Exploration of accurate calculation methods for pore size distributions

For porous carbons, which often have a range of pore sizes, the pore size distribution (PSD) is one of the most important characteristics and is currently evaluated by using kernel fitting methods represented by non-local density functional theory (NLDFT). However, it is a well-known problem that the mismatch between the ideal infinite smooth surface of the slit pore model and the actual carbon surface, including amorphous, corrugated, and defect structures causes errors in the PSDs, especially around 1 nm. We report new kernels for N2 and Ar adsorption at 77 K and 87 K, respectively, derived from Monte Carlo (MC) simulations based on a carbon slit-pore model that considers heterogeneity due to surface roughness. It shows better results than the NLDFT and Grand Canonical Monte Carlo (GCMC) smooth surface models but there is still a problem. The article explains how to evaluate the exact pore size distribution, including its relation to the adsorption isotherm.