Virus Inactivation by Homogeneous and Heterogeneous Fenton-like Processes

Human enteric viruses in water sources cause a great public health risk. Conventional disinfection treatments are not able to completely inactive viruses. However, advanced oxidation processes (AOPs) have recently been shown to effectively inactivate pathogens. One of the most promising AOPs is the Fenton process. In the framework of this work, the main objective was to characterize the fate of viruses upon inactivation by homogeneous and heterogeneous Fenton and Fenton-like processes, as well as to elucidate the mechanisms governing virus inactivation by these processes. MS2 coliphage, a commonly used surrogate for human enteric viruses, was used as the model organism. Virus inactivation by homogeneous, Cu- or Fe-catalyzed Fenton(-like) reactions, were studied at neutral pH. The effect of the metal (1-10 µM) and H2O2 (3-50 µM) concentrations, HO• production and sunlight on virus inactivation was investigated. Virus inactivation followed first-order kinetic with respect to the H2O2 concentration for both treatments. The influence of the metal concentrations was more complex. For the Cu/H2O2 system, it was found that inactivation was governed by soluble Cu. In contrast, for the Fe/H2O2 system, the colloidal Fe was involved in inactivation rather than dissolved iron. Sunlight only affected the Fe/H2O2 system. HO• production rates measured by electron spin resonance (ESR), could not account for the observed inactivation in Fe/H2O2 system. Other oxidants, such as ferryl species, must therefore play a role. Overall, our results have shown that virus inactivation by Cu- and Fe- catalyzed Fenton reaction may serve as an efficient disinfection method. Virus inactivation by the heterogeneous Fenton process was carried out via iron(hydr-)oxide particles, such as hematite (α-Fe2O3), goethite (α-FeOOH), magnetite (Fe3O4) and amorphous iron (Fe(OH)3), in batch reactors at circumneutral pH. The influence of adsorption and sunlight exposure on the survival of MS2 was investigated. Both mass-based and surface-area normalized pseudo-second order adsorption rate constants followed the same trend of α-FeOOH > α-Fe2O3 > Fe3O4 ≈ Fe(OH)3. Virus adsorption onto all particles was only partly reversible. In addition to irreversible adsorption, adsorption to three of the particles studied (α-FeOOH, Fe3O4, Fe(OH)3) caused slight virus inactivation (85%, 77%, 97%, respectively). Exposure of particle-adsorbed viruses to sunlight and H2O2 resulted in efficient inactivation, whereas inactivation was negligible for suspended viruses. The observed first-order inactivation rate constants were 1.44×10-3, 1.09×10-3, 0,58, 1.48 min-1 for α-Fe2O3, α-FeOOH, Fe3O4 and Fe(OH)3, respectively. Our results showed that in the heterogeneous Fenton system, inactivation was mainly attributed to a particle-mediated photo-Fenton-like reaction. Finally, the extent of genome and protein damage of MS2 coliphage during inactivation by the homogeneous Cu/H2O2 and Fe/H2O2 /sunlight systems were studied. The results showed that both damage to the genome and the capsid protein may contribute on virus inactivation by both treatments. The patterns of damage were different, even though the same oxidant (HO•) was present in both systems, indicating the source of the oxidant is important. For the Cu system, the extent of genome damage was similar to that of inactivation, indicating that inactivation may occur via single-hit kinetics. In contrast, for the Fe system, genome damage was very extensive in comparison to inactivation, consistent with multi-hit inactivation kinetics. For both systems, the most susceptible region of the capsid protein was peptide segment 84-106, which is located on the capsid outer surface. The other regions of the protein are not likely to be involved in inactivation. Overall, our findings suggested that both genome and protein oxidation by Cu and Fe systems may play a role in inactivation, and that the determination of molecular-level mechanisms governing inactivation can be assessed by MALDI-TOF-MS and qPCR.