Electrochemical ammonia removal and disinfection of aquaculture wastewater using batch and flow reactors incorporating PtRu/graphite anode and graphite cathode
Ammonia removal and disinfection are two major problems in aquaculture systems, which require clean and reliable water to support long-term growth and health of target animals. In this study, we report electrochemical ammonia removal and disinfection of wastewater from an aquaculture farm (Mari’s Gardens) in Hawaii. First, we attempted to reproduce the work of Zollig and co-authors, who reported that direct ammonia oxidation can occur between 1 V and 1.6 V vs SHE on a graphite electrode in a solution (pH = 9.0) containing 1 M NaClO4, 0.25 M NH4ClO4, and 0.085 M NaCl. Our results, however, show that direct ammonia oxidation is unlikely to occur, at least at significant rates, on a graphite electrode in aqueous solutions (pH = 9.0) containing 0.7 M Na2SO4, 0.1 M (NH4)2SO4, and 0.02 M NaCl. We tentatively attribute this discrepancy to the different physico-chemical characteristics of graphite electrodes made by different manufacturers. Second, PtRu/graphite electrodes were prepared using a pulsed electrodeposition method, and electrode activity towards ammonia removal and disinfection was examined in both synthetic and real aquaculture wastewater using batch and flow reactors. The PtRu catalyst was partially oxidized at the beginning of electrolysis, and a significant increase in the electrode activity towards indirect ammonia oxidation was observed. Ammonia removal was slow when NaCl concentration was 0.66 mM, but the addition of NaCl (up to 20 mM) led to a drastic increase in the ammonia removal rate, indicating that ammonia removal proceeds via indirect oxidation. The ammonia removal rate depends primarily on NaCl concentration and current density and is independent of the initial ammonia concentration and solution pH. The ammonia removal rates can be modeled by pseudo zero-order kinetics, and a linear correlation can be drawn between the ammonia removal rate (k, mg L−1 min−1) and the product of NaCl concentration ([Cl-], mM) and current density (j, mA/cm2): k = 0.0047 [Cl-] j (R2 = 0.99). Free chlorine (Cl2, HOCl, and OCl-) was not detected in the solution until the complete removal of ammonia. Combined chlorine (NH2Cl, NHCl2, and NCl3) was measured at concentrations of 2–15 mg/L (as Cl2) during the ammonia removal process but was eliminated as soon as ammonia was depleted and an excess of free chlorine was available. Our detailed findings on the formation of both free chlorine and combined chlorine are significant to the mechanistic study of indirect ammonia oxidation. Ammonia removal experiments in synthetic and real aquaculture wastewater showed similar results. However, ammonia removal in the flow reactor took about three times longer than that in the batch reactor under similar conditions, likely due to hydrodynamic mixing differences. In addition, it was found that E. coli bacteria can be completely inactivated (5-log reduction) within a short time (e.g., 5 min).