Cellulose acetate is widely used in the manufacture of commercial water purification membranes. Although commercially available cellulose acetate membranes have high flux through membrane, they suffer limitations such as susceptibility to microbial attack and low chlorine resistance due to the presence of naturally occurring cellulose acetate backbone structure. Throughout filtration this leads to fouling due to the presence of natural organic matter (NOM) in the feed solution which coordinates with the hydrophobicity of the membrane. Modification of membrane surfaces can lead to improved filtration properties such as amplified permeate flux and high resistance to fouling. In order to reduce the hydrophobic interactions and thereby fouling due to NOM, hydrophilic poly(ethylene glycol) (PEG) monomer chains were attached to membrane via in situ graft polymerization. Different methods of grafting poly (ethylene glycol) (PEG) chains to commercially available cellulose acetate ultrafiltration membrane were considered and compared with respect to permeate flux and fouling prevention. Grafting was attained by forming a reactive radical on the membrane surface by oxidation followed by attaching PEG chains. These chain lengths and density were controlled by using a chain transfer agent.

Graft polymerization was carried out in two different methods named bulk method and drop method. In bulk method membrane samples were completely immersed in the reactive reagents for the modification to take place. Samples were dissolved first in oxidizing agent for 10 minutes then in 10% PEG solution for 5 minutes finally in chain transfer agent foe 2.5 minutes. All the reagents were in vigorous stirring motion during the modification in this process. In another method called Drop method, a membrane sample was placed flat on a glass sample holder and solutions containing the oxidizing agent were added to the membrane sheet drop-wise so that the entire sheet was filled with the oxidizing solution. After ten minutes the oxidizing solution was replaced by PEG solution for polymerization then chain transfer agent was added drop wise for controlling chain length. Energy Dispersive X-ray Spectroscopy, Scanning Electron Microscopy, Atomic Force Microscopy and Fourier Transform Infrared Spectroscopy were performed to assess the receptivity of membrane to the modification via chemical grafting.

Three different feed solutions were used to characterize the modification. Dextran solution was used to determine the effect of modification on uncharged particulate matter, while a modeled sea water solution to determine the influence of graft polymerization on NOM during filtration through cellulose acetate membranes. In these experiments sea water was simulated by forming an aqueous solution composed of 2 mg/l of each Suwannee River Fulvic and Humic Acids, along with 0.1 mM CaCl2 as a representative of divalent cations, 0.1 mM NaHCO3 as buffer system, 1M NaCl as background electrolyte and 1 mg/l of SiO2. Protein filtration was also performed to evaluate the modification influence on ultrafiltartion.

Drop method modification of the membrane resulted in better flux than bulk modified and virgin membranes during ultrafiltration of dextran solution. It also resulted in 10% increase in the rejection capacity than that of virgin membrane. Drop modification of membrane also led to modest decrease in the roughness of the membrane which reduces the membrane susceptibility to fouling. Only bulk modification was used to polymerize the membranes in the case of ultrafiltration of modeled sea water for the ease of use. Different sets of filtration runs were performed such as 1 minute, 5 minutes, 15 minutes and up to 6 hours to determine the fouling patterns due to natural organic matter such as humic and fulvic acids present in the feed solution. In this case modification led to the increase in the permeability of the membrane and finer fouling patterns during filtration.