Effects of Natural Organic Matter on Herbicide Adsorption to Activated Carbon

Doktorsavhandling, 1997

The presence of pesticides in the aquatic environment represents a potential threat to water quality and human health. Increasingly stringent drinking water quality directives focus on the need for efficient pesticide removal processes. Granular activated carbon (GAC) has proved the most effective and reliable barrier to sporadic, seasonal variations of pesticides, although it has now been established that natural organic matter (NOM) hinders adsorption through preloading and competitive adsorption. The effect of preloading GAC with NOM and the competitive interaction of NOM with the herbicides atrazine and MCPA were studied. Preloading and competitive interaction were considered as two separate mechanisms and were investigated through Freundlich adsorption isotherms, partition coefficients and site energy distribution theory. Adsorption tests with herbicides were carried out with preloaded and non-preloaded carbon in combination with NOM from lake water, treated water from two treatment plants, groundwater, ultrapure water and synthetic organic compounds including polystyrene sulphonate (PSS) standards. The results demonstrate that competitive interaction overrides the effect from preloading, even at low NOM concentrations. These findings can be explained by site energy distribution theory, where the activated carbon has a heterogeneous surface with a distribution of adsorption site energies. Characterisation of NOM through a pilot-scale GAC column as specific UV absorbance showed the preferential removal of UV-absorbing NOM compared to total organic carbon characterised NOM. The competitively absorbing component is probably the small, hydrophobic aromatic fraction of NOM. Analysis of the molecular weight (MW) distribution in the waters used in the study through high-performance size-exclusion chromatography demonstrated that a low MW refractory NOM fraction accounts for competitive interaction during GAC adsorption and that this fraction is not affected by conventional water treatment. These findings were confirmed by adsorption studies of atrazine in the presence of PSS standards. There appeared to be no evidence of pore blockage by larger molecules during atrazine adsorption to GAC. Prediction of GAC performance in a treatment plant for herbicide removal requires a consideration of both NOM preloading and NOM competitive interaction. The candidate GAC's should be precrushed and site-preloaded and adsorption isotherms for target compounds carried out in the presence of the relevant NOM. Further improvement in the design of GAC filters to allow the preferential adsorption of pesticides or the reduction of NOM interference seems an unproductive approach, because the competing fraction is always in excess of target compounds, even in groundwater. Further research might focus on nanofiltration-GAC which seems well poised to solve the competitive effects presented in this thesis.