Superfine Powdered Activated Carbon (S-PAC) Coupled with Microfiltration for the Removal of Trace Organics in Drinking Water Treatment

Anthropogenic contaminants - such as pharmaceuticals and personal care products - are an area of emerging concern in the treatment of drinking water. An integrated activated carbon membrane coating consisting of superfine powdered activated carbon (S-PAC) with particle size near or below one micrometer was explored to enhance removal of trace synthetic organic contaminants (SOCs) from water. S-PAC was chosen for its fast adsorption rates relative to conventionally sized PAC and atrazine was chosen as a model SOC. S-PAC and microfiltration membranes have a symbiotic relationship; membrane filtration separates S-PAC from water, while S-PAC adds capacity for a membrane process to remove soluble components. Three aspects of S-PAC in conjunction with membranes were examined, fouling by S-PAC on the membrane, effects of S-PAC production on material parameters, and modeling of S-PAC adsorption with and without a membrane. Fouling caused by carbon particles can result in marked reduction of filtration rate and an increased cost of operation. Since larger carbon particles foul less than smaller particles, while smaller carbons have faster adsorption performance, states of carbon aggregation were tested for filtration. Particles aggregated using the coagulant ferric chloride resulted in improved flux, while aluminum sulfate and polyaluminum chloride resulted in the same or worse filtration rates. A calcium chloride control showed that increased effective particle size via divalent bridging was very successful in reducing fouling. While particle size increased with conventional coagulants, the unflocculated metal precipitates likely contributed to membrane fouling. The methods of producing S-PAC determine material properties that affect both adsorption and filtration performance. In-house S-PACs - including multiple sizes of several carbon types - were prepared by wet bead milling and measured for both physical and chemical material parameters. Physical parameters, aside from particle size, did not change deterministically with milling duration, although stochastic changes were observed. Chemical measurements revealed a heavily oxidized external particle surface resulting from a high energy milling environment. Surfaces of interior pores appeared to be unaffected. Adsorption via batch kinetics and adsorption via S-PAC coating were modeled with analytical and computational models, respectively, using experimental data produced from the in-house S-PACs. The experimental data showed that removal of atrazine by S-PAC membrane coating correlated most strongly to a combination of oxygen content and the specific external surface area, while membrane fouling correlated to particle size and the specific external surface area. Batch kinetics data were modeled with the homogeneous surface diffusion model (HSDM) while membrane coating data were modeled with computational fluid dynamics (CFD). The fitted models required isotherm parameters indicative of an adsorbent with more capacity than was measured for S-PAC experimentally. Lastly, surface diffusion coefficients were neither constant nor varied with any measured material parameter. However, both model parameters correlated with overall atrazine removal, which indicates that model fits are related to performance, but it is not yet clear how they are connected