Abstract
The present study aimed to employ a synergistic approach integrating experiments and molecular dynamics simulation to investigate the swelling of PET films caused by SC-CO2 and sorption of CO2 within the PET films. The rate of volume change (ΔV/V) of PET films increased with temperature elevation (50–140°C) and pressure elevation (10–27.5 MPa) which were often used in dyeing process. The sorption concentration decreased with rising temperature at the lower pressure, while at high pressure, it increased concomitantly with temperature elevation. The values of glass transition temperature (Tg) predicted by ΔV/V were 75.5, 69.7, and 60°C , and fitted by fractional free volume (FFV) were 75.8, 67.0, and 56.7°C , respectively. In addition, it was found that the swellability of PET was dependent upon the interaction between CO2 and the carboxyl group of PET chains. Eventually, the model based on backpropagation neural network optimized by Sparrow search algorithm (SSA-BPNN) exhibited robust predictive capabilities, yielding results consistent with experimental data for training set and testing set. Therefore, it can be concluded that the SSA-BPNN model is verified to be an efficient quantitative tool to predict the ΔV/V caused by SC-CO2 and sorption concentration within PET films in this work.
Get full access to this article
View all access options for this article.