An energy-efficient Magneĺi phase Ti 4 O 7 reactive electrochemical membrane (REM) was applied for the oxidation of perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS). Approximately 5-log removal of these compounds was achieved in a single pass through the REM with residence times of ∼11 s at 3.3 V/SHE for PFOA and 3.6 V/SHE for PFOS. The permeate concentrations of PFOA and PFOS were <86 and 35 ng L −1 , from initial concentrations of 4.14 and 5 mg L −1 , respectively. The highest removal rates were 3415 ± 203 μmol m −2 h −1 for PFOA and 2436 ± 106 μmol m −2 h −1 for PFOS at a permeate flux of 720 L m −2 h −1 (residence times of ∼3.8 s). The levels of energy consumption (per log removal) to remove PFOA and PFOS to below the detection limits were 5.1 and 6.7 kWh m −3 , respectively. These values are the lowest reported for electrochemical oxidation and approximately an order of magnitude lower than those reported for other technologies (i.e., ultrasonication, photocatalysis, vacuum ultraviolet photolysis, and microwave−hydrothermal decomposition), demonstrating the promise of the REM technology for water treatment applications.

Aromatic polyamide thin-film composite membranes are widely used in reverse osmosis (RO) and nanofiltration (NF) due to their high water permeability and selectivity. However, these membranes undergo biofouling and can degrade and eventually fail during free chlorine exposure. To better understand this effect, the reactivity of the polyamide monomer (benzanilide (BA)) with free chlorine was tested under varying pH and chloride (Cl − ) conditions. The kinetic results indicated that the current existing mechanisms, especially the Orton rearrangement, were invalid. Revised reaction pathways were proposed where BA chlorination was driven by two independent pathways involving the anilide ring and amide nitrogen moieties. The ability for one moiety to be chosen over the other was highly dependent on the pH, Cl − concentration, and the resulting chlorinating agents (e.g., Cl 2 , HOCl, OCl − , and Cl 2 O) generated. Species-specific rate constants for BA with Cl 2 , OCl − , and HOCl equaled (7.6 ± 0.19) × 10 1 , (1.7 ± 1.5) × 10 1 , (2.1 ± 0.71) × 10 −2 M −1 s −1 , respectively. A similar value for Cl 2 O could not be accurately estimated under the tested conditions. The behavior of these chlorinating agents differed for each reactive site such that OCl − > HOCl for N-chlorination and Cl 2 > HOCl > OCl − for anilide ring chlorination. Experiments with modified monomers indicated that substituent placement largely affected which reactive site was kinetically favorable. Overall, such findings provide a predictive model of how the polyamide monomer degrades during chlorine exposure and guidance on how chlorine-resistant polyamide membranes should be designed.

No abstract

No abstract