This
study introduces a new approach for constructing atomistic
models of nanoporous carbon by randomly distributing carbon atoms
and pore volumes in a periodic box and then using empirical and ab initio molecular simulation tools to find the suitable
energy-minimum structures. The models, consisting of 5000, 8000, 12000,
and 64000 atoms, each at mass densities of 0.5, 0.75, and 1 g/cm3, were analyzed to determine their structural characteristics
and relaxed pore size distribution. Surface analysis of the pore region
revealed that sp atoms exist predominantly on surfaces and act as
active sites for oxygen adsorption. We also investigated the electronic
and vibrational properties of the models, and localized states near
the Fermi level were found to be primarily situated at sp carbon atoms
through which electrical conduction may occur. Additionally, the thermal
conductivity was calculated using heat flux correlations and the Green–Kubo
formula, and its dependence on pore geometry and connectivity was
analyzed. The behavior of the mechanical elasticity moduli (Shear,
Bulk, and Young’s moduli) of nanoporous carbons at the densities
of interest was discussed.