Sang Kyu Kwak1, HuanCong Huang1, and Jayant K. Singh2. (1) Division of Chemical and Biomolecular Engineering, Nanyang Technological University, 62 Nanyang Drive, Singapore, 637459, Singapore, (2) Chemical Engineering, Indian Institute of Technology, Kanpur, Kanpur, India
Importance of cylindrical geometry has been highlighted in the fields of micro- to mesoporous materials. In particular, many studies have been conducted to understand the properties of confined fluids, such as thermophysical properties, packing of fluid molecules and phase behavior, which highly depend on the pore size, yet little known about the concrete pore dependence on the packing structure and fluid-solid coexistence and phase transition. To grasp the fundamentals, we use molecular dynamics simulation via considering model fluids, hard-sphere and square-well, in the hard repulsive cylindrical pore. The spreading pressure of the fluids is investigated for wide range of densities from dilute fluids to solid and pore diameters from 1.1σ to 8.0σ. The fluids have clear fluid-solid coexistence state, of which freezing and melting densities are lower than that of the bulk. We also investigate the structure properties of the hard sphere fluids by the study of radial density profiles and global bond order parameters. The possible fluid-solid transition is monitored by the sharp change of the bond order parameter W6, which we found more sensitive to the structure change other than the other bond order parameters under cylindrical confinement. Both the studies of pressure and bond order parameters show that the fluid behaviors have distinct difference for small pore size (approximately in the range from 2.2σ to 3.5σ) and large pore size (larger than 3.5σ). For the small pores there may exist a stable crystal-like structure while hardly observed in large pores. The pore size dependence of these properties shows that it was not a continuous but discrete function. The phase coexistence of fluid-solid under variable pore diameters was investigated by implementing the grand-canonical transition-matrix Monte Carlo with the histogram-reweighting and finite-size scaling methods.