Photoinduced disinfection in sunlit natural waters: Measurement of the second order inactivation rate constants between E. coli and photogenerated transient species

Abstract

This work uncovers the implications of the estimation of exogenous inactivation rates for E. coli after the initial lag phase, and presents a strategy for the determination of the second-order inactivation rate constants (k2nd) of these bacteria with relevant transient species promoted by solar light in natural waters. For this purpose, specific precursors were considered (nitrate ion, rose bengal, anthraquinone-2-sulfonate) as well as the respective photo-generated transient species (i.e., hydroxyl radical (•OH), singlet oxygen (1O2) and triplet excited states). Under these conditions and by using suitable reference compounds (acesulfame K and 2,4,6-trimethylphenol in different series of experiments), the k2nd values were obtained after developing a proper competition kinetics methodology. The k2nd values were (2.5 ± 0.9) × 1011, (3.8 ± 1.6) × 107 and (1.8 ± 0.7) × 1010 M−1 s−1 for the inactivation of E. coli by •OH, 1O2 and the triplet state of anthraquinone-2-sulfonate (3AQ2S*), respectively. The measurement of a reaction rate constant that is higher than the diffusion-control limit for small molecules in aqueous solution implies that bacteria behave differently from molecules, e.g., because of the large size difference between bacteria and the transients. The obtained k2nd values were used for the modeling of the bacteria inactivation kinetics in outdoor systems (both water bodies and SODIS bottles), limited to the exponential decay phase that follows the initial lag time. Afterwards, the role of dissolved organic matter (DOM) as precursor of transient species for bacterial elimination was systematically studied. The interaction of different sunlight wavelength regions (UVB, UV-A, blue, green and yellow light) with Suwannee river (SW) and Nordic Lake organic matter (ND) was tested, and the photoinduced disinfection exerted by DOM isolates (SW DOM, Suwannee River Humic Acid, Suwannee River Fulvic Acid or Pony Lake Fulvic Acid) was compared. It was not possible to achieve a complete differentiation of the individual contributions of DOM triplet states (3DOM*) and 1O2 to bacterial inactivation. However, the application of competition kinetics to E. coli under solar irradiation in the presence of SW led to a k2nd value of (2.17 ± 0.40) × 1010 M−1 s−1, which is very near the value for inactivation by 3AQ2S* and suggests that the latter behaved very similar to SW-3DOM* and was a good 3DOM* proxy in the present case. The determination of the second-order inactivation rate constants of E. coli with •OH, 3DOM* and 1O2 represents a significant progress in the understanding of the external inactivation pathways of bacteria. It also allows predicting that, after the lag phase, 1O2 would contribute to photoinactivation to a far lesser extent than •OH and 3DOM*.