N-nitrosamines (NAs) are a class of strongly carcinogenic compounds with high contamination levels in aquatic products, posing significant risks to human health. Recent random inspections in China have revealed that excessive N-nitrosodimethylamine (NDMA) content has become a major factor disqualifying processed aquatic products, seriously hindering the healthy development of the aquatic product processing industry. Therefore, elucidating the formation mechanisms and influencing factors of NAs in aquatic products, along with controlling their levels through processing technologies and multidisciplinary approaches, has become a key research focus. This review systematically examined various formation pathways of NAs in aquatic products, pretreatment methods, and analytical techniques compliant with national food safety standards. It further summarized comprehensive strategies for NAs control, including reduction of NAs precursor intake, blocking of NAs formation pathways, and decomposition of existing NAs. The paper aimed to address the excessive NAs levels in aquatic products, thereby ensuring the quality and safety of aquatic products and promoting sustainable development of the industry.

Food safety is crucial for national economic development and public health. With the development of science and technology, traditional detection methods is difficult to meet the supervision requirements of food safety in complex scenarios. In recent years, artificial intelligence (AI), with its high efficiency, high accuracy, powerful data processing capabilities, and precise identification and predictive analysis functions, has gradually been applied to research fields such as food flavor analysis, food safety risk assessment and early warning, and authenticity identification. The application of AI technology has improved the timeliness and reliability of food quality and safety testing, and has built a prevention and control system covering the entire industry for regulatory departments through intelligent decision-making assistance systems, becoming the core technical support for ensuring “safety on the tip of the tongue”. This paper mainly introduced the main algorithms of machine learning and deep learning algorithms, and reviewed the research progress of AI algorithms in food quality and safety. Additionally, it looks ahead to the existing problems and future research trends of the application of AI algorithms in food quality and safety. The aim is to offer references for further promoting the research and practice of AI technology in the field of food quality and safety, and to provide research ideas for the field of food quality and safety detection.

Tea is deeply loved by consumers for its unique aroma and health benefits, which is why tea quality and safety have garnered widespread attention. The quality and safety of tea involves the quality and safety of tea, as well as issues such as pesticide residues, heavy metals, harmful microorganisms, and fluorine. It is closely related to all aspects of tea cultivation, processing and storage. This paper started with tea-related standards and focuses on tea quality and safety issues such as pesticide residues, heavy metals, harmful microorganisms and fluoride, etc. It summarized the pretreatment of tea samples, normal detection technologies, rapid detecting methods, and relevant research progress. The aim is to provide a comprehensive knowledge framework for the tea industry, enhance quality and safety standards, and promote the sustainable development of the tea industry.

Objective To investigate false positive interference from spices in hot pot base materials during the detection of thebaine using mass spectrometry, identify interfering matrix components, and optimize analytical methods. Methods The 62 kinds of commonly used spices in hot pot bases were collected. The false positive issues screened by BJS 201802 Determination method of morphine, codeine, papaverine, noscapine and thebaine in food were further analyzed by Fourier transform ion cyclotron resonance mass spectrometry combined with positive pressure electrospray ionization full mass spectrometry scanning. The potential interfering substances were locked through ion fragment analysis and verified. Results The results showed that the detection found interference peaks with retention times similar to those of thebaine in the extracts of bay leaves and pepper, which could lead to false positive results. The molecular weight of the interfering substance was 311.3812 ([M+H]+), with a deviation of less than 2 ppm from the theoretical value of thebaine (311.375), suggesting that it was an isomer. Through the detection of the suspected interfering substance, it was verified that Launobine and Pisatin had no interfering activity. Conclusion The alkaloid substances in spices have significant interference with the detection of thebaine. It is suggested to improve the specificity and accuracy of the detection by optimizing the mass spectrometry conditions, introducing nuclear magnetic resonance technology and perfecting the pretreatment process to avoid false positive misjudgment.

Objective To establish a method for the determination of the migration of 14 kinds of ultraviolet absorbers in aqueous food simulants and chemical alternative solvents (95% ethanol aqueous solution and isooctane) by high performance liquid chromatography. Methods The ultraviolet absorbers migrated into the aqueous food simulants were extracted twice with n-hexane by vortexing and shaking, and the combined extract was evaporated to dryness under reduced pressure and re-dissolved in tetrahydrofuran. The ultraviolet absorbers migrated into the chemical alternative solvents were evaporated to dryness and re-dissolved in tetrahydrofuran. A C₈ column was used for separation, with acetonitrile and water as the mobile phase and gradient elution. Detection was performed using a ultraviolet detector at a wavelength of 300 nm. Results The limits of detection of 14 kinds of ultraviolet absorbers were as low as 0.008 mg/L, and the limits of quantitation were 0.020 mg/L. The correlation coefficients of the 14 kinds of target compounds were above 0.999 within the linear range of 0.20-5.0 mg/L. The 3 levels of ultraviolet absorbers (0.02, 0.05 and 0.50 mg/L) were added to the food simulants or chemical alternative solvents, and 6 replicate experiments were conducted. The results showed that the recovery rates of each compound at each level were within the range of 82.4%-108.0%, and the relative standard deviations were less than 10%. Conclusion This method is accurate, efficient and simple, and can be used to simultaneously determine the migration amounts of 14 kinds of ultraviolet absorbers in aqueous simulants and chemical alternative solvents.

Objective To establish a method for determination the fat content in flavored fermented milk containing thickeners by papain-gravimetric method. Methods Papain was used to reduce the adsorption of fat by casein in dairy products. The amount of papain added was 0.5 g, the enzymatic hydrolysis temperature was 65 ℃, and the hydrolysis time was 20 min. A mixture of ether and petroleum ether with equal volumes [25:25 (V:V)] was used to extract fat twice from the flavored fermented milk hydrolysate containing thickening agents. The extraction solvent was evaporated in a water bath, and the fat content was quantified by gravimetric method after drying in an oven. Results The determination of fat content in flavored fermented milk with thickener flavor was consistent with the label value, with a relative standard deviation of 0.40% and a maximum deviation of 0.05 g/100 g, which met the precision requirements of GB 5009.6—2016 third method for determination of fat in foods. Conclusion This method is simple to operate and conform to the concept of green analytical chemistry, compared with GB 5009.6— 2016 third method, it reduces the total volume of mixed ethers used, thereby minimizing harm to human health and environmental pollution. With good reproducibility, it can effectively determine the fat content in flavored fermented milk containing thickeners, providing robust technical support for quality control of such products.

Objective To detect and analyze the content of mineral elements in wines of different varieties and years from the Helan Mountain East Slope region, and construct a model for identifying wine varieties and years based on mineral elements using multivariate statistical analysis methods. Methods Grapes of 5 varieties, namely, Zinfandel, Merlot, Malbec, Cabernet Sauvignon and Marselan, were collected from a winery in Qingtongxia, Ningxia, and were processed into wines using the same method. Wines of the Cabernet Sauvignon variety from 6 years (2018—2023) were collected from a winery in Yongning, Ningxia. The content of 54 kinds of mineral elements including B, Bi, Be, etc. in the wines was determined by inductively coupled plasma mass spectrometry. A model for identifying wine varieties and years was constructed based on the differences in mineral element content among different varieties and years of wines. Results A total of 53 mineral elements were detected in wines of different varieties, and 47 kinds of mineral elements including As and B showed extremely significant differences among varieties (P<0.01). All 54 kinds of mineral elements were detected in wines of different years, and all showed extremely significant differences among years (P<0.01). The correct discrimination rate of the model for identifying wine varieties and years based on Fisher discriminant analysis was 100%. The 9 kinds of mineral elements, namely, As, Cs, Dy, Mo, Na, Sb, U, Yb and Fe, were identified as the differential elements for different varieties of wines, and 11 kinds of mineral elements, namely, Cu, Li, Na, Sb, Tl, U, V, Y, Yb, Ag and Pt, were identified as the differential elements for wines of the same variety from different years. The correct discrimination of wines from the same origin but of different varieties and years was achieved based on mineral elements. Conclusion Mineral elements show significant differences among wines of different varieties and years, and can be used for the identification of wine varieties and years from the same region.

Objective To establish a pre-amplified quantitative real-time polymerase chain reaction (qPCR) method for the detection of Oncorhynchus mykiss source components. Methods The study developed an enhanced pre-amplified qPCR method, which was based on the validation and improvement of the existing standard qPCR method and compared the performance of both methods. Results Specificity testing of the standard qPCR method revealed cross-reactivity with closely related species, such as Coho salmon and Atlantic salmon. The established improved pre-amplified qPCR method, which incorporated a pre-amplification step, exhibited high specificity and no cross amplification was observed for Coho salmon, Atlantic salmon and other closely related species. The amplification efficiencies of the standard qPCR and the pre-amplified qPCR methods were 90.25% and 104.91%, respectively. The lowest limits of detection of the 2 kinds of methods were 2.99×10^-3 ng/μL (0.83-8.87 pg/μL, 95% confidence interval) for the standard qPCR and 1.50×10^-4 ng/μL (0.09-0.27 pg/μL, 95% confidence interval) for the pre-amplified qPCR. Conclusion The method established in this study has the characteristics of strong specificity, high sensitivity, and is suitable for authenticity and adulteration detection of Oncorhynchus mykiss source components in food products.

Objective To analyze the key factors affecting constant weight in the direct drying method and to optimize the experimental conditions for vacuum drying, distillation, and Karl Fischer methods as specified in the GB 5009.3—2016 National food safety standard-Determination of moisture in foods. Methods The determination of moisture content in samples was conducted according to GB 5009.3—2016. The study altered experimental conditions to explore key factors affecting constant weight in the direct drying method, optimized conditions for vacuum drying by adjusting vacuum pressure, and compared the effects of different reagents (xylene, Karl Fischer reagent) on detection results. Results In the direct drying method, sample temperature, cooling time in the desiccator and the degree of sample pulverization were identified as key factors affecting constant weight. For vacuum drying method, increasing the vacuum to -0.1 MPa and adding an effective desiccant to the drying box could rapidly evaporate moisture from the sample, improving detection efficiency. In the distillation method, the preparation and purification of analytical-grade toluene or xylene could be omitted. The Karl Fischer method could use a low-toxicity, pyridine-free Karl Fischer reagent. Conclusion The 4 detection methods in this standard are optimized separately to solve problems such as difficulty in achieving constant weight after repeated drying and the use of toxic and harmful reagents. The rationalization suggestions provide in this study have increased the operability of the standard and improved the testing efficiency of the laboratory.

Objective To understand the levels of veterinary drug residues in chickens and eggs in the diet of Xinjiang residents, and analyze and assess the health risks associated with exposure to veterinary drug residues in this region. Methods Veterinary drug residues were detected in chicken and egg samples from supermarkets and farmers’ markets in Xinjiang Uygur Autonomous Region from 2021 to 2023 by liquid chromatography-tandem mass spectrometry (LC-MS/MS), and then combined with the data on the food intake of the residents of Xinjiang Uygur Autonomous Region, and then assessed the risk of dietary exposure by point assessment method. Results The detection rate of 19 kinds of veterinary drug residues in chicken and egg samples in Xinjiang from 2021 to 2023 was 23.3% (63/270), and the exceedance rate was 6.3% (17/270), and the difference in the detection rate of florfenicol was statistically significant by chi-square test (P=0.002), and the differences in detection and exceedance rates in the Northern, Southern and Eastern regions of Xinjiang were not statistically significant. The food safety index for chicken meat in Xinjiang were much smaller than 1, and the food safety index for veterinary drug residues in eggs were all less than 1, but the food safety index for ofloxacin, enrofloxacin, ciprofloxacin, diclazuril and florfenicolide were much greater than 1. Conclusion The risk of dietary exposure to veterinary drug residues for chicken meats and eggs in Xinjiang is small and in the controllable range.