Prediction of Density and Viscosity of Supercritical CO2-Expanded Lactate Esters for Pharmaceutical Extraction from Water Using Molecular Dynamics Simulation, Thermodynamic Modeling, and Machine Learning Study
Journal of Chemical & Engineering Data · 12.11.2026 · 71 (11) · 4568-4586
DOI: 10.1021/acs.jced.6c00372 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Chemical & Engineering Data, online erschienen am 07.10.2026. Verfasst von Guoxiao Cai und 5 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Gefördert von: Anhui Provincial Department of Education, Ministry of Education, Culture, Sports, Science and Technology, Japan International Cooperation Agency.
Themen: Künstliche Intelligenz in der Pflege, Simulation in der Pflegeausbildung
Abstract (Originalsprache)
Supercritical CO2-expanded lactate esters (CXLEs) are tunable, green solvents for separation applications. However, required thermophysical data on density and viscosity across the homologous series remain scarce. A multiscale computational framework integrating the Peng-Robinson equation of state (PR EoS), molecular dynamics (MD), density functional theory (DFT), and machine learning (ML) was developed to elucidate the structure–property–extraction relationships in CXLEs. The PR EoS density reference model was validated against literature density data for CO2-expanded ethyl lactate, yielding an AARD of 0.78%. At 313.15 K and 20.0 MPa, ReaxFF-MD predicted densities of CO2-expanded methyl lactate (CXML), ethyl lactate (CXEL), and butyl lactate (CXBL), yielding an overall AARD of 8.44% across the CO2 mole fractions (x1) from 0.10 to 0.95 compared with the PR EoS. ReaxFF-MD provided the first viscosity predictions, revealing a systematic CO2-induced reduction in viscosity. The RK Ridge and Minimal Log-Ridge predicted density and viscosity, respectively, for CO2-expanded propyl lactate (CXPL), pentyl lactate (CXPenL), and hexyl lactate (CXHL) using a limited three-ester training set, achieving R2 values of 0.94 and 0.97 and LOOCV AARDs of 5.39% and 7.95%, respectively. Preliminary MD trajectories further illustrated finite-time pharmaceutical transfer from water to CXEL. These results guide the rational design of CXLE solvents with tunable thermophysical properties for extraction.
Förderung
- Anhui Provincial Department of Education
- Ministry of Education, Culture, Sports, Science and Technology
- Japan International Cooperation Agency
- Anhui Provincial Teaching Research Quality Engineering Project
- University Talent Research Initiation Fund Project
Zitieren
APA
Cai, G., Tao, X., Xiong, W., Zhou, S., Hao, F., & Kong, C. Y. (2026). Prediction of Density and Viscosity of Supercritical CO2-Expanded Lactate Esters for Pharmaceutical Extraction from Water Using Molecular Dynamics Simulation, Thermodynamic Modeling, and Machine Learning Study. Journal of Chemical & Engineering Data, 71(11), 4568-4586. https://doi.org/10.1021/acs.jced.6c00372
Vancouver
Cai G, Tao X, Xiong W, Zhou S, Hao F, Kong CY. Prediction of Density and Viscosity of Supercritical CO2-Expanded Lactate Esters for Pharmaceutical Extraction from Water Using Molecular Dynamics Simulation, Thermodynamic Modeling, and Machine Learning Study. Journal of Chemical & Engineering Data. 2026;71(11):4568-4586. doi:10.1021/acs.jced.6c00372
BibTeX
@article{Cai2026,
author = {Cai, Guoxiao and Tao, Xun and Xiong, Wei and Zhou, Songlin and Hao, Fang and Kong, Chang Yi},
title = {Prediction of Density and Viscosity of Supercritical CO2-Expanded Lactate Esters for Pharmaceutical Extraction from Water Using Molecular Dynamics Simulation, Thermodynamic Modeling, and Machine Learning Study},
journal = {Journal of Chemical & Engineering Data},
year = {2026},
volume = {71},
number = {11},
pages = {4568--4586},
doi = {10.1021/acs.jced.6c00372}
}
EUVIMED