Monitoring the microparticle transfer process in wastewater treatment systems is crucial for improving treatment performance. Supervised deep learning methods show high performance to automatically detect particles, but they rely on vast amounts of labeled data for training. To overcome this issue, we proposed a semi-supervised learning (SSL) method based on the Simple framework for Contrastive Learning of visual Representations (SimCLR), to detect microparticles free from sludge and attached to sludge. First, we pre-trained a ResNet50 backbone by SimCLR, to extract features from much unlabeled data (1,000 images). Then, we constructed a Mask R-CNN architecture based on the pre-trained ResNet50, and fine-tuned it on a small quantity of labeled data (≈200 images with ≈600 annotated particles) in supervised learning fashion. We showcased its performance and practical applicability for microscopy images obtained from the water lab of TU Delft. The results demonstrate that the SSL methods obtain a significant improvement in mean average precision of up to 5% compared to the conventional supervised learning method, when a limited amount of labeled data is available (e.g., 91 labeled images). Furthermore, these methods improve the average precision for detecting attached particles by over 12%. With the detection results from the SSL methods, we measured the attachment efficiency of microparticles to sludge under varying mixed liquor suspended solids concentration and aeration intensity. The precise measurements demonstrate the effectiveness and practical applicability of the SSL method in facilitating long-term monitoring of particle transfer processes in biological wastewater treatment systems.

A large decrease in permeability is often observed during the filtration of nano-sized colloids, while fouling is widely regarded as the main explanation for this phenomenon. The osmotic pressure or concentration polarization (CP) of colloids can also contribute to the flux decline. However, the contribution of CP to flux loss cannot be determined by the traditional CP model. In this study, the effect of fouling and CP/osmotic pressure on flux was distinguished. The CP values of polyethylene glycol (PEG) and silica-colloids were determined by the osmotic pressures near the membrane surface and in the feed. The CP induced by colloids accounted for 43–95% of the flux loss in our experiments. Silica exhibited higher CP values (127–460), compared to 7–71 for PEG. This was attributed to the slower back diffusion caused by the larger colloids, as evidenced by the diffusion coefficients of 4.30 × 10⁻¹¹ m²/s for silica (10 nm) and 1.45 × 10⁻¹⁰ m²/s for PEG (2.9 nm). Although the CP was mitigated by increasing the cross-flow velocity, CP values of 31 and 250 were observed for PEG and silica at high Reynolds number of 7317, respectively. The experimentally obtained CP values were also compared with those calculated by the film diffusion model.

This work focuses on the photocatalytic removal of recalcitrant organic pollutants in water treatment. Based on facile precipitation reaction, we fabricated a photocatalyst (PbCrO₄) in single crystals that present evident response to visible light and employed the catalyst in the photocatalytic decomposition of microcystin-LR (MC-LR). In the degradation test using the nanorods with prepared PbCrO₄ photocatalyst, a 100% removal efficiency (27 min reaction) and a kinetics constant of 0.1356 min⁻¹ were achieved. Such a high performance of PbCrO₄ in photocatalytic conversion of MC-LR was ascribed to its high carrier separation efficiency, positive valence band (VB) position, and good delocalization of VB and conduction band (CB). The test of electron spin-resonance resonance (ESR) demonstrated that excessive free [rad]OH radicals were produced during the PbCrO₄ photocatalysis of MC-LR. The density functional theory (DFT) and LC/MS/MS technology were employed to ascertain the intermediates during the MC-LR photocatalytic degradation. The major intermediates were resulted from the attack of hydroxyl radicals to the ADDA side chains of MC-LR structure. This study provides a proof-of-concept strategy to develop effective photocatalysts to efficiently produce [rad]OH radicals for the visible-light induced photocatalytic degradation of MC-LR in water.