67639 research outputs found
Sort by
Microbial Dynamics and Stability in a Reverse Osmosis Drinking Water System: Methodological Advances and Applied Microbial Characterization
No full textThe microbiological stability of reverse osmosis (RO)–produced drinking water is challenged by its ultra-low biomass, residual disinfectants, and minimal nutrients. Despite growing interest in drinking water microbiology, specialized research on chlorinated, low-biomass RO systems remains scarce. This dissertation fills key gaps by (i) optimizing DNA-based methods for ultra-low-biomass samples, (ii) evaluating the effect of remineralization on microbial regrowth, and (iii) characterizing biofilm formation on shower plumbing materials in full-scale RO distribution.
In Chapter 2, we assessed DNA recovery from RO tap water by varying filtration volumes, testing dechlorination, and spiking with E. coli. We found that low cell densities (≈10²–10³ cells/mL) plus residual chlorine often yield DNA below sequencing thresholds. These results underscore the need for standardized pre-assessment protocols in chlorinated, low-biomass systems.
Chapter 3 compared five filter membranes (mixed-ester cellulose, polycarbonate, polyethersulfone, and PVDF) and introduced an incubation step to boost cell density. A polycarbonate 0.2 μm membrane outperformed others, delivering the highest DNA yield and lowest 16S background. We also demonstrated that multiple negative controls are essential for distinguishing true community signals from contamination in amplicon sequencing of low-biomass samples.
In Chapter 4, hydrostatic trials and a six-month simulated network examined magnesium remineralization (up to 50 mg/L) in RO water. Magnesium did not enhance overall bacterial growth—carbon remained the limiting nutrient—but did reshape biofilm communities, enriching taxa such as Piscinibacter and suppressing Legionella. These findings suggest that magnesium can be added safely for corrosion control without compromising biological stability.
Chapter 5 monitored biofilms on shower hoses and filters under real-use conditions. Flexible hoses supported richer and more even biofilm communities than inline filters, though both were dominated by oligotrophic genera (Pelomonas, Blastomonas, Porphyrobacter). Minimal organic carbon leached from the materials, confirming that nutrient limitation governs microbial growth in these niches.
Together, these studies advance our understanding of microbial ecology in RO-treated water and its distribution, highlighting how analytical methods, remineralization practices, and material properties influence water quality. The results inform improved monitoring and management strategies—such as protocol standardization, targeted remineralization, and material selection—to mitigate biofilm-related risks in domestic water systems. Future work should explore nutrient dynamics, flow conditions, and engineered biofilm control
Resonant acoustic mixing enables solvent-less amide coupling in solid-phase peptide synthesis
No full textSolid-phase peptide synthesis (SPPS) is the backbone of modern peptide production. However, it relies heavily on relatively toxic solvents and generates significant waste, limiting its sustainability and scalability. To address these limitations, we report the first fully solvent-less peptide coupling protocol for SPPS enabled by Resonant Acoustic Mixing (RAM), representing a step toward greener peptide manufacturing. This method eliminates bulk solvent use, reagent pre-dissolution, and pre-activation during coupling by using mechanical agitation to drive efficient amide bond formation. Optimized conditions (95g acceleration, 5 min, 1.5 equiv. Fmoc-amino acids) afford rapid and clear reactions with high conversion and purity. Notably, no external solvent is added during coupling; instead, residual solvent retained from resin pre-swelling creates a localized microenvironment sufficient for in situ activation. Compared to conventional SPPS, this protocol significantly reduces solvent and reagent use, reaction time, and waste. Process Mass Intensity (PMI) calculations show clear improvements, highlighting the method's environmental and economic benefits. This approach was validated by synthesizing two bioactive peptides (IKVAV and Angiotensin 1–7) in high yield and purity, and further demonstrated excellent scalability in a tenfold scale-up.This work was financially supported by the King Abdullah University of Science and Technology (KAUST), Saudi Arabia, Office of Sponsored Research (URF/1/4405 and URF/1/4726). We acknowledge Dr Srinivas Banala for helpful discussions
Probabilistic Shaping For Color-Shift Keying: Spectral-Efficient Transceiver Design and Prototype
No full textVisible light communication (VLC) systems often rely on uniformly distributed constellations, which can lead to suboptimal performance and a reduction in spectral efficiency (SE). To address this limitation, we propose a novel spectrally efficient modulation scheme that leverages probabilistic shaping (PS) to enhance the SE of VLC systems. The proposed scheme is based on the color-shift keying (CSK) modulation format for a quadrichromatic LED (QLED)-based system. We derive both the capacity and transmission rate (TR) for the proposed scheme, and adapt the TR based on the optical signal-to-noise ratio (OSNR) by optimizing the distribution of constellation symbols and forward error correction (FEC) coding rate, thus ensuring optimal system performance under varying channel conditions. Furthermore, we introduce an algorithm to compute the optimal capacity-approaching input distribution. To validate the proposed scheme, a QLED CSK system prototype was developed and experimentally tested. We evaluated the performance of the proposed scheme in terms of SE and frame error rate (FER) under different OSNR levels, and compare it against the conventional uniform-based scheme. The results demonstrated that the proposed scheme achieves a 20% improvement in SE over the uniform-based scheme
<i>N</i> , <i>N</i> -Difluoroarylsulfonamides: A Reagent for Chemodivergent Trifunctionalization of Unactivated Alkenes to Construct Quinoline Isosteres
No full textAlthough the replacement of aromatic rings with C(sp3)-rich isosteres has proven essential for improving the reactivity of drug candidates, the development of aza-arene-related isosteres, especially for privileged quinolines, remains limited due to the lack of efficient and controllable synthetic methods. Herein, we report a cost-effective Cu(I)-catalyzed chemodivergent trifunctionalization protocol for unactivated alkenes, which leverages the integrated versatile chemical reactivity of N,N-difluoroarylsulfonamides. This strategy enables the de novo synthesis of two types of topologically complex bridged sultams that are otherwise difficult to access using existing methods. The synthetic utility of our approach is demonstrated through the late-stage functionalization of biologically relevant targets and the synthesis of 15N-labeled molecules. More essentially, preliminary biological studies have identified that the synthesized bridged sultams serve as three-dimensional isosteres of quinoline, exhibiting enhanced bioactivity for cancer treatment. Comprehensive mechanistic studies further elucidate this reaction pathway.This project was sponsored by the Natural Science Foundation of Shandong (ZR2024MB088), Natural Science Foundation of Henan (252300421183, 242301420061, 242300420532), National Natural Science Foundation of China (21801250, 22201062, 22401077), the Fundamental Research Funds for the Central Universities (No. 25CX02006A), and Program for Science & Technology Innovation Talents in Universities of Henan Province (26HASTIT004). S.N. acknowledges the financial support from the Guangdong Basic and Applied Basic Research Foundation (2024A1515010323, 2025A1515011907) and the open research fund of Songshan Lake Materials Laboratory (2023SLABFN16). We sincerely thank Prof. Dr. Pengwei Xu (East China Normal University) for his very helpful discussions
Toward Robust Multimodal Egocentric Video Understanding
No full textEgocentric, or first-person, video offers a powerful medium for capturing and understanding natural human behavior in real-world environments. With the growing availability of wearable cameras and multimodal sensors, there is increasing interest in developing intelligent systems that can process, localize, and interpret activities from this unique perspective. However, egocentric video understanding presents fundamental challenges: videos are often long and unstructured, captured in diverse and dynamic scenes, and rich in sensory signals that may be noisy, incomplete, or constrained by real-world limitations such as privacy and power.
This thesis explores key problems in egocentric video understanding through the lens of multimodal learning, large-scale dataset curation, and robust modeling techniques for deployment under real-world constraints. It comprises four core contributions. First, we propose OWL, a method that integrates audiovisual temporal context to improve action localization in egocentric video, demonstrating significant performance gains on two large-scale datasets. Second, we document contributions to the creation of two landmark datasets, Ego4D and Ego-Exo4D, that enable scalable benchmarking of first-person perception tasks, including efforts in diverse data collection and energy-efficient activity recognition. Third, we address the practical issue of missing modalities in multimodal learning by introducing the Missing Modality Token (MMT), a transformer-compatible mechanism that allows robust inference even when inputs are partially absent. Finally, we present MiDl, a self-supervised, online test-time adaptation framework that dynamically adapts to modality incompleteness without requiring retraining or labeled data.
Together, these contributions advance the field of egocentric video understanding across both algorithmic and infrastructural dimensions. This work lays the foundation for building perception systems that are not only accurate and multimodally rich, but also scalable, resilient, and ready for deployment in the wild
Effectiveness of Max-Pooling for Fine-Tuning CLIP on Videos
No full textCLIP is a powerful spatial feature extractor trained on a large dataset of image-text pairs. It exhibits strong generalization when extended to other domains and modalities. However, its extension to videos is challenged by the need for additional temporal modeling. While recent works have attempted to bridge this modality gap through the integration of complex modules, we apply a simple and modular approach to enhance CLIP's video understanding on action recognition tasks. In its standard application, CLIP processes each video frame independently, restricting its ability to associate features across frames. To address this, we apply frame-wise max-pooling on the tokens within the transformer layer to construct a new set of tokens that aid the model in extracting temporal information better. We then use max-pooling to aggregate the features into a single video feature. We evaluate the effectiveness of this approach on different action recognition benchmarks, showing that max-pooling is able to help fine-tune the model to extract the features important for temporal modeling. Furthermore, we show that the max-pooling of tokens is particularly useful when applied to the last few layers of the model, which are typically more specialized and abstract for capturing high-level image features. To the best of our knowledge, we are able to achieve SOTA on the base-to-novel and few-shot benchmarks on the Something-SomethingV2 dataset
Elucidating the role of heterojunction in pristine non-fullerene acceptor organic solar cells
No full textNon-fullerene acceptors (NFAs) are rapidly transforming organic solar cell (OSC) performance and stability, yet the operational principles of pristine NFA devices remain underexplored. Here, we reveal that interfacial energetics, rather than bulk properties, dominate charge generation and recombination in pristine NFA-based photoactive layers (PALs). Although recent studies suggest spontaneous charge generation within the NFA bulk, our findings demonstrate that charge generation and extraction predominantly occur at the hole transport layer (HTL)/NFA interface, mimicking bilayer device behavior. Moreover, while CuSCN forms favorable interfaces that sustain long-lived charges and enhance photocurrent, PEDOT:PSS exhibits poor energy level alignment and a high trap density, leading to severe recombination losses via triplet exciton formation. Introducing as little as 2 wt% donor polymer surpasses the PAL percolation threshold, forming donor–acceptor interfaces that enhance photon utilization, reduce injection barriers, and improve charge transport. Our results not only challenge current interpretations of charge generation in pristine NFA devices but also establish new design principles for simplified, scalable single-component OSCs suited for next-generation semitransparent photovoltaics, including building-integrated photovoltaics (BIPV) and agrivoltaics.This work was supported by King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under award no: OSR-CARF/CCF-3079 and ORA-CRG10-2021-4681. The authors greatly acknowledge the device fabrication and characterization facilities at the KAUST Solar Center. O. J. S. acknowledges funding from the Research Council of Finland through project no. 357196. Y. H. thanks the research grant from KAUST global postdoc fellowship (award No. ORA-2023-6014)
A Report of a Child with SEC31A-Related Neurodevelopmental Disorder
No full textSEC31A-related neurodevelopmental disorder (Halperin–Birk syndrome) was recently identified in two siblings who shared the phenotype of profound developmental delay, structural brain defects, spastic quadriplegia with multiple contractures, seizures, dysmorphism, and optic nerve atrophy. Both patients died during childhood. In this study, we identified an additional patient who suffers from global developmental delay and seizures. Genetic analysis inclusive of whole exome and genome sequencing identified a homoallelic variant in the SEC31A (p.Cys453Trp). Various in silico classifiers predicted a deleterious effect of the replacement of cystein with tryptophan at the 453rd position. Protein–protein interaction (PPI) network analysis of SEC31A revealed high-confidence interactions with SEC13, SEC23A, and SEC23B, suggesting potential regulatory roles in these processes. Structural analysis of the SEC31A–SEC13 interaction and the Cys453Trp mutant in SEC31A predicted that the stability of coat protein complex II would be compromised. Our findings support the clinical correlation of SEC31A variants with neurodevelopmental disorder.We would like to thank the patient and his family for their kind participation in our study. We also thank the Genomic Medicine Center of Excellence (GMCoE) for allowing us to utilize the core facilities there. We thank the King Salman Center for Disability Research for their generous grant support.This research was funded by the King Salman Center for Disability Research, grant number: KSCDR-RAC: 2180 004.
Institutiona
Ethanol/ethyl acetate azeotropic mixture gas-phase separation using MOF-801-modified monolithic columns.
No full textOrganic monolithic columns have traditionally been used for macromolecule separations; however, enhancing their performance for small molecule and azeotrope separations remains challenging. Incorporating particles into monolithic matrices has emerged as a promising strategy to overcome these limitations, particularly in gas chromatography (GC). In this work, a composite monolithic column comprising a minimal loading (0.2 wt%) of zirconium-based MOF-801 particles embedded within a divinylbenzene (DVB) polymer framework was successfully fabricated and evaluated under low operating pressure (0.2 MPa). The short 15 cm × 0.25 mm i.d. MOF-801@DVB columns enabled rapid, high-efficiency separation of light hydrocarbons, achieving sub-minute separation of linear alkane mixtures with excellent repeatability (RSD% 0.38-1.16 %) over ten injections and operational stability across >2000 runs. Thermodynamic characterization revealed that MOF-801 incorporation enhanced the dispersive surface energy of the monolith while maintaining high permeability. McReynolds constants confirmed a polar interaction character, and a clear enthalpy-entropy compensation behavior was observed during adsorption. The monolith demonstrated successful separation of ethanol/ethyl acetate azeotropic mixtures, achieving a selectivity of 8.3 at 120 °C, surpassing benchmark materials like ZIF-8. Zero-coverage enthalpy analysis indicated stronger guest-host interactions in the MOF-801@DVB system, attributed to enhanced dispersion forces and hydrophilic framework effects. In addition, inter-batch reproducibility across three independently prepared columns was confirmed (RSD% 1.41-6.29 %), demonstrating the reliability of the fabrication approach. This study underscores the potential of integrating MOF particles into organic monolithic matrices to expand their application in small molecule GC separations, providing a cost-effective, high-stability platform for rapid gas-phase separations driven by thermodynamic control.This work was funded by the Ongoing Research Funding program - Research Chairs (ORF-RC-2025-1504), King Saud University, Riyadh, Saudi Arabia
Real-Time Control and Assessment of Grid Flexibility With Large-Scale Energy Assets Using HIL Framework
No full textThe transition towards renewable energy dominated power systems has introduced significant challenges in ensuring grid stability. To address these challenges, this thesis proposes a real-time hardware-in-the-loop (HIL) framework for control and assessment of grid flexibility enabled by large-scale energy assets, particularly hydrogen electrolyzers and grid-scale battery energy storage systems (BESS). A multi-time-scale frequency response (MTSFR) model is developed for alkaline water electrolysis (AWE) plants, capturing dynamic interactions between the electrolyzer stack and auxiliary components such as pumps and thermal subsystems. A high-fidelity neural network (HiFiNN) model is trained using detailed process simulations in Aspen HYSYS and integrated with a model predictive controller (MPC) to optimize frequency regulation performance while maintaining process stability. The proposed framework is validated through simulations and real-time controller-hardware-in-the-loop (CHIL) experiment to validate its feasibility for practical deployment. Complementing this, a cyber-physical BESS testbed is developed by interfacing the OPAL-RT real-time simulator with a physical controller like Raspberry Pi (RPi) exposed to realistic communication networks emulated in EXataCPS. The testbed facilitates validation of voltage support strategies for BESS under high renewable penetration scenarios. Case studies demonstrate that the proposed framework enhances system resilience through fast frequency and local voltage regulation, offering a validated pathway for deploying advanced control algorithms and accelerating the transition towards flexible, stable, and decarbonized power grids