Relating complex solute mixture characteristics to membrane fouling using flow field flow fractionation

Flow Field Flow Fractionation (Fl-FFF) was investigated as a tool to better understand complex solute mixture properties and relating them to membrane fouling. A standard synthetic solute mixture was developed for use in studies with Fl-FFF, cross flow filtration, and solute property characterizations to ensure that the composition was reliable and had consistent qualities. A mélange of components including colloidal silica and whey protein were characterized with standard techniques and were found to be realistic representation of natural water qualities.; A parametric sensitivity analysis of Fl-FFF assessed the reliability and reproducibility of the apparatus as a precision analytical tool representing the idealization of cross-flow filtration. An optimized protocol using surfactant-laden eluant and known diameter particle standards was used to determine theoretical solutions for channel thickness within ±5.2% uncertainty.; An advection-dispersion transport model is shown to provide ‘ideal’ (non-interacting) solute residence time distributions (RTDs) for comparison with ‘real’ RTDs obtained experimentally using standard mixtures at different cross-field velocities and solution ionic strengths. Chemical reactor theory is applied to the Fl-FFF system and a solute RTD moment analysis based on a particle diameter probability density function is used to extract characteristic properties rather than uniquely defined characteristics of the standard solute mixture. Mode, mean, and variance were found to provide reproducible results while skew and kurtosis had uncertainties greater than 40% due primarily to failure of the advection-dispersion transport model under high cross-field velocity conditions.; Constant flux cross-flow filtration studies investigated the effect of transport hydrodynamics (solvent flux to solute back diffusion (J/k) ratios), solution ionic strength, and feed water mass concentration for filtration using a regenerated cellulose ultrafiltration membrane.; A semi-empirical resistances-in-series modeling hypothesis was tested to exploit characteristic properties of solute mixtures based on RTD analysis and correction parameters based on the mean and variance of solute RTDs. The analytical protocol demonstrated that RTD analysis is a useful technique to describe colloid properties but it was not capable of accurately describing the complex nature of protein mélanges. A correction parameter based on RTD measurements was used to modify the mass transfer coefficient and was demonstrated to be sensitive to composition and solution ionic strength. Another correction parameter based on the moments obtained from ‘real’ and ‘ideal’ conditions was used to modify the specific layer resistance and was demonstrated to be sensitive to solute-solute and solute-membrane interactions induced by solution ionic strength conditions.