N-Nitrosodimethylamine Removal From Drinking Water By UV-Treatment: A Direct Comparison of Different UV Technologies

摘要

Four different lamps that produce ultraviolet light (UV) were investigated for treating N-nitrosodimethylamine (NDMA) in drinking water: low- and medium-pressure mercury (Hg) arc lamps (LP and MP, respectively), and two different pulsed lamps. Experiments were carried out in batch (120 mL, 3.5 cm pathlength) using collimated beam geometries, a synthetic drinking water at pH 8.1, and an initial NDMA concentration of 100 ppb. Although high for drinking water, this concentration is low enough that clean pseudo-lst order kinetics are observed. The UV dose-based removal rate constants range from 2.29 cm~2/J (LP) to 3.50 cm~2/J (pulsed) when the irradiance is specified over the 200 to 300 nm range. Because of experimental differences, the time-based rate constants are not directly comparable between lamp types. To obtain 2-log removal of NDMA, a much lower germicidal dose (approximately 350 mJ/cm~2) is apparently required with the pulsed lamps than with either LP or MP (700 and 1000 mJ/cm~2, respectively). The addition of 100 ppm H_2O_2 has no beneficial effect on the electrical efficiency of any of the lamps. The two pulsed lamps gave different removal rates, and the observed trends cannot be explained by their spectral output from 200 to 300 nm. It is suspected that photochemical reactions at wavelengths below 200 nm are a significant factor for these lamps.