Molecular modeling of gaseous mercury species adsorption and transport in nanoporous silica membranes

The separation of Hg and HgO from industrial flue gas is an issue of much interest in the control of heavy metal emissions, and its feasibility for kinetic and equilibrium separation of Hg and HgO from N2 using disordered silica is theoretically investigated here using a combination of pore level transport modelling and effective medium theory. It is found that the classical Knudsen model generally leads to considerable over-prediction of the effective diffusion coefficient in the nanoporous silica compared to the more accurate Oscillator model that considers fluid-solid interaction. It is seen that nanoporous silica membranes are kinetically selective to nitrogen, the major component of flue gas, relative to Hg and HgO. Nevertheless, the kinetic selectivity for N2 is low, and only about 2-3, indicating that the kinetic separation of mercury species from flue gas is not practically feasible. On the other hand, the equilibrium selectivity of HgO over N2 is over 50 at 0.75 nm modal radius, while that for Hg exceeds 30 at 0.3 nm radius, suggesting this method of separation to be appropriate. Narrow pore silica materials of uniform nanopore size of about 0.3-0.75 nm are found to be optimal for the removal of mercury species.