A leading challenge in drinking water treatment is to remove small-sized viruses from the water in a simple and efficient manner. Multi-walled carbon nanotubes (MWCNT) are new generation adsorbents with previously demonstrated potential as filter media to improve virus removal. This study therefore aimed to evaluate the field applicability of MWCNT-filters for virus removal in water containing natural organic matter (NOM) as co-solute to viruses, using batch equilibrium experiments. Contrary to previous studies, our results showed with MS2 bacteriophages single-solute systems that the affinity of MWCNT for MS2 was low, since after 3 h of equilibration only 4 log₁₀ reduction value (LRV) of MS2 (20 mL at an initial concentration of 10⁶ PFU MS2/mL) were reached. Single solute experiments with Suwannee river NOM (SRNOM) performed with environmentally-relevant concentrations showed MWCNT surface saturation at initial SRNOM concentrations between 10 and 15 mgC/L, for water pH between 5.2 and 8.7. These results suggested that at NOM:virus ratios found in natural waters, the NOM would competitively suppress virus adsorption onto MWCNT, even at low NOM concentrations. We confirmed this expectation with SRNOM-MS2 co-solute experiments, which showed an exponential decrease of the MS2 LRV by MWCNT with an increase in the initial SRNOM concentration. More interestingly, we showed that pre-equilibrating MWCNT with a SRNOM solution at a concentration as low as 0.4 mgC/L resulted in a LRV decrease of 3 for MS2, due to the formation of a negatively charged SRNOM adlayer on the MWCNT surface. Complementary batch experiments with natural NOM-containing waters and competition experiments with SRNOM in the presence of CaCl₂ confirmed that the presence of NOM in waters challenges virus removal by MWCNT-filters, irrespective of the concentration and type of NOM and also in the presence of Ca²⁺. We therefore conclude that MWCNT-filters produced with commercially available pristine MWCNT cannot be considered as a viable technology for drinking water virus removal.

Both multi-walled carbon nanotubes (MWCNTs) and metal or metal oxides have demonstrated virus removal efficacy in drinking water applications. In this study, MWCNTs were coated with copper(I) oxide (Cu₂O) using three distinct synthesis procedures (copper ion attachment, copper hydroxide precipitation, and [Cu(NH₃)₄]²⁺ complex attachment) and virus removal efficacy (using MS2 bacteriophages) was evaluated. All synthesis procedures resulted in the presence of adsorbed, nanosized Cu₂O particles on the MWCNTs, shown using X-ray diffraction. Further, transmission electron microscopy confirmed uniform copper(I) oxide distribution along the MWCNTs for all three materials. Virus removal efficacy was assessed for all three synthesised composites both before and after material conditioning (filtering for at least 24 h/280 mL/h), and accounting for additional MS2 inactivation in the permeate due to continued copper inactivation from dissolved/desorbed copper in permeate (time-control). Material conditioning influenced virus removal, with the first litres of water containing higher concentrations of copper than the sixth litres of water, suggesting excess or non-bonded copper species dissolve from filters. Higher copper dissolution was observed for water at pH 5 than at pH 7, which decreased with time. Copper dissolution most likely caused an associated decrease in copper adsorbed to MWCNTs in the filters, which may explain the observed lower MS2 removal efficacy after conditioning. Additionally, the time-control study (immediately after filtration as compared to 2 h after filtration) highlighted continued MS2 inactivation in the permeate over time. The obtained results indicate that the synthesis procedure influences virus removal efficacy for MWCNTs coated with copper oxides and that virus removal is likely due to not only virus electrostatic adsorption to the coated MWCNTs, but also through antiviral properties of copper which continues to act in the permeate. In conclusion, it is highly important to revise the methods of testing filter materials for virus removal, as well as procedure for virus concentration evaluation.

A lab-scale membrane bioreactor was fed with fresh urine collected from source separation public toilets to perform full nitrification, urinary nitrogen stabilization and effluent sanitation. After acclimation, a nitrogen loading rate (NLR) of 1.2 kg TN per m³ d⁻¹ was reached with an average nitrification rate of 98.3%. Dissolved organic matter (DOM) transformation was studied within the MBR to identify biomass activity indicators that could be used for process optimization. By combining three-dimensional excitation emission fluorescence (3DEEM) and liquid chromatography coupled with organic carbon detection, nitrogen detection and UV detection (LC-OCD-OND-UVD), it was possible to investigate the transformation and fate of urine DOM during biological reactions and membrane separation. Results show that urine is mainly composed of low molecular weight molecules (<1000 Da) easily consumed by the active biomass. This consumption came with the production of soluble microbial products (SMP): biopolymers and humic substances. Significant correlations were observed between 3DEEM indicators and the NLR values. Biopolymers were mainly retained by the membrane and tended to accumulate in the reactor, which was of major interest since they were identified as DOM consumed by the active biomass in case of system instabilities. Humic substances were slightly retained by the membrane and mostly implied in irreversible fouling. This original approach allowed highlighting the quick active biomass acclimation under challenging conditions of high strength ammonia solutions such as urine.

The goal of this study is to help filling the research gaps linked to the on-line application of fluorescence spectroscopy in wastewater treatment and data processing tools suitable for rapid correction and extraction of data contained in three-dimensional fluorescence excitation-emission matrix (3DEEM) for real-time studies. 3DEEM was evaluated for direct quantification of Effluent Organic Matter (EfOM) fractions in full-scale MBR bulk supernatant and permeate samples. Principal Component Analysis (PCA) was used to investigate possible correlations between conventional Lowry and Dubois methods, Liquid Chromatography coupled to Organic Carbon and Organic Nitrogen Detection (LC-OCD-OND) and 3DEEM. 3DEEM data were analyzed using the volume of fluorescence (Φ) parameter from the Fluorescence Regional Integration (FRI) method. Two mathematical correlations were established between LC-OCD-OND and 3DEEM data to quantify protein-like and humic-like substances. These correlations were validated with supplementary data from the initial full-scale MBR, and were checked with samples from other systems (a second full-scale MBR, a full-scale conventional activated sludge (CAS) and a laboratory-scale MBR). While humic-like correlation showed satisfactory prediction for a second full-scale MBR and a CAS system, further studies are required for protein-like estimation in other systems. This new approach offers interesting perspectives for the on-line application of 3DEEM for EfOM quantification (protein-like and humic-like substances), fouling prediction and MBR process control.