An investigation of the density of nano-confined subcritical/supercritical water

In various applications such as supercritical water gasification, oil and gas production, and energy harvesting and storage, subcritical/supercritical water is often confined within material pores. Investigating the density and critical parameters of high-temperature and high-pressure water in confined spaces can help us better understand the behavior of water within the pore space. In this study, a topological model consisting of two baths and a carbon nanotube (CNT) is built to examine the effect of temperature (300 K & 600–1173 K), pressure (1 atm & 20–30 MPa), and tube diameter (9.49–50.17Å) on confined density, and used machine learning (ML) to develop a predictive application (APP). The ML model demonstrates excellent predictive performance, achieving an R² of 0.9962 on the test set. Furthermore, the mean impact value (MIV) is used to evaluate the impact of independent variables on confined density, and find that temperature has the most significant effect. The results indicate that the critical temperature and pressure of water confined within different CNTs are not significantly different from those of bulk water, but the confined critical density is slightly lower than the bulk critical density.