most appropriate I.L. for this application is a challenging task due to the variety of anion and cation structures. In the current study, the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases has been evaluated using the Quantitative Structure–Property Relationship (QSPR) method. A dataset comprising of 112 ternary systems (namely, I.L., benzene, and aliphatic hydrocarbon) was compiled after an extensive review of literature. The primary dataset consists of 17 anions, 20 cations, and 12 aliphatic hydrocarbons. Therefore, the impact of the structure of anion, cation, or aliphatic hydrocarbon on the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases has been investigated. The linear QSPR models were constructed using Multiple Linear Regression (MLR). The statistical evaluation of the final linear model showed that the constructed model (R2 = 0.900) has an acceptable capability to predict the mole fraction of benzene in the I.L.-rich phase. Additionally, non-linear QSPR models were developed using Genetic Programming (GP) and Artificial Neural Network (ANN) machine learning methods. The statistical evaluation of the GP model (R2 = 0.927) and ANN model (R2 = 0.939) showed that non-linear models had slightly higher prediction accuracy compared to the linear model. The final QSPR model was developed using the BELe3 cation descriptor which is a 2D Burden eigenvalues descriptor and HTm anion descriptor which is a 3D GETAWAY descriptor. After model construction, the selected molecular descriptors of anion and cation structures has been interpreted. The results showed that the size and the electronegativity of the atoms in the anion and cation structure are probably important parameters that affect the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases. Additionally, the anion shape can be considered as an effective parameter in the benzene extraction process. |
| format | Article |
| id | doaj-art-37923c3eb8e64afb85f0f6c98577e8a8 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-37923c3eb8e64afb85f0f6c98577e8a82024-12-29T12:24:54ZengNature PortfolioScientific Reports2045-23222024-12-0114112410.1038/s41598-024-79639-xMolecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR methodMahdieh Amereh0Ali Ebrahimpoor Gorji1Mohammad Amin Sobati2School of Chemical Engineering, Iran University of Science and Technology (IUST)School of Chemical Engineering, Iran University of Science and Technology (IUST)School of Chemical Engineering, Iran University of Science and Technology (IUST)Abstract Benzene separation from hydrocarbon mixtures is a challenge in the refining and petrochemical industries. The application of liquid–liquid extraction process using ionic liquids (I.Ls) is an option for this separation. The selection of the most appropriate I.L. for this application is a challenging task due to the variety of anion and cation structures. In the current study, the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases has been evaluated using the Quantitative Structure–Property Relationship (QSPR) method. A dataset comprising of 112 ternary systems (namely, I.L., benzene, and aliphatic hydrocarbon) was compiled after an extensive review of literature. The primary dataset consists of 17 anions, 20 cations, and 12 aliphatic hydrocarbons. Therefore, the impact of the structure of anion, cation, or aliphatic hydrocarbon on the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases has been investigated. The linear QSPR models were constructed using Multiple Linear Regression (MLR). The statistical evaluation of the final linear model showed that the constructed model (R2 = 0.900) has an acceptable capability to predict the mole fraction of benzene in the I.L.-rich phase. Additionally, non-linear QSPR models were developed using Genetic Programming (GP) and Artificial Neural Network (ANN) machine learning methods. The statistical evaluation of the GP model (R2 = 0.927) and ANN model (R2 = 0.939) showed that non-linear models had slightly higher prediction accuracy compared to the linear model. The final QSPR model was developed using the BELe3 cation descriptor which is a 2D Burden eigenvalues descriptor and HTm anion descriptor which is a 3D GETAWAY descriptor. After model construction, the selected molecular descriptors of anion and cation structures has been interpreted. The results showed that the size and the electronegativity of the atoms in the anion and cation structure are probably important parameters that affect the benzene distribution between the aliphatic hydrocarbon-rich and I.L.-rich phases. Additionally, the anion shape can be considered as an effective parameter in the benzene extraction process.https://doi.org/10.1038/s41598-024-79639-xQSPRIonic liquidsLiquid–liquid equilibriaBenzene extractionMachine learning |
| spellingShingle | Mahdieh Amereh Ali Ebrahimpoor Gorji Mohammad Amin Sobati Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method Scientific Reports QSPR Ionic liquids Liquid–liquid equilibria Benzene extraction Machine learning |
| title | Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method |
| title_full | Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method |
| title_fullStr | Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method |
| title_full_unstemmed | Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method |
| title_short | Molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via QSPR method |
| title_sort | molecular insights on the solvent screening for the benzene extraction from fuels using ionic liquids via qspr method |
| topic | QSPR Ionic liquids Liquid–liquid equilibria Benzene extraction Machine learning |
| url | https://doi.org/10.1038/s41598-024-79639-x |
| work_keys_str_mv | AT mahdiehamereh molecularinsightsonthesolventscreeningforthebenzeneextractionfromfuelsusingionicliquidsviaqsprmethod AT aliebrahimpoorgorji molecularinsightsonthesolventscreeningforthebenzeneextractionfromfuelsusingionicliquidsviaqsprmethod AT mohammadaminsobati molecularinsightsonthesolventscreeningforthebenzeneextractionfromfuelsusingionicliquidsviaqsprmethod |