Phi Long Nguyen

The Ni/Cu ₂ O electrocatalyst, featuring improved *NO ₂ → *NO conversion and an enriched amount of surface-chemisorbed hydroxyl/water species, exhibits enhanced electrochemical nitrate-to-ammonia reduction performance.

Botulism is a potentially fatal illness in humans and many animals caused by the botulinum neurotoxins (BoNTs) produced mainly by Clostridium botulinum, which can be found in a wide variety of foods. The treatment of this illness requires antitoxin drugs given to the patients within a 48 h window to be effective. Therefore, there is a strong need for the development of rapid and sensitive biosensors for the early detection of the toxins in foods. In this work, by using comprehensive atomistic computer modeling and simulation, we investigate aptamers that bind BoNT serotypes A, B, and C to serve as recognition elements in the biosensor design for cross-serotype detection. Known aptamers targeting BoNT/A and BoNT/C are investigated in this work to understand whether they can cross-react with other serotypes. Our results highlight the importance of structural relaxation of the aptamers, which the standard docking procedure cannot adequately capture. During molecular dynamics, most of the aptamer structural relaxation correlates with binding to the toxins. Among the aptamers chosen, we found two aptamers A4 and A5 that bind equally strongly to all serotypes. C5 binds equally strongly to its designated target of serotype C, but it binds significantly weaker to other serotypes. These most promising aptamers, based on the computational analyses, are all ssDNA aptamers and share a common feature of a DNA double helix of 5–6 nucleotide base pairs. For all aptamers investigated, there are two binding sites for an aptamer on BoNT, which are located on opposite sides of the protein. This suggests an experimental setup of a sandwich array to allow aptamers to bind to both binding sites for optimal sensitivities. Our computational results provide molecular mechanisms of binding and assist in further optimization and design of new aptamers for these botulinum neurotoxins.

Robust sensor fusion remains a fundamental challenge for autonomous mobile robots operating under degenerate sensing conditions, such as wheel slippage on low-friction surfaces and visual degradation in feature-poor environments. Conventional approaches typically rely on fixed noise models or heuristic threshold-based mechanisms to detect sensor failures, limiting generalization and theoretical consistency. This paper proposes an Eigenvalue-Adaptive Fusion (EAF) framework that formulates sensor reliability as a probabilistic quantity derived from information-theoretic principles. Specifically, measurement quality is characterized through the spectral analysis of the observed Fisher Information Matrix (FIM). The framework incorporates a latent reliability model, from which an adaptive weighting mechanism is derived. This mechanism admits a Bayesian interpretation as approximate posterior inference over sensor reliability. The resulting formulation connects information theory, observability, and probabilistic inference within a unified framework. The proposed method eliminates heuristic thresholds and enables continuous adaptation to varying sensing conditions while preserving realtime performance. Experimental results on an autonomous robot in real-world indoor environments demonstrate improved robustness and trajectory consistency compared to conventional fusion strategies and wheel odometry only. Ultimately, this framework establishes a principled connection between sensor fusion, information theory, and probabilistic inference, while maintaining real-time performance.

Simultaneous wireless information and power transfer (SWIPT) with intelligent reflecting surfaces (IRS) enhances wireless performance but requires solving discrete phase optimization under hardware constraints. This letter studies joint energy delivery and interference suppression in a dual-transmitter IRS-aided SWIPT system with 1-bit elements. A hybrid binary grey wolf-particle swarm optimization algorithm directly optimizes IRS phases in the discrete domain. Simulations under standardized channels demonstrate up to 15.0 bit/s/Hz spectral efficiency and 36.0 dB received power, outperforming conventional binary and continuous-relaxation methods while remaining robust to channel state information errors, scaling, interference constraints, and mobility.

The intrinsic material properties play a more decisive role in governing electrocatalytic performance than the presence of abundant oxygen vacancies.

This study presents an enhanced tribo‐hygroelectric generator (THEG) capable of efficiently converting mechanical and atmospheric water energies into direct current (DC) power. The device utilizes water‐infused porous cellulose (WIPC) as the active layer, sandwiched between two electrodes with different work functions (e.g., aluminum and copper). Electricity generation originates from the synergistic effects of triboelectric charge excitation and Schottky junction–induced charge separation at the Al/WIPC interface. The hydrogen‐bonded network formed between water molecules and cellulose functional groups establishes an efficient pathway for charge transport, while mechanical compression reduces the Schottky barrier height, enhances carrier density, and lowers internal resistance. As a result, the optimized THEG achieved a peak current of 4.46 mA and a current density of 11.15 A/m ² , representing a threefold enhancement over the previous sliding‐type design. Although the output voltage of a single cell is modest, a series of 10 cells generated 4.2 V—sufficient to power small electronic devices such as light‐emitting diodes (LEDs), calculators, and DC motors. Moreover, the THEG demonstrated excellent sensing performance for applied pressure and compression rate, exhibiting strong linear correlations (R ² ≥ 0.99). These findings underscore the potential of THEG as a sustainable and multifunctional platform for sustainable energy harvesting, self‐powered sensing, and practical low‐voltage power applications.

Conventional direct methanol fuel cells (DMFCs) are fundamentally limited by their dependence on noble-metal catalysts and the inevitable release of CO2. Overcoming these barriers requires the design of non-precious alternatives capable of selectively steering methanol oxidation toward formate, avoiding greenhouse gas emissions, but it remains challenging to control. In this work, we present a CO2-free DMFC architecture that utilizes ultra-dispersed CuFe clusters anchored on plastic-recycled hierarchically porous carbon (CuFe/RHPC). This system achieves exceptional electrocatalytic metrics, including a maximum power density of 127.8 mW/cm2 and a current density of 613.83 mA/cm2 at a cell voltage of 0.23 V. Notably, the device demonstrates robust long-term stability over 200 h with 96.9% current-density retention and 99.63% Faraday efficiency for formate synthesis. Mechanistic insights from density functional theory (DFT) reveal that the electronic d-orbital coupling between Cu and Fe facilitates selective partial oxidation by inhibiting the deep oxidation path to CO2. This strategy offers a dual-benefit approach for sustainable energy conversion and waste upcycling through high-performance, non-noble bimetallic electrochemistry.

Developing sustainable and cost-effective electrocatalysts for the oxygen evolution reaction (OER) is essential for efficient water splitting. Herein, Ni-doped Co₃O₄ supported on graphene (Ni-Co₃O₄@G) is synthesized using cobalt recovered from spent lithium-ion batteries via a hydrothermal–calcination approach. The composite exhibits a hierarchical porous structure with uniformly dispersed nanoparticles on conductive graphene sheets, providing abundant active sites and enhanced charge transfer. XPS analysis confirms mixed valence states (Co²⁺/Co³⁺ and Ni²⁺/Ni³⁺) and rich oxygen defects, contributing to improved catalytic activity. Electrochemical results in alkaline media show that Ni-Co3O4@G achieves superior OER performance, delivering a low overpotential (1.66 V vs. RHE at 10 mA cm⁻²), a small Tafel slope (107.52 mV dec⁻¹), and excellent stability over 3000 cycles. This work demonstrates a green strategy for converting recycled cobalt into efficient OER electrocatalysts for sustainable hydrogen production.