Abstract
The present study aims at the extraction of a polyaromatic hydrocarbon from fuel oils using a novel low-cost deep eutectic solvent (DES). The DES is prepared by mixing the hydrogen bond acceptor (HBA; methyltriphenylphosphonium bromide, MTPB) and hydrogen bond donor (HBD; ethylene glycol) at a molar ratio of 1:4. The liquid-liquid equilibrium is then measured at ambient condition. The classical molecular dynamic (MD) simulation technique is then employed to investigate and compare the experimental phase behavior of a DES-quinoline-heptane ternary system. For performing the MD simulations, two experimental feed points are considered which gave high selectivity and distribution coefficient values. The interaction energies of different species and the structural properties such as radial distribution functions, average number of hydrogen bonds, and spatial distribution functions (SDFs) are then computed. It is found that the cation within the HBA, namely, MTP, possesses favorable interactions with quinoline when compared to HBD or anion (Br). MTP here acts as a HBA and contributes to the hydrogen bonding with quinoline, which results in higher experimental selectivity values. The calculations of SDFs further reveal the fact that the DES molecules are evenly distributed around the active sites of the quinoline molecule, whereas heptane molecules are found to be distributed around the nonactive sites of the aromatic ring.
| Original language | English |
|---|---|
| Pages (from-to) | 4006-4015 |
| Number of pages | 10 |
| Journal | Journal of Physical Chemistry B |
| Volume | 122 |
| Issue number | 14 |
| DOIs | |
| State | Published - Apr 12 2018 |
| Externally published | Yes |
Funding
The author P.K.N. acknowledges the fellowship grant from the Department of Science and Technology (DST), Government of India for supporting this work through INSPIRE fellowship program vide grant no. DST/INSPIRE Fellowship/2015/ IF150175. Computational time from PARAM Ishan Supercomputer facility of IIT Guwahati is highly acknowledged. The authors are also acknowledged to Debashis Kundu for his valuable suggestions in MD simulation systems.