Competitive adsorption of trace organic compounds by PAC in membrane filtration systems

Natural organic matter (NOM) can severely compromise adsorption of trace organic compounds. Due to the complex nature of NOM, it has been a great challenge to describe competitive adsorption of trace organic compounds in natural water. The objectives of this study were to understand the mechanisms of competitive adsorption and the important factors that affect them, and to develop a competitive adsorption model to predict trace compound removal from natural water when PAC is added to a membrane filtration reactor.; A small fraction of NOM was found to displace adsorbed atrazine and cause a continuous decrease in atrazine adsorption capacity in a continuous flow PAC/MF system. Considering the size exclusion effect, adsorption properties of NOM with different molecular weight were evaluated. The small NOM was found to be preferably adsorbed by a microporous PAC while larger NOM was better adsorbed by a PAC with a large fraction of mesopores. Both fractions were responsible for the lower atrazine removal efficiency than what batch adsorption parameters would predict. Model compounds were used to investigate competitive effects of these two NOM fractions. The small, strongly competing model compound was found to reduce atrazine adsorption capacity by direct competition for sites, but it had no effect on atrazine adsorption kinetics. The large competing compound, on the contrary, had a small effect on atrazine adsorption equilibrium, but severely blocked carbon pores and slowed the adsorption kinetics when the PAC was preloaded. A fraction of relatively large NOM molecules showed the same effect of pore blockage.; A three-component kinetic adsorption model was developed to describe competitive adsorption of a trace organic compound in the presence of NOM. NOM was simplified into two fractions: a strongly competing fraction that directly competes for adsorption sites, and a large, pore blocking fraction that blocks carbon pores. The model predictions were verified by experimental data obtained using model compounds. Model simulations for different operating conditions were made to optimize the PAC/MF system.