Knowledge Bank

UV-based water disinfection inactivates pathogens by damaging organisms' DNA. It differs from chlorine disinfection because it does not form toxic disinfection byproducts and is effective against chlorine-resistant pathogens. UV treatment systems typically use low pressure (LP) or medium pressure (MP) mercury lamps. LP lamps emit a peak wavelength of 254 nm and MP lamps emit wavelengths across the UV and visible light spectrum. LP lamps inactivate organisms by forming cyclobutane pyrimidine dimers (CPDs) on DNA. CPDs can be repaired by the light-activated photolyase enzyme in a process called photorepair, resulting in reactivation of microorganisms. To minimize impacts of photorepair, larger fluences of UV radiation are used in water treatment and residual chlorine is added after treatment. Krypton-chloride (KrCl) excimer lamps emit a peak wavelength of 222 nm and have been shown to damage proteins in laboratory settings. Studies have shown that short-wavelength UV light and the KrCl lamp lessen photorepair after disinfection. This research demonstrates that the KrCl lamp causes more damage to the photolyase enzyme during UV disinfection and prevents repair of bacteria compared to the LP lamp. E. coli suspended in solution were exposed to LP and KrCl lamps in a lab bench set up. After UV exposure, E. coli samples were moved under a fluorescent lamp in photorepair ("light") experiments or shielded from any light in baseline control ("dark") experiments. The inactivation and repair of E. coli samples were found by plating and counting culturable cells before UV exposure, after UV exposure to a given fluence, and after a given amount of repair. ELISA methods were used to compare concentrations of CPDs and photolyase in samples. The results show that both lamps inactivated E. coli with similar kinetics. A Geeraerd model estimated that the inactivation rate for the LP and KrCl lamp were 1.23 and 1.32 cm2 mJ−1, respectively. The CPD ELISAs showed that while both lamps induced fluence-dependent CPD formation, the LP lamp did more DNA damage than the KrCl lamp at a given fluence. An ELISA developed for this study found that while the LP lamp had no fluence-dependent impact on photolyase concentration, the KrCl lamp caused the concentration of photolyase in E. coli to decrease. Although repair experiments did not show a large difference between light and dark repair of samples exposed to either lamp, the KrCl samples exhibited less repair overall. The log repair rates with respect to photorepair fluence were 0.0013 and −0.0010 cm2 mJ-1 for the LP and KrCl lamps, respectively. Although CPD and photolyase concentrations measured during repair had large variations between duplicate samples, CPD-DNA decreased at a higher rate in LP samples than KrCl samples and photolyase concentration increased in LP samples but decreased in KrCl samples The UV fluence–response results suggest that protein damage was part of the disinfection mechanism of the KrCl lamp in these experiments. It was also shown that samples exposed to the KrCl lamp decreased in cell concentration during repair. CPD concentrations during repair decreased more in LP samples, though results had large variability. These results indicate that the KrCl lamp could be used in UV water treatment systems to prevent photorepair after treatment and lessen the amount of residual chlorine needed.