Adsorption interactions of s-triazine herbicides and natural organic matter by activated carbon

Predictive information is lacking on treatment performance in natural waters where adsorptive interactions of s-triazine herbicides (atrazine and cyanazine), trace organic contaminants, and natural organic matter (NOM) render a great deal of uncertainty in the carbon capacity and adsorption kinetics. Therefore, the objectives of research were to quantify the effect of NOM adsorption by GAC on the removal of atrazine and cyanazine from drinking water, to evaluate the adsorptive interactions between herbicides at low concentrations, and to compare the laboratory results with GAC filter-adsorber performance at the Higginsville Water Treatment Plant, Missouri.;Atrazine and cyanazine exhibit similar adsorption capacity. Therefore, atrazine and cyanazine adsorption can be evaluated as "Total" s-triazine adsorption. Individual s-triazine adsorption in dual-component systems can be predicted by conducting Total s-triazine isotherm experiments as in a single-component system. The initial s-triazine concentrations have a significant impact on the adsorption isotherms. Atrazine adsorption decreases with cyanazine present, and vice versa. Isotherm data for the two-component systems (atrazine and cyanazine) are predicted well by the Ideal Adsorbed Solution Theory model with the single-component Freundlich parameters.;For the ratios of atrazine metabolite (hydroxyatrazine; deethylatrazine, and deisopropylatrazine) to atrazine concentrations normally found in natural water (up to 50%), atrazine is favored over the metabolites by carbon adsorption. Also, atrazine is more adsorbable than trihalomethanes. An increase in total trihalomethanes Co (47 to 225 mug/L) slightly decreased atrazine adsorption (<4%). For trihalomethane levels normally found in water plants, the effect of trihalomethanes on atrazine adsorption is negligible.;For natural water, atrazine adsorption is reduced at lower pH's in which NOM adsorption is increased. A large fraction of the NOM in Higginsville Lake raw water has a molecular weight less than 3K Dalton. In comparison to higher molecular weight NOM, this fraction yields greater adsorption competition where atrazine adsorption decreased 23% with GAC and 39% with PAC. Greater competitive effects of NOM on atrazine adsorption occur at low initial atrazine concentration. Adsorption of atrazine and cyanazine by GAC is not strongly influenced by the inorganic background matrix.;Newly installed GAC filter-adsorbers (7.5-min empty-bed contact time) showed no detection or very low levels of s-triazine herbicides in the first six months of service. Older GAC filter-adsorbers (more than three year service) effluent had higher effluent s-triazine herbicide levels, but still less than 3 mug/L, and their removals of s-triazines ranged from 30 to 60%. GAC filter-adsorbers initially showed high reductions of UV254, but declined to 20% removal over long-term operation. NOM reduction and removal of s-triazines by anthracite filters are negligible. With the effluent from the anthracite and GAC filters blended together, parallel operations of GAC contactors can effectively control s-triazine herbicides in finished water and extend GAC bed life. Both field and batch studies conclude that the capacity of GAC for atrazine is much higher than with PAC.;In using the Freundlich parameters in models, precaution should be exercised between the test conditions and the actual field conditions. Site-specific studies are recommended to actually determine adsorption performance.