The effective treatment of iodine-containing wastewater is essential for environmental protection, as elevated iodide concentrations can adversely affect ecosystems and human health. Traditional methods for iodide monitoring and pH adjustment often involve time consuming multi step procedures, limiting their efficiency in real time wastewater management. In this study, we introduce a deep learning assisted analytical strategy that enables the simultaneous determination of iodide concentration and pH from a single fluorescence lifetime decay curve. The approach is based on the dual sensitivity of fluorescein sodium, which exhibits distinct fluorescence responses in different environments: protonation occurs under acidic conditions, deprotonation under alkaline conditions, and dynamic fluorescence quenching is triggered by iodides. By integrating time correlated single-photon counting (TCSPC) with a hybrid convolutional neural network-residual network-random forest regression (CNN-ResNet-RF) model, we achieve rapid and accurate prediction of both parameters without chemical separation or additional reagents. The proposed model not only overcomes the nonlinear coupling between pH and iodide effects on fluorescence lifetime but also demonstrates strong generalization performance in simulated wastewater matrices. This work provides a robust, reagent-free, and high-throughput analytical platform for real-time monitoring and process control in iodine-containing wastewater treatment.

Sensitive, rapid, and cost-effective detection of Fe³⁺ is urgently required for biological research, food safety, and environmental monitoring. Herein, novel yellow-fluorescent carbon dots (CDs) were successfully synthesized via a facile one-step solvothermal method using ursolic acid and o-phenylenediamine as green precursors. The as-prepared CDs, with an average particle size of 4.82 nm, exhibit bright yellow fluorescence (λem = 540 nm) and high selectivity toward Fe³⁺. The fluorescence intensity is linearly quenched by Fe³⁺ in the range of 0–150 µmol/L, with a low detection limit of 0.096 µmol/L; meanwhile, the fluorescence lifetime remains unchanged, indicating a static quenching mechanism. The proposed sensor displays excellent stability and anti-interference performance, and has been successfully applied to Fe³⁺ detection in real spinach and lake water samples, with satisfactory recoveries of 98

This study presents a novel ratiometric fluorescent sensor based on nitrogen and sulfur co‑doped carbon quantum dots (N, S‑CQDs) for detection of azithromycin (AZM), a widely used veterinary macrolide antibiotic whose residues in dairy products and aquatic environments pose potential health risks. The N, S‑CQDs were synthesized in a one‑step hydrothermal process using 2,5‑diaminobenzenesulfonic acid as the sole precursor, affording a high quantum yield of 39.7

To tackle the challenge of heavy-metal determination is with overlapping fluorescence responses, a multichannel fluorescent sensor array was constructed using three silver nanoclusters (AgNCs) with different surface ligands [histidine (His), poly(methyl vinyl ether-alt-maleic acid) (PMVEM), and poly(methacrylic acid) (PMAA)]. The fluorescence responses toward seven typical metal ions (Al3+, Cu2+, Fe3+, Hg2+, Cr3+, Cd2+, and Pb2+) at multiple concentration levels were converted into normalized spectral fingerprints and analyzed using an automated machine learning framework (AutoGluon), which integrates Top-K feature selection, multiclass ion classification, and ion-specific concentration regression. The conventional single-probe LODs ranged from 0.16 to 28.29 µM, and the AutoGluon-assisted workflow achieved 100

Neotame is a widely used artificial sweetener in milk beverages. However, excessive intake may pose potential health risks, necessitating the development of reliable detection methods. In this work, carbon dots (CDs) were designed for fluorescence enhancement detection of neotame in milk beverages. The CDs were synthesized via a one-pot solvothermal reaction at 180 degrees C for 8 h using 2-hydroxy-3-naphthalic acid and phthalaldehyde as precursors in ethanol. The obtained CDs demonstrated excellent storage stability, maintaining their performance for over six months at 12 degrees C under light-proof and sealed conditions, which exhibited their cost-effectiveness and practical utility. Upon the addition of neotame to a CDs solution diluted with ethanol, a new fluorescence emission peak emerged at 515 nm. Systematic investigation revealed that the fluorescence enhancement is attributed to modification of the surface defect states of the CDs, which leads to an increased radiative transition rate. In milk beverages, the fluorescence intensity at 515 nm showed a good linear response to neotame concentrations ranging from 2 to 100 mu M, with a calculated detection limit of 1.55 mu M (0.58 mg/L). The proposed method exhibits high specificity and strong anti-interference capability, and its application potential in milk beverages has been verified.

Blue-fluorescent carbon dots (CDs) were synthesized via a one-step hydrothermal method using 3,4-diaminobenzoic acid and citric acid as precursors. The as-prepared CDs were comprehensively characterized, revealing abundant surface functional groups, uniform particle size, a negatively charged surface, and excitation-independent emission behavior. By varying the ratio of the precursors, the isoelectric point of the CDs could be adjusted, thereby regulating their binding capacity with other substances. A positively charged fluorescent probe was constructed by electrostatically assembling the CDs with the cationic polymer polyethylenimine (PEI). This post‑synthetic surface modification strategy offered better controllability and tunability compared to directly using PEI as a precursor during CD synthesis. Furthermore, this probe was applied to RNA detection, demonstrating a good linear response with a detection limit of 65 µg/mL. Notably, it exhibited negligible response to structurally similar DNA, confirming its high selectivity for RNA. This study provides ideas for exploring and improving surface modification methods for CDs, and has broad prospects.

This study employed Raman spectroscopy combined with characteristic-peak-intensity-ratio analysis and a novel difference-spectral-statistical-mapping (DSSM) method to elucidate dynamic functional group transformations in flaxseed oil during various frying temperatures (100-185 °C) and repeated frying frequencies (1-5 times). The multiple linear regression models of I969/I1266 with I969/I1300 and I969/I1441 exhibit high correlation coefficients (R2) between 0.93 and 0.94, suggesting that the changes in trans-CC are closely related to the variations in cis-CC and methylene (CH2), which provides key insights for optimizing industrial detection models and fundamental oxidation kinetics research. DSSM identified hidden oxidation features, including epoxy groups and conjugated trienes, visualizing their evolution trends, offering critical spectroscopic evidence for assessing the quality degradation of frying oils. These methods provide theoretical support for real-time and rapid monitoring of the dynamic changes of functional groups in frying oil, taking a key step towards industrial level real-time monitoring of Raman spectroscopy technology.

Fluorescent metal nanoclusters (FMNCs) have attracted significant attention in fluorescence sensor research due to their unique optical properties. Herein, glutathione-protected copper nanoclusters (GSH-CuNCs) were rapidly synthesized within 10 min by direct mixing of glutathione and copper nitrate solutions under weakly acidic conditions (pH 5-6). This as-prepared GSH-CuNCs exhibit a pronounced aggregation-induced emission (AIE) effect in ethanol solution. The generated fluorescence excited at-330 nm was found to be sensitive to the presence of chromium ions (Cr3+), which can be significant quenched by the addition of Cr3+. A fluorescent probe was designed for on-off detection of Cr3+ based on as-synthesized GSH-CuNCs. A linear relationship between fluorescence intensity and Cr3+ concentration was observed, with a detection limit of 600 nM. This probe is simple and rapid to fabricate, enabling fast and sensitive detection of Cr3+. It holds significant potential for practical applications in environmental analysis.

A surface & hybull;enhanced Raman scattering (SERS) composite substrate was designed and prepared based on Ag nanoparticle & hybull;composited ZIF & hybull;8 & hybull;coated perovskite nanocrystals (Ag@CsPbBr3@ZIF & hybull;8). This composite substrate exhibited excellent water stability and, compared to a pure Ag substrate, significantly enhanced the Raman signal of pyrene molecules in aqueous solution. Using spectral analysis, a fitting equation was derived, revealing that the composite substrate enabled highly sensitive detection of pyrene with a detection limit of 8.58 mu g L-1. In spiked recovery tests, the recovery rate of pyrene ranged from 97.6% to 109.2%, with a relative standard deviation (RSD) between 1.12% and 7.91%. These results indicate that the substrate possesses good reproducibility and specificity for pyrene detection.

Circularly permuted yellow fluorescent protein (cpYFP)-based biosensors are powerful tools for biological sensing but are inherently pH-sensitive. This complicates quantitative analysis because the fluorescence signal convolves the effects of the target analyte with ambient pH fluctuations, hindering accurate disentanglement of their individual contributions. Achieving simultaneous monitoring of analyte concentration and pH is therefore essential to unlock the full potential of these biosensors. Here, using the cpYFP-based NADH biosensor SoNar as a model, we present a deep learning framework for the concurrent determination of NADH concentration and pH from its time-resolved fluorescence spectra. We address the critical issue of limited experimental data by developing a hybrid Generative Adversarial Network (GAN) augmented with a Long Short-Term Memory (LSTM) module for high-fidelity spectral data augmentation. This approach significantly enhances prediction accuracy. Our work establishes a robust strategy for dual-parameter quantification with cpYFP-based biosensors and provides a data-efficient solution that reduces reliance on extensive experimental trials.

In this study, nitrogen-doped carbon nanodots (N-CDs) with temperature and fluorescence sensing were prepared via hydrothermal method using L-lysine and ethylenediamine as precursors. The synthesized N-CDs exhibited spherical morphology with sizes ranging from 2.8 to 5.2 nm, with an average diameter of 4.03 nm. Maximum fluorescence emission was observed at 390 nm upon excitation at 320 nm, with the excitation spectrum closely overlapping the absorption spectrum of tinidazole (TNZ). In the temperature range of 20 50 °C, the fluorescence intensity of N-CDs decreased linearly with the increase of temperature. TNZ was detected based on inner filter effect (IFE) using N-CDs as a fluorescent probe. The fluorescence quenching degree had a good linear correlation with the TNZ concentration in the range of 1 100 µM (r = 0.9970), and the detection limit was 0.362 µM. In addition, the detection limits of other nitroimidazole antibiotics, including metronidazole (MNZ), Ornidazole (OMZ) and Seknidazole (SNZ), were 0.324 µM, 0.345 µM and 0.341 µM, respectively. Importantly, this method exhibits minimal interference from ions present in milk and has been validated in real milk samples, with recovery rates ranging from 92.56 to 107.27

Tetracycline antibiotics, valued for potent antibacterial effects, are widely used in livestock but raise concerns over unsafe residues in milk. In this study, a surface-enhanced Raman spectroscopy (SERS) method based on an amino-functionalized covalent organic framework (COF) was developed for the detection of trace tetracycline (TTC) and oxytetracycline (OTC) in milk. The COF material acted as a selective ligand, effectively mitigating interference from the milk matrix. In addition, hydrogen bonding between the COF and the antibiotic molecules contributed to the chemical enhancement of the Raman signal. The proposed method achieved low detection limits of 0.05 μg/L for TTC and 0.07 μg/L for OTC. Owing to their structural similarity, principal component analysis was employed for dimensionality reduction, combined with a support vector machine classification algorithm, which enabled accurate discrimination of the antibiotics with a classification accuracy of 100 % based on the spectral data.

Carbon dots are a new type of luminescent nanomaterial with advantages such as fast response, high sensitivity, and good biocompatibility, which makes carbon dots have great application prospects in drug delivery, optoelectronic devices, biological imaging, and other fields. However, most of the luminescent mechanisms are controversial. In this work, we revealed the luminescence mechanism of carbon dots through fluorescence spectroscopy and transient absorption spectroscopy. The added iron ions bind with carbon dots on the surface to form more active sites, and the increase in active sites makes the detection of cysteine more sensitive, with a detection limit of 1.033 mu M. The research of this work not only provides ideas for exploring and improving the mechanism of carbon dot luminescence, but also provides new methods for biological detection, with great application prospects.

The detection of pefloxacin (PEF) in milk is of significant importance for food safety. Traditional methods for detecting PEF often rely on techniques such as high-performance liquid chromatography (HPLC), which are complex and time-consuming. This article introduces a surface-enhanced Raman scattering (SERS) substrate based on a composite of covalent organic frameworks (COFs) and silver nanoparticles (AgNPs) for the rapid detection of PEF in milk. After testing PEF in both water and milk, the detection limits were calculated to be 6.31 μg/L for water and 82 μg/L for milk, with correlation coefficients of 0.991 and 0.998, respectively. Furthermore, the average recovery rate of PEF in milk ranged from 96.25

Photoluminescent copper nanoclusters have broad application prospects in lighting and display. However, the difficulty of photoluminescence wavelength regulation greatly limits its practical application. Now, more and more studies show that the photoluminescence of metal nanoclusters cannot be simply attributed to the quantum confinement effect of the metal core, metal-metal, metal-ligand and ligand-ligand interactions play a pivotal role in the emission process. Achieving the effective regulation of these weak interactions in metal nanoclusters is the current research focus in this field. Self-assembly is an effective strategy to regulate these weak interactions in metal nanoclusters. The variation of the spatial assembly structure of metal nanoclusters will affect the charge and energy transfer process, and then affect the photoelectric properties of metal nanoclusters. Although many strategies have been proposed to regulate the assembly structure of metal nanoclusters, most of the proposed strategies need to be carried out under heating conditions, which is not conducive to the large-scale production of metal nanocluster assemblies and limits its practical application. In this paper, copper nanocluster assemblies with yellow and blue emission have been successfully synthesized by a solvent-regulated strategy at room temperature. Different solvent environments lead to different assembly modes of copper nanoclusters, and finally two typical assembly structures of nanosheet and nanorod are formed. In a high boiling solvent, such as dibenzyl ether, the solubility and fluidity of copper clusters are poor, which is conducive to the formation of loose nanosheet assembly structure. In contrast, in a low boiling solvent, such as n-hexane, the solubility and fluidity of copper clusters are better and the clusters tend to form a dense nanorod assembly structure. The spacing of clusters in the assembly plays a pivotal role in its photoluminescence performance. Compared with the single-layer nanosheet structure, the nanorod structure adds the additional interlayer interaction. This results in additional Cu(I)...Cu(I) interaction and the increasing of spacing between adjacent sulfhydryl ligands on the cluster surface. Finally, the emission wavelength of copper nanoclusters blue shifted from ca. 550 nm to ca. 490 nm. This solvent-regulated synthesis strategy is simple, easy to operate and short in time consuming, which is conducive to the large-scale synthesis of copper nanoclusters. In addition, light-emitting diodes (LEDs) with different emission colors based on the synthesized metal nanocluster assemblies were successfully prepared.

Tylosin (TYL) is widely used to treat respiratory and intestinal infections in cattle and sheep. However, its residues in dairy products may pose risks to human health. Therefore, it is necessary to establish a sensitive, rapid and highly selective method for TYL detection. Herein, a method for detecting TYL in milk using enrofloxacin (ENR) as a fluorescence probe was developed. We found that TYL effectively quenched the fluorescence of ENR through a synergistic mechanism involving electrostatic interactions and the inner filter effect (IFE). The fluorescence quenching rate exhibited a linear correlation with TYL concentrations in the range of 0.25-25 mg/L, with a low detection limit (LOD) of 10.96 mu g/L. This method innovatively employs ENR as a fluorescence probe to detect TYL, establishing the first feasible approach for reciprocal antibiotic detection while avoiding the complex synthesis and modification processes required for conventional probes. Furthermore, the method was successfully validated in milk.

In this paper, a lanthanide complex-based fluorescent sensor Tb(4-MBA) was developed for the selective recognition of diabetic ketoacidosis (DKA) and the diabetes biomarker β-hydroxybutyric (β-Hb). β-Hb significantly enhanced the fluorescence emission of the Tb(4-MBA) complex at 539  nm. Based on the analysis of the surface electrostatic potential distribution and time-resolved spectra, we speculate that in the reaction system of β-Hb with Tb(4-MBA), β-Hb and Tb(4-MBA) may form a complex through hydrogen bonding interactions, which brings β-Hb closer to Tb3+ and thus reduces the non-radiative energy loss of the solvent molecules to Tb3+ and significantly enhances the Tb(4-MBA) fluorescence intensity. The linear range of Tb(4-MBA) for β-Hb was 2-55 μM, and the limit of detection (LOD) was 50.6 nM. This sensor has high sensitivity and selectivity and shows great potential in the field of screening and diagnosis of diabetes mellitus and DKA.

Accurate monitoring of acid value (AV) is critical for edible oil quality control, yet traditional chemometric methods often face limitations in handling complex spectral data. This study combines Raman spectroscopy with deep learning, including Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Transformer, to explore their potential in improving the accuracy and efficiency of AV quantification during the thermal oxidation of palm oil. The results showed that all three deep learning models outperformed traditional chemometric methods in predictive accuracy. The CNN-LSTM model achieved the best performance, with a predicted coefficient of determination (Rp2) of 0.9978, a mean square error of prediction (RMSEP) of 0.0015, and a residual predictive deviation (RPD) of 21.21. This method demonstrates the effectiveness of Raman spectroscopy-driven deep learning for precise AV monitoring and holds promise for further validation with more diverse indicator datasets, providing a novel technical reference for edible oil quality control.

In this study, N, S co-doped carbon dots (N, S-CDs) with enhanced quantum yields were synthesized via a onestep hydrothermal method, using sodium citrate and glutathione as precursors. Formaldehyde (FA) induces aggregation of N, S-CDs, resulting in fluorescence quenching and the color of the solution turning pink. The Stern-Volmer equation fitting revealed that the quenching mechanism was static quenching at low FA concentrations, while both static and dynamic quenching existed at high FA concentrations. Using N, S-CDs as probe, FA could be detected under three modes: fluorescence intensity, colorimetric, and fluorescence lifetime. Linear detection ranges were successfully established within the FA concentration ranges of 0.1-50/200-600 mu M, 1-100 mu M, and 200-500 mu M, respectively. Acceptable recoveries ranging from 97.78 % to 104.12 % were achieved in real milk samples, and the relative standard deviation (RSD) value was less than 3.84 %. The detection of FA in mushroom, vermicelli and baby cabbage matched the standard Nash method, indicating that this method has great potential for FA detection in real samples.