NH2-PEG-AuNR-Based Surface-Enhanced Raman Spectroscopy for Rapid Detection of AFB1: Dark Tea Safety Assessment

NH2-PEG-AuNR-Based Surface-Enhanced Raman Spectroscopy for Rapid Detection of AFB1: Dark Tea Safety Assessment | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article NH2-PEG-AuNR-Based Surface-Enhanced Raman Spectroscopy for Rapid Detection of AFB1: Dark Tea Safety Assessment Shuci Cao, Yifan Zuo, Qianfeng Yang, Yongning Wei, Qisheng Liu, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7144899/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract There is a risk of aflatoxin B1 (AFB1) contamination during dark tea processing, transportation, and storage. Given the potent carcinogenicity of AFB1, there is an urgent need to develop rapid and sensitive detection methods to address the requirements of tea safety monitoring. A rapid and trace detection method for AFB1 based on surface-enhanced Raman spectroscopy (SERS) using NH2-PEG-AuNRs was established and successfully applied to dark tea. The established method exhibited a linear dynamic range (LDR) for the AFB1 standard solution from 10⁻¹ to 10⁻¹⁷ ng/mL, with a limit of detection (LOD) of 10⁻¹⁷ ng/mL. The partial least-squares (PLS) prediction model achieved an accuracy of 99.36%. For dark tea infusions, the LDR extended from 10⁻¹ to 10⁻⁹ ng/mL, and the LOD of 10⁻⁹ ng/mL. The PLS model achieved an accuracy of 97.43%, and recovery rates ranged from 98.54–102.54%. This establishes a robust methodology for quantifying AFB1 in dark tea. Biological sciences/Biochemistry Biological sciences/Biological techniques Biological sciences/Biotechnology Physical sciences/Chemistry Earth and environmental sciences/Environmental sciences Aflatoxin B1 Dark tea Amino-terminally modified polyethylene glycol gold nanorods Surface-enhanced Raman scattering Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Dark tea is widely consumed globally because of its distinctive flavour and numerous health-promoting components 1 , 2 . The pile fermentation process of dark tea is strongly influenced by microorganisms, which play a crucial role in determining the quality and flavour profile of dark tea. Aspergillus spp. and Penicillium spp are the main microorganisms involved in this fermentation process 3 . Microbial toxin contamination can severely affect the quality and safety of dark tea. The risk of such contamination arises from mycotoxins produced during pile fermentation, spoilage-like substances produced during mould decomposition, and microbial toxins generated through interactions between environmental microorganisms and those present in the tea that occur when tea is improperly stored 3 . Aflatoxin B1 (AFB1) accounts for 75% of the mycotoxin contamination that occurs in food and feed 4 . Aflatoxins, which are primarily produced by Aspergillus fungi, particularly Aspergillus flavus and Aspergillus parasiticus , can be classified into four major subtypes: AFB1, AFB2, AFG1, and AFG2. Among these, AFB1 is the most common aflatoxin monomer and has the highest toxicity 5 . AFB1 is physicochemically stable, has poor water solubility, and is readily soluble in organic solvents such as methanol, ethanol, and acetonitrile. Regarding its structure, it contains a double-furan ring with a double bond that undergoes epoxidation, which facilitates its interaction with nucleic acids, proteins, and other biological macromolecules, leading to toxic effects. Additionally, AFB1 contains an oxonaphthalene–octanone (coumarin) ring that contributes to AFB1’s carcinogenicity. Because of its potent toxicity and carcinogenicity, AFB1 has been classified as a class I carcinogen by the International Agency for Research on Cancer 6 . Therefore, a sensitive and accurate method for the rapid detection of AFB1 in dark tea must be developed to ensure product safety and support the sustainable development of the tea industry. Numerous analytical techniques have been developed for detecting AFB1, including high performance liquid chromatography (HPLC) 7 , enzyme-linked immunosorbent assay 8 and liquid chromatography-mass spectrometry (LC-MS) 9 . Although these assays provide highly accurate quantitative results, they are often complex, time-consuming, and costly, and therefore, they cannot be used for rapid, on-site detection. In consideration of these limitations, a growing number of studies has focused on developing nanomaterial-based sensors for AFB1 detection. For instance, Wang developed a high performance electrochemiluminescence sensor based on ultra-stable perovskite quantum dots@ZIF-8 composites for the detection of AFB1 in maize 10 . Additionally, Zhang developed a composite gold nanophotoelectrochemical aptamer–based sensor for AFB1 detection and successfully applied it in peanuts and wheat 11 . Although electrochemical techniques offer advantages such as high sensitivity and rapid response times, electrochemical sensors are highly susceptible to interference from complex sample matrices, which can lead to cross-reactivity and reduced detection accuracy. Chen identified AFB1 in wheat by using a composite nanocolourimetric sensor array 12 , whereas Lu developed a colorimetric sensor based on gold nanoparticles (AuNPs) and a smartphone for detection of AFB1 in beans 13 . This nanosensor-based colorimetric method relied on evaluation of colour changes before and after a reaction. Although they achieved visualisation, the colorimetric results were highly influenced by ambient lighting conditions and instrument variability, resulting in low sensitivity. Dou developed a fluorescence method by using a ratiometric NMOFs-Aptasensor based on fluorescence resonance energy transfer (FRET) for detecting AFB1 in maize 14 . Similarly, Lu developed a FRET-based fluorescence method by using CdZnTe quantum dots and AuNPs for detecting AFB1 and successfully achieved such detection in peanuts 15 . Fluorescence methods enable multichannel detection of AFB1 through fluorescence signals at different wavelengths, which can enhance accuracy and reliability. However, these methods typically require complex sample pretreatment, which limits their practical application. In consideration of these constraints, a nanosensor must be developed for trace detection of AFB1 that is both simple to operate and resistant to interference from sample matrices. Surface-enhanced Raman scattering (SERS) is an advanced spectroscopic technique that provides a unique molecular ‘fingerprint’ for an analyte. SERS offers the benefits of accurate trace detection, high resolution, and minimal sample consumption. Among the various SERS substrates that have been proposed, nanomaterials composed of gold and silver are the most efficient hotspot sensors and are widely employed for the trace detection of AFB1 in food 16 . AFB1 detection using precious metal nanomaterials is generally conducted using one of two methods. The first involves the direct detection of AFB1 by using nanomaterials; however, this approach is often associated with signal instability and poor reproducibility. The second method involves modifying nanomaterials to facilitate the detection of target molecules through physical or chemical adsorption on the substrate surface. This approach has gradually replaced direct detection. Notably, aptamer-modified nanomaterial sensors are highly sensitive receptors for specific target detection. Wu detected AFB1 in peanuts by using SH-aptamer-modified AuNP dimers as SERS aptamer sensors 17 . Additionally, He employed SERS aptamers comprising SH-cDNA-modified Fe3O4@Au NFs and SH-Apt-modified Au-4MBA@Ag NSs to detect AFB1 in peanut oil 18 . In addition to aptamer-based approaches, immunosensors have been widely employed in SERS-based detection; these sensors use antibody–antigen interactions to identify targets. Li developed a DSNB-modified AuNP SERS immunosensor for the detection of mycotoxins in maize, rice, and wheat 19 , and Zhang employed DTNB-labelled SRES immunosensors for the detection of mycotoxins in maize, rice, and wheat 2 . Although these sensors have considerably improved detection accuracy, antibody–antigen- and aptamer-based methods are complex. Furthermore, research focused on the trace detection of AFB1 in tea matrices is scarce. Such detection is difficult because of the composition of tea extracts, which form a complex background for detection. Therefore, a simple and easy-to-use method must be developed that enables rapid and accurate quantitative analysis of AFB1 in tea matrices. In consideration of this, the present study developed a simple and rapid method for trace detection of AFB1 in dark tea extracts by using SERS. This approach involved the preparation of highly sensitive, reproducible, and stable SERS substrates, namely, amino-terminally modified polyethylene glycol gold nanorods (NH 2 -PEG-AuNRs). This method enables detection through the use of the hydrogen bonding interactions between NH 2 -PEG-AuNRs and AFB1 (Scheme 1). By integrating Raman spectroscopy with emerging nanotechnology, this study established a simple, sensitive, and accurate method for detecting AFB1 to ensure tea safety. The study findings can be applied to enhance food safety measures and advance the tea industry. 2. Results and Discussion 2 .1 Characterisation of NH 2 -PEG-AuNRs The morphology and homogeneity of NH 2 -PEG-AuNRs were analysed through TEM and scanning electron microscopy. As presented in Figure 1AB, the NH 2 -PEG-AuNRs exhibited a uniform size distribution, with an average size of approximately 20 nm × 80 nm. The elemental composition of the NH 2 -PEG-AuNRs was determined through energy-dispersive X-ray spectroscopy, as indicated in Figure 1C. The analysis confirmed that the samples were mainly composed of Au (red), C (green), and O (blue). A UV-vis absorption spectrum of the NH₂-PEG-AuNRs (Figure 1D) revealed transverse plasmon resonance absorption peaks corresponding to the transverse surface plasmon resonance (TSPR) and the longitudinal plasmon resonance (LSPR). The TSPR peak was observed at approximately 508 nm, whereas the LSPR peak occurred at approximately 808 nm. Additionally, the surface chargeability of the NH 2 -PEG-AuNRs was characterised using zeta potential maps. As presented in Figure 1E, the zeta potential of the NH 2 -PEG-AuNRs was 34.97 mV. 2 .2 SERS Activity of NH 2 -PEG-AuNRs The SERS sensitivity of the NH 2 -PEG-AuNR substrates was evaluated using R6G at various concentrations (10 −3 , 10 −4 , 10 −5 , 10 −6 , 10 −7 , 10 −8 , 10 −9 , 10 −10 , and 10 −11 mol/L). The corresponding SERS spectra for different R6G concentrations are presented in Figure 2A, which reveals characteristic peaks of R6G at 612, 772, 1184, 1311, 1361, 1506, and 1647 cm −1 . The peaks at 612, 1311, 1361, 1506, and 1647 cm −1 were attributable to C–C ring stretching vibrations, whereas those at 772 and 1184 cm −1 were associated with C–H bending vibrations 20 . When the concentration of R6G was greater than 10 −11 mol/L, the characteristic peaks of R6G molecules at 1361 and 1506 cm −1 were still clearly visible. As indicated in Figure 2B, a linear relationship was noted between the R6G characteristic peak intensity at 1506 cm −1 and the concentration of R6G, yielding the equation y = −310.20 x + 3501.76, with a high correlation coefficient of R 2 = 0.9960. This strong linear correlation confirms the high sensitivity of the NH2-PEG-AuNRs prepared in this study. The reproducibility of the NH2-PEG-AuNR substrates was investigated using R6G (10 −7 mol/L) as a model analyte. As presented in Figure 2C, the Raman signal intensity mapping plot of 121 evenly distributed sites within a 100 μm × 100 μm square matrix at 1361 cm −1 revealed a uniform colour distribution. Figure 2D presents the SERS spectra of 20 randomly selected sites on the substrate. The RSD of the Raman signal intensity at 1506 cm −1 was calculated as 4.08%. These findings confirm that the NH2-PEG-AuNR substrates exhibit high reproducibility. The stability of the NH2-PEG-AuNR substrate was evaluated by monitoring the SERS signals of R6G (10 −7 mol/L) every 15 days for 4 months. As presented in Figure 2E, the characteristic peaks of the SERS spectra remained consistent across all time points, indicating that the NH2-PEG-AuNRs are structurally stable and meet the requirements for practical applications. Variations in signal intensity with time over 120 days were recorded, and the RSD was calculated to be 9.75%. This relatively low variation indicates that the NH2-PEG-AuNR substrate prepared in this study has excellent stability. The SERS and Raman spectra of R6G, presented in Figure 2F, illustrate the substantial signal enhancement effect of the NH2-PEG-AuNR substrates on R6G. The EF of the NH2-PEG-AuNRs was calculated using eq. (1) as 4.51 × 10 7 . This result confirms the excellent performance of the NH₂-PEG-AuNRs in SERS detection. 2 .3 Optimisation of the NH2-PEG-AuNR Substrate Mixing Ratio for AFB1 Detection To investigate the potential for enhanced Raman signal detection of AFB1 by using NH2-PEG-AuNR substrates, the Raman spectra of AFB1 solid powder and SERS spectra of AFB1 ethanol solution were obtained (Figure 3A). A comparison of these spectra revealed distinct AFB1 characteristic peaks located at 524, 775, 1075, 1144, 1267, 1496, and 1550 cm −1 , with particularly strong response signals at 1144, 1267, and 1496 cm −1 . To investigate the feasibility of detecting AFB1 in tea matrices, the SERS spectra of AFB1 in ethanol solution, the SERS spectra of AFB1 in tea extracts (at equivalent concentrations), the SERS spectra of tea extracts, and the SERS spectra of the NH2-PEG-AuNR substrate itself were compared, and the results are presented in Figure 3B. The SERS signal intensity of AFB1 in the tea extract was significantly weaker than that of AFB1 in ethanol solution. However, the tea extract itself exhibited no distinct characteristic peaks, and the NH2-PEG-AuNR substrate itself had no effect on the detection results. This suggests that the tea extract matrix may influence the quantity of AFB1 captured by the NH2-PEG-AuNR substrate, although it does not appear to alter the SERS characteristic peaks of AFB1. Further investigation into the influence of different dark tea extracts spiked with AFB1 was conducted, with the corresponding SERS spectra displayed in Figure 3C. Notably, the AFB1 characteristic peak at 1267 cm −1 remained clearly visible across all samples. These findings indicate that variations in tea extract matrices did not exert any obvious effect on AFB1 detection. To assess the effect of different mixing ratios of AFB1 and NH2-PEG-AuNR substrates on Raman signal intensity, a systematic investigation was conducted, as illustrated in Figure 3D. The results indicate that the Raman signal intensity of AFB1 was the highest at a volume ratio of 1:1. However, when the volume ratio either increased or decreased, the signal intensity declined. This may be attributable to the optimal formation of hotspots at the suitable volume ratio, where the interaction between the substrate and the target molecule is most effective. Therefore, a mixing ratio of 1:1 was selected as the optimal condition for subsequent detection experiments. 2 .4 Establishment of AFB1 Standard Curve 2 .4.1 Establishment of the AFB1 Standard Curve in Ethanol Solutions The SERS spectra of AFB1 ethanol solution were recorded across 17 concentrations ranging from 10 −1 to 10 −17 ng/mL, with the lowest instrumental limit of detection (LOD) being 10 −17 ng/mL, as indicated in Figure 4A. Repeated experiments confirmed that the intensity of the spectral peak at 1267 cm −1 was stable, and a linear relationship was observed in the specific concentration range. As presented in Figure 4B, the Raman signal intensity at 1267 cm −1 gradually weakened as the concentration of AFB1 was reduced. A linear correlation was observed between the peak intensity at 1267 cm −1 and the AFB1 concentration in the range of 1 × 10 −1 to 1 × 10 −17 ng/mL, following the equation: y = −57.69 x + 1009.87, R 2 = 0.9971. In the field of rapid detection of toxins, studies regarding SERS-based toxin detection have generally struggled to achieve direct detection using nanomaterials 18 . In the current study, NH2-PEG-AuNRs were employed to bind AFB1. This method offers advantages such as a broad detection range and low detection limit for AFB1 quantification. The experimental results of this study confirmed the high sensitivity of the method for AFB1 detection in standard solutions, providing a strong foundation for its application in the quantitative detection of AFB1 in real samples. 2 .4.2 Establishment of Standard Curve for AFB1 in Pu-erh Tea Extracts The NH2-PEG-AuNR substrates prepared in this study could effectively detect the SERS spectra of AFB1 specimens. Conventional methods for detecting AFB1 in tea are often subject to interference from the complex matrix of tea infusions 3 . However, NH2-PEG-AuNRs can selectively adsorb the AFB1 molecules in tea extracts onto the substrate, which improves SERS detection. Figure 4C presents the SERS spectra of Pu-erh tea extract spiked with AFB1 at concentrations ranging from 1 × 10 −1 to 1 × 10 −9 ng/mL, with a total of 9 concentrations, with an instrumental LOD of 1 × 10 −9 ng/mL. Figure 4D presents the correlation curve between the AFB1 concentration in Pu-erh tea extracts and the intensity of characteristic SERS peak at 1267 cm −1 . Figure 4D presents the correlation curve between the AFB1 concentration in Pu-erh tea extract and the intensity of the characteristic SERS peak at 1267 cm −1 . The established linear regression equation is given by y = −20.41x + 214.52. The correlation coefficient was R 2 = 0.9953, which indicates that this method can effectively detect different concentrations of AFB1 in Pu-erh tea extract. Moreover, the detection limit and sensitivity achieved in this assay comply with regulatory standards established by China and the European Union 4 . 2 .5 Development of a Quantitative Prediction Model for AFB1 2 .5.1 Development of a Quantitative Prediction Model for AFB1 in Standard Solutions In this study, a PLS regression model was developed for the quantitative prediction of AFB1 concentrations in standard solutions. The model’s performance metrics, presented in Table 1, reveal an Rc value of 0.9924, Rp value of 0.9936, and an RPD of 8.93, as calculated using eq. (4). These high statistical values indicate that the PLS model exhibits excellent predictive accuracy, rendering it suitable for the accurate prediction of AFB1 in standard solutions. 2 .5.2 Development of a Quantitative Prediction Model for AFB1 in Real Samples A PLS quantitative prediction model was developed for the quantification of AFB1 in Pu-erh tea extracts. The model’s performance metrics, presented in Table 1, indicate the Rc value of 0.9750, Rp value of 0.9743, and RPD of 4.40, as calculated using eq. (4). These results confirm the high predictive accuracy and reliability of the model for the quantitative prediction of AFB1 in Pu-erh tea extracts. Compared with traditional detection methods 21,22 , the SERS-based approach using NH2-PEG-AuNRs developed in the present study offers the advantages of simplicity, speed, and efficiency for rapid detection of AFB1 in Pu-erh tea extracts. This study established a novel method for rapid quantitative prediction of AFB1 in Pu-erh tea extracts. 2 .5.3 External Validation of the Quantitative Prediction Model The developed PLS quantitative prediction model was applied to spectral data collected from four types of dark tea extracts. The results, summarised in Table 1, indicate that the RPD value for all tested samples exceeded 1.9, confirming the broad applicability of the assay for various dark tea matrices. The recovery rates were calculated as the ratio of the AFB1 concentration detected in the tea extract to the actual spiked concentration, and the results are presented in Table 2. The spiked recovery rates ranged from 98.54% ± 5.55% to 102.54% ± 3.71%, with an RSD of <8.06%. These results indicate the high accuracy and precision of the developed assay, highlighting its robustness and reliability for the quantitative determination of AFB1 in real tea samples. Table 1. PLS quantitative forecasting model results. Samples Model parameter Training sets Prediction sets RPD RMSECV Rc RMSEP Rp AFB1 ethanol solution LVs=7 0.60 0.9924 0.55 0.9936 8.93 Labelling of Pu-erh tea infusion with AFB1 0.58 0.9750 0.59 0.9743 4.40 Qianliang tea 0.58 0.9750 0.35 0.9431 2.39 Fuzhuan tea 0.58 0.9750 0.35 0.9528 2.38 Kangzhaun tea 0.58 0.9750 0.41 0.9165 2.03 Liubao tea 0.58 0.9750 0.43 0.9013 1.91 Table 2. Determination and recovery of AFB1 concentration in different dark teas (n=15). Samples Additive concentration （-lg C ng/mL） Detected concentration （-lg C ng/mL） Recovery rate （%） RSD （%） Qianliang tea 0 0 5.00 4.99±0.26 99.84±5.18 5.19 6.00 6.15±0.22 102.54±3.71 3.62 7.00 7.08±0.23 101.08±3.23 3.20 Fuzhuan tea 0 0 5.00 5.13±0.31 102.63±6.15 6.00 6.00 6.14±0.23 102.49±3.91 3.82 7.00 7.03±0.23 100.44±3.24 3.23 Kangzhaun tea 0 0 5.00 4.96±0.40 99.14±7.99 8.06 6.00 6.19±0.24 103.23±3.96 3.84 7.00 7.07±0.21 101.05±2.94 2.91 Liubao tea 0 0 5.00 4.93±0.28 98.54±5.55 5.63 6.00 6.11±0.28 101.91±4.72 4.63 7.00 7.02±0.23 100.32±3.23 3.22 2.6 Discussion AFB1, a contaminant produced by Aspergillus species, is among the most toxic mycotoxins and is a potential contaminant in the processing, transportation, and storage of Pu-erh tea. In this study, a SERS-based approach to detecting AFB1 was developed that uses the hydrogen bonding interactions between NH2-PEG-AuNRs and AFB1. This method enables rapid and accurate detection of AFB1 in complex tea matrices. Most SERS sensors detect AFB1 in samples through physical or chemical adsorption. However, sensors that rely on chemical adsorption often have the disadvantages of complex preparation processes and limited adsorption efficiency. As indicated in Table S1, the method proposed in the current study offers a broader detection range, lower detection limit, and higher efficiency, addressing a critical gap in the literature with respect to AFB1 detection in tea matrices. Au@AgNPs have previously been employed as a signal-enhancing substrate, with melamine used as an adsorbent; they facilitate AFB1 adsorption in tea oil through hydrogen bonding with a linear range of 10 −4 to 10 −7 mol/L and a detection limit 10 −8 mol/L 23 . However, tea leaf extracts have a more complex background than tea oil does, and therefore, AFB1 detection is more challenging in teas. Compared with Fe3O4@AuNFs-cDNA and Au-4MBA@AgNSs-Apt 18 , NH2-DNA1-CS-Fe3O4 and SH-DNA2-ADANR 24 , NH2-Rh-Au@Ag CSNPs, and AuNP dimers-MXenes assembly aptamer sensors 17 , the NH2-PEG-AuNR-enhanced substrate proposed in this study involves a simpler preparation process and faster AFB1 binding kinetics. Moreover, compared with a previously proposed DSNS-AuNP immunosensor 19 , the NH2-PEG-AuNR substrate prepared in this study is more stable and has a 4-month validity. The method proposed in this study yielded satisfactory results in the detection of AFB1 in tea samples. However, in the real world, microbial toxins are often complex and diverse, and this study did not explore simultaneous detection of multiple microbial toxins. However, in theory, the substrate prepared in this study can adsorb other toxins and enhance the Raman fingerprints of other toxins through hydrogen bonding interactions. Nevertheless, the binding energy between the substrate and mycotoxins and the potential for cross-reactivity among different mycotoxins remain uninvestigated. In the future, we will calculate the binding energies between the substrate and different mycotoxins by using density functional theory and investigate the ability of the substrate to simultaneously detect different mycotoxins in the presence of multiple mycotoxins. The present study successfully developed a quantitative method for detecting AFB1 by using SERS with NH2-PEG-AuNRs and established a rapid and effective approach to AFB1 analysis of dark tea. The proposed method involves a simple pretreatment process; overcomes key challenges associated with aptamer sensors, which have limited binding efficiency for AFB1; and overcomes key challenges associated with immunosensors, which exhibit instability. Additionally, this method involves a broader detection range and a lower detection limit than conventional approaches do. By enhancing the precision and reliability of AFB1 detection, this study makes a notable contribution that can improve the quality control and safety assessment of tea products. This method has potential for broader application in the field of food safety; it can be used to detect other toxins in complex food matrices. This study successfully developed a SERS technique using NH2-PEG-AuNRs that enables rapid and sensitive detection of AFB1 in tea. The NH2-PEG-AuNRs demonstrated excellent SERS activity, leading to a broad linear detection range of 1 × 10 −1 to 1 × 10 −17 ng/mL, with R 2 = 0.9971 and LOD = 1 × 10 −17 ng/mL for AFB1 in standard solutions. Additionally, a PLS quantitative prediction model was established that had a prediction accuracy of 99.36% and an RPD value of 8.93. For the detection of AFB1 in tea extracts, the proposed method yielded a linear range of 1 × 10 −1 to 1 × 10 −9 ng/mL, an R 2 value of 0.9953, and an LOD of 1 × 10 −9 ng/mL. A separate PLS quantitative prediction model for AFB1 in Pu-erh tea extract was developed that had a prediction accuracy of 97.43% and an RPD value of 4.40. External validation experiments further confirmed the robustness of the model, with prediction accuracies exceeding 90.00%, RPD values above 1.9, and recovery rates between 98.54% ± 5.55% and 102.54% ± 3.71%. Notably, the entire detection process can be completed in 20 min, with this including sample–substrate mixing and drying time, and online detection can be completed in less than 1 min. The method’s broad detection range, low detection limit, and rapid analysis time address a key gap in AFB1 detection in tea by offering a means of rapid detection. This approach holds substantial potential for application in tea processing. 3. Methods 3.1. Materials and Reagents The dark tea used in this study was Pu-erh ripe tea (Menghai Tea Factory, Dayi Pu-erh Ripe Tea, 2008), Qianliang tea (Baishaxi Tea Factory Limited Liability Company, Hunan Province, 2007), porcupine brick tea (Anhua, Yiyang, Hunan Province, China Tea Hunan Anhua Tea Factory Limited Company, 2014), Kang brick tea (Ya’an, Sichuan Province, 2012), and Liu Baotong tea (Wuzhou, Guangxi Province Maosheng Tea Co. Ltd., 2006). AFB1 (99.5%) was purchased from the Yunmo Quality Inspection Standard Material Centre. NH 2 -PEG3400-NH 2 (97%) and R6G (95%) were purchased from Shanghai Yi En Chemical Technology Co. Ethanol was purchased from Shanghai Zhan Yun Chemical Co. The following chemical reagents, all of which were of analytical grade, were obtained from Sinopharm Chemical Reagent Co. Ltd.: CTAB, HAuCl 4 -4H 2 O, NaBH 4 , NaOL, AgNO 3 , ascorbic acid, HCl, H 2 SO 4 , NaOH. 3.2. Preparation and Characterisation of NH 2 -PEG-AuNRs The method used to synthesise NH 2 -PEG-AuNRs was adapted from a previously reported method, with some modifications 26 . First, the synthesised gold nanorods were dispersed in an amino-polyethylene glycol-amino (NH 2 -PEG-NH 2 ) solution and allowed to react for 12 h to achieve surface modification. Following functionalisation, the solution was centrifuged at 6000 rpm for 20 min, and the supernatant was discarded. The remaining lower layer was redispersed in ultrapure water and centrifuged again under identical conditions. This purification process was repeated three times to ensure the removal of excess reagents and byproducts. The final NH 2 -PEG-AuNRs were stored at 4°C in the dark. The ultraviolet-visible (UV-vis) spectra of the NH 2 -PEG-AuNRs were measured using an enzyme labeller (SpectraMax M2, Molecular Devices, USA) in the detection range of 300–1000 nm, with a step size of 1 nm. The morphological structure of the NH₂-PEG-AuNRs was examined using transmission electron microscopy (TEM) at an accelerating voltage of 180.0 KV (HT7700, Hitachi, Japan). The elemental composition of the synthesised nanorods was analysed using an ultra-high resolution field emission scanning electron microscope (SU8600, Hitachi) equipped with an energy-dispersive spectrometer. The zeta potential of the NH 2 -PEG-AuNRs was measured using a nanoparticle sizer (Zetasizer Lab, Malvern Panalytical, UK). Each measurement was performed in triplicate, and the instrument automatically determined average values. To evaluate the applicability and accuracy of the newly developed method for the rapid quantitative detection of AFB1 in Pu-erh tea extracts, four representative dark teas were selected for comparison: Qianliang tea, Kangzhi tea, Poria tea, and Liubao tea. Each dark tea extract was spiked with AFB1 at three different concentrations: 1 × 10 − 5 , 1 × 10 − 6 , and 1 × 10 − 7 ng/mL. The spiked tea extracts were then mixed at a 1:1 ratio with the NH2-PEG-AuNR substrate for SERS detection. 3.3. Validation of SERS Performance of NH 2 -PEG-AuNRs The sensitivity, reproducibility, stability, and enhancement factor (EF) of the NH 2 -PEG-AuNRs as a SERS substrate were evaluated using Raman spectrometry (Lab RAM HR Evolution, Horiba Jobin Yvon, France). R6G was used as a reference analyte, with a 785-nm laser used in the detection range of 400–1800 cm − 1 . The detection time was 8 s, the accumulation number was 1, and the power was 25%. To determine the sensitivity of the NH₂-PEG-AuNRs, a series of R6G standard solutions with concentration gradients of 10 − 3 , 10 − 4 , 10 − 5 , 10 − 6 , 10 − 7 , 10 − 8 , 10 − 9 , 10 − 10 , and 10 − 11 mol/L were prepared. Each solution was mixed in a 1:1 ratio with NH 2 -PEG-AuNRs, and 15 spectra were recorded per concentration. The correlation coefficient between the Raman signal intensity and R6G concentration was analysed to confirm the sensitivity of the NH 2 -PEG-AuNRs. For reproducibility testing, 10 − 7 mol/L R6G was mixed with NH 2 -PEG-AuNRs in a 1:1 ratio, and 20 randomly selected measurement points were analysed. The reproducibility of the NH 2 -PEG-AuNRs was confirmed by calculating the RSD values of the signal intensities across these measurement points. To assess the stability of the NH 2 -PEG-AuNRs, a Raman assay was performed every 15 days over a 4-month period by mixing 10 − 7 mol/L R6G with NH 2 -PEG-AuNRs in a 1:1 ratio, and 15 valid spectra were recorded at each time point. The RSD values of the signal intensities across the time points were calculated. The EF of the NH 2 -PEG-AuNRs as a SERS substrate was calculated using the following Eq. 2 7 : $$\:\text{E}\text{F}=\left({\text{I}}_{\text{S}\text{E}\text{R}\text{S}}/{\text{C}}_{\text{S}\text{E}\text{R}\text{S}}\right)/\left({\text{I}}_{\text{R}\text{S}}/{\text{C}}_{\text{R}\text{S}}\right)$$ 1 where I SERS and C SERS denote the Raman signal intensity and concentration of R6G bound to the SERS substrate, respectively, and I RS and C RS denote the signal intensity of the Raman spectra and concentration of R6G in its free form, respectively. 3.4 SERS Acquisition of AFB1 3.4.1 Optimisation of Detection Conditions To prepare an AFB1 standard solution, AFB1 powder was completely dissolved in ethanol; this was followed by serial 10-fold dilutions from 10 − 1 to 10 − 17 ng/mL. A total of 17 samples were prepared and stored in sealed containers at − 20°C. To determine the optimal mixing ratio between AFB1 and NH 2 -PEG-AuNRs, a 10 − 6 ng/mL AFB1 standard solution was mixed with NH 2 -PEG-AuNRs at different ratios (3:1, 2:1, 1:1, 1:2, and 1:3) in centrifuge tubes. The mixtures were then ultrasonicated for 10 min, and 5 µL of the mixture was aspirated and dripped onto gold-plated slides, which were dried in an oven at 55°C. After, Raman spectra were collected using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France) by using a 785-nm laser with an acquisition range of 400–1800 cm − 1 , an acquisition time of 1 s, and an accumulation number of 1. A total of 15 spectra were collected for each set of experiments and averaged. The optimal mixing ratio was determined by comparing the signal intensity across the average spectra. 3.4.2 SERS Detection of AFB1 in Standard Solutions On the basis of the optimal mixing ratio, AFB1 standard solutions of varying concentrations were thoroughly mixed with NH2-PEG-AuNRs in centrifuge tubes. The mixtures were ultrasonicated for 10 min, and 5 µL of the mixture was aspirated and dropped on a gold-plated slide, which was dried in an oven at 55°C. Raman spectra were acquired using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France). 3.4.3 SERS Detection of AFB1 in Real Samples To prepare tea extracts spiked with AFB1, 5 g of each type of tea (i.e., Pu-erh tea, Liu Bao tea, Qian Liang tea, Kang brick tea, and Por brick tea) was weighed using an electronic balance and transferred to a 500-mL conical flask. Subsequently, 250 mL of boiling water (95–100°C) was added, and the mixture was steeped for 5 min. The resulting tea infusions were filtered, collected into clean conical flasks, and rapidly cooled to room temperature. AFB1 standard solutions were added to the Pu-erh tea extract to achieve a series of concentrations ranging from 10 − 1 to 10 − 9 ng/mL in 10-fold decrements, with a total of nine samples obtained with different concentrations. The AFB1 standard solution was similarly added to the Liubao tea, Qianliang tea, Kang brick tea, and Por brick tea extracts to obtain final concentrations of 10 − 5 , 10 − 6 , and 10 − 7 ng/mL, respectively. On the basis of the optimal mixing ratio, AFB1 standard solutions of varying concentrations were thoroughly mixed with NH 2 -PEG-AuNR substrates in centrifuge tubes. The mixtures were ultrasonicated for 10 min, and 5 µL of the mixture was aspirated and dripped onto gold-plated slides, which were dried in an oven at 55°C. Raman spectra were acquired using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France) by using a 785-nm laser at a detection time of 8 s, a cumulative number of 2, and a detection range of 400–1800 cm − 1 . In total, 15 spectra were collected for each sample concentration. 3.5 Data Processing and Analysis 3.5.1 Establishment of the AFB1 Standard Curve To quantify AFB1, a correlation curve was constructed using the Raman signal intensity at 1267 cm − 1 as the y-axis and the AFB1 concentration as the x-axis. For the AFB1 standard solutions, a total of 17 samples with varying concentrations were analysed, with 15 valid spectral data points per sample, for a total of 255 spectral data points. For the Pu-erh tea extracts spiked with AFB1, a total of 9 samples with different concentrations were tested, and 15 valid spectral data points were obtained for each sample, for a total of 135 data points. The standard curves of AFB1 in the standard solution and Pu-erh tea extract were established using this method. 3.5.2 Development and External Validation of the AFB1 Quantitative Prediction Model Partial least-squares (PLS) regression models were developed to quantitatively predict AFB1 concentrations in both standard solutions and spiked Pu-erh tea extracts by using SERS with NH 2 -PEG-AuNRs as the substrate. To construct these models, the spectral data were first divided into calibration and prediction sets in a 2:1 ratio, with Raman spectral data serving as input variables and AFB1 concentrations as output variables. For external validation, the PLS prediction model was applied to four types of dark tea extracts, namely Qianliang tea, Kang brick tea, Por brick tea, and Liu Bao tea. Each extract was spiked with 3 concentrations of AFB1, and 15 valid Raman spectra were collected per concentration. This resulted in a dataset of 45 spectral measurements per tea extract. The developed PLS model was then applied to these datasets to assess its predictive accuracy. PLS regression is a widely used technique in quantitative predictive modelling. It establishes a quantitative forecasting model by predicting one or more dependent variables (output variables) from known independent variables (input variables). External validation of a PLS quantitative forecasting model involves applying an established PLS model to a new dataset that has not been used during model development. This dataset is referred to as a test or validation set 28 . The metrics used to evaluate the performance of this quantitative prediction model are the coefficient of determination (Rc) for the training set and its root mean square error of calibration, the coefficient of determination (Rp) for the prediction set and its root mean square error of prediction (RMSEP), and the residual prediction deviation (RPD). RPD is defined as the ratio of the standard deviation of the samples in the prediction set to the RMSEP. In general, RPD < 1.4 indicates that the developed model has poor prediction performance and the model is unreliable; 1.4 < RPD < 1.8 indicates that the developed model can be used for rough prediction; 1.8 < RPD 2 indicates that the model has excellent quantitative prediction ability. In this study, the model evaluation indices were calculated as follows 29 : $$\:\text{R}=\sqrt{1-\frac{\sum\:_{\text{i}=1}^{\text{n}}{({\text{y}}_{\text{i}}-{\widehat{\text{y}}}_{\text{i}})}^{2}}{\sum\:_{\text{i}=1}^{\text{n}}{({\text{y}}_{\text{i}}-{\stackrel{-}{\text{y}}}_{\text{m}})}^{2}}}$$ 2 $$\:\text{R}\text{M}\text{S}\text{E}=\sqrt{\frac{\sum\:_{\text{i}=1}^{\text{n}}{({\text{y}}_{\text{i}}-{\widehat{\text{y}}}_{\text{i}})}^{2}}{\text{n}}}$$ 3 $$\:\text{R}\text{P}\text{D}=\frac{\text{S}\text{D}}{\text{R}\text{M}\text{S}\text{E}\text{P}}$$ 4 where \(\:{y}_{i}\) and \(\:{\widehat{y}}_{i}\) are the true value and the model predicted value of the first sample in the training or prediction set, respectively; \(\:{\stackrel{-}{y}}_{m}\) denotes the average of the true values of the samples in the training or prediction set; and n denotes the number of samples in the training or prediction set. 3.6 Software All spectral data in this study were baseline corrected and smoothed for noise reduction by using Labsix Sigma software (Horiba Scientific, NJ, USA) with uniform parameters. All maps were plotted in Origin 2023b (OriginLab Corp, MA, USA), and the PLS model was developed and implemented in MATLAB 2014a (Mathworks, Natick, USA). Declarations Competing interests All authors declare no financial or non-financial competing interests. Author Contribution S.C., Y.Z. designed experiments and data analyses; S.C., Y.Z., Q.Y., andY.W. performed the experimental work; Q.L. provided technical support; S.C.,Y.Z. analyzed data and wrote the manuscript; J.N., Q.L. supervised theproject. All authors reviewed the manuscript. Acknowledgements This study was financially supported by National Nature Science Foundation of China (32202543), National Key Research and Development Program (2021YFD1601102), National Key Research and Development Program (2023YFD1601300), Key Research and Development Program of Anhui Province (202104h04020023), the earmarked fund for CARS (CARS-19). Data availability The data that support the findings of this study are available on request from the corresponding author upon reasonable request. References Ding, Q. et al. Comparison of hypoglycemic effects of ripened pu-erh tea and raw pu-erh tea in streptozotocin-induced diabetic rats. Rsc Advances, 9(6), 2967-2977 (2019). Zhang, W. et al. Multiplex SERS-based lateral flow immunosensor for the detection of major mycotoxins in maize utilizing dual raman labels and triple test lines. Journal of Hazardous Materials, 393, 122-348 (2020). Zhang, X. et al. Fungal flora and mycotoxin contamination in tea: Current status, detection methods and dietary risk assessment-A comprehensive review. Trends in Food Science & Technology, 127, 207-220 (2022). Fan, L. et al. Dummy molecularly imprinted solid-phase extraction-SERS determination of AFB1 in peanut. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 288, 122-130 (2023). Ostry, V., Malir, F., Toman, J., & Grosse, Y.. Mycotoxins as human carcinogens-the IARC monographs classification. Mycotoxin Research, 33(1), 65-73(2017). Hassan, F. et al. Aflatoxin B1 in Rice: Effects of storage duration, grain type and size, production site, and season. Journal of Food Protection, 85(6), 938-944 (2022). Gholizadeh, S. et al. Ultrasound-assisted solvent extraction combined with magnetic ionic liquid based-dispersive liquid-liquid microextraction for the extraction of mycotoxins from tea samples. Journal of Food Composition and Analysis, 114, 104-831 (2022). Gazioglu, I., & Kolak, U. Ochratoxin a levels in food and beverage samples from Turkey. Acta Alimentaria, 47(2), 189-194. https://doi.org/10.1556/066.2018.47.2.7 Cina, M.et al. Martinez, L. D., & Cerutti, S. Development of a novel UHPLC-MS/MS method for the determination of ochratoxin A in tea. Heliyon, 7(4), e06-663 (2021). Wang, Q. et al. Li, J., Deng, Q., Zhang, X., Wang, S., & Chen, M. High-performance electrochemiluminescence sensors based on ultra-stable perovskite quantum dots@ZIF-8 composites for aflatoxin B1 monitoring in corn samples. Food Chemistry, 410, 135-325 (2023). Zhang, B. et al. Simple "signal-on" photoelectrochemical aptasensor for ultrasensitive detecting AFB1 based on electrochemically reduced graphene oxide/poly(5-formylindole)/Au nanocomposites. Biosensors & Bioelectronics, 134, 42-48(2019). Chen, Z. et al. Discrimination of toxigenic and non-toxigenic Aspergillus flavus in wheat based on nanocomposite colorimetric sensor array. Food Chemistry, 430, 137-048(2024). Lu, L., Yu, R., & Zhang, L. AFB1 colorimetric aptamer sensor for the detection of AFB1 in ten different kinds of miscellaneous beans based on gold nanoparticles and smartphone imaging. Food Chemistry, 421, 136-205 (2023). Dou, X. et al. Construction of a nanoscale metal-organic framework aptasensor for fluorescence ratiometric sensing of AFB1 in real samples. Food Chemistry, 416, 135-805(2023). Lu, X. et al. Target-driven switch-on fluorescence aptasensor for trace aflatoxin B1 determination based on highly fluorescent ternary CdZnTe quantum dots. Analytica Chimica Acta, 1047, 163-171 (2019). Ding, S. et al. Electromagnetic theories of surface-enhanced raman spectroscopy. Chemical Society Reviews, 46(13), 4042-4076(2017). Wu, Z.et al.. Ti3C2Tx MXenes loaded with Au nanoparticle dimers as a surface-enhanced raman scattering aptasensor for AFB1 detection. Food Chemistry, 372, 131-293(2022). He, H.et al. Bridging Fe3O4@Au nanoflowers and Au@Ag nanospheres with aptamer for ultrasensitive SERS detection of aflatoxin B1. Food Chemistry, 324, 126-832(2020). Li, Y. et al. Microarray surface enhanced raman scattering based immunosensor for multiplexing detection of mycotoxin in foodstuff. Sensors and Actuators B-Chemical, 266, 115-123(2018). Duy Vu, T. et al. A facile paper-based chromatography coupled Au nanodendrite on nickel foam for application in separation and SERS measurement. Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy, 313, 124-137(2024). Medina, Det al. Current role of modern chromatography and mass spectrometry in the analysis of mycotoxins in food. Trac-Trends in Analytical Chemistry, 135, 116-156(2021). Zhou, H. et al. Mycotoxins in tea ((Camellia sinensis (L.) Kuntze)): Contamination and dietary exposure profiling in the Chinese population. Toxins, 14(7), 452 (2022). Chen, W. et al. Optimum synthesis of Au@Ag nanoparticle as plasma amplifier to detect trace concentration of AFB1 via object-binder-metal SERS method. Journal of Food and Drug Analysis, 30(4), 603-613(2022). Chen, Q. et al. A large raman scattering cross-section molecular embedded SERS aptasensor for ultrasensitive Aflatoxin B1 detection using CS-Fe3O4 for signal enrichment. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 189, 147-153(2018). Jiao, T. et al.. Aggregation triggered aflatoxin B1 determination in foodstuff employing 5-aminotetramethylrhodamine decorated gold-silver core-shell nanoparticles in surface enhanced raman scattering. Sensors and Actuators B-Chemical, 331, 129-424(2021). Deng, Y. et al. Trimetallic Au@PtPd mesoporous nanorods as efficient electrocatalysts for the oxygen reduction reaction. Acs Applied Energy Materials, 1(9), 4891-4898(2018). Li, L. et al. Rapid detection of multiple colorant adulteration in Keemun black tea based on hemp spherical AgNPs-SERS. Food Chemistry, 398, 133-841(2023). Wang, Y. et al. Evaluating taste-related attributes of black tea by micro-NIRS. Journal of Food Engineering, 290, 110-181(2021). Viegas, T. et al. Determination of quality attributes in wax jambu fruit using NIRS and PLS. Food Chemistry, 190, 1-4(2016). Scheme Scheme 1 is available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterials.docx floatimage1.png scheme 1 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7144899","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":492088609,"identity":"99a2f393-e4d6-4a25-b667-55843a2cfdc3","order_by":0,"name":"Shuci Cao","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Shuci","middleName":"","lastName":"Cao","suffix":""},{"id":492088610,"identity":"93070a63-fe21-4ff4-8580-b31595fa7c7e","order_by":1,"name":"Yifan Zuo","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Yifan","middleName":"","lastName":"Zuo","suffix":""},{"id":492088611,"identity":"d2d8a98e-6ebd-4cea-8464-6b73bb4546ee","order_by":2,"name":"Qianfeng Yang","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Qianfeng","middleName":"","lastName":"Yang","suffix":""},{"id":492088612,"identity":"d117f39e-6245-488c-a07e-a9ecf5583ddd","order_by":3,"name":"Yongning Wei","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Yongning","middleName":"","lastName":"Wei","suffix":""},{"id":492088613,"identity":"13d0a2af-8836-4a93-b743-3325d510fd5d","order_by":4,"name":"Qisheng Liu","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Qisheng","middleName":"","lastName":"Liu","suffix":""},{"id":492088614,"identity":"4e26cd92-f183-4f98-8210-f759dc4cead1","order_by":5,"name":"Jingming Ning","email":"","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Jingming","middleName":"","lastName":"Ning","suffix":""},{"id":492088615,"identity":"ad463488-acee-40ad-92f4-5d41f5f74572","order_by":6,"name":"Luqing Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuUlEQVRIiWNgGAWjYBACPmbmBoYPDAfAHAmitLAxMzYwziBNCwNjAzMPaVrYGRtv29TckTc4wHzwNg+DnZxuA2GHNVvnHHtmuOEAW7I1D0OysdkBwlrapHPYDjNuOMBjJg10YeI2orRY/Dtsv+EA/zcStDC2HU4E2sJGtJZmy96+w8kzD7MZW84xIMIv/PyHD9748e2wbd/x5oc33lTYyRHUAgKQ6GAGEQZEKEdoGQWjYBSMglGACwAAOIM7gFZuTNsAAAAASUVORK5CYII=","orcid":"","institution":"Anhui Agricultural University","correspondingAuthor":true,"prefix":"","firstName":"Luqing","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-07-17 04:53:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7144899/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7144899/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87802287,"identity":"1c519488-6ad6-4d4c-909c-186bf7ad0b9e","added_by":"auto","created_at":"2025-07-29 08:01:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":463647,"visible":true,"origin":"","legend":"\u003cp\u003eCharacterisation of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. (A) TEM images of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. (B) SEM images of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. (C - E) EDS elemental analysis diagram of Au, C and O elements in NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs nanostructures. (F) UV absorption spectra of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. (G) Zeta potential maps of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/e67d43768a7627958755e24e.png"},{"id":87803649,"identity":"c4706f64-b330-437b-a361-ec68c9a5b9ef","added_by":"auto","created_at":"2025-07-29 08:09:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":624558,"visible":true,"origin":"","legend":"\u003cp\u003eCharacterisation plot of the properties of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. (A B) SERS spectra of different concentrations of R6G based on NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs substrates and corresponding fitted curves. (C) Raman signal intensity mapping of 121 different sites equally spaced within a 100 μm × 100 μm square matrix at 1361 cm\u003csup\u003e-1\u003c/sup\u003e. (D) SERS spectrograms of 20 randomly collected points. (E) SERS intensity of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs -based R6G at 1361 cm\u003csup\u003e-1\u003c/sup\u003e and 1506 cm\u003csup\u003e-1\u003c/sup\u003e collected every 15 days. (F) Comparison of raman and surface-enhanced Raman spectra of R6G.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/fc8c5e08c48150ce54751f7b.png"},{"id":87802291,"identity":"a2e4ebc8-6de6-4d9c-bd42-f0048356ef54","added_by":"auto","created_at":"2025-07-29 08:01:47","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":194276,"visible":true,"origin":"","legend":"\u003cp\u003eOptimisation of the mixing ratio of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs with samples. (A) Comparison of raman and surface-enhanced Raman spectra of AFB1. (B) Surface-enhanced Raman spectra of different substances based on NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs substrates. (C) Surface-enhanced Raman spectra of AFB1 in different dark tea infusion. (D) Surface-enhanced Raman spectra of AFB1 and NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs with different mixing ratios.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/cfd5997c8190fcad259df4a5.png"},{"id":87802288,"identity":"4f4c8579-09a0-4a93-8d09-84a8472130d5","added_by":"auto","created_at":"2025-07-29 08:01:47","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":374672,"visible":true,"origin":"","legend":"\u003cp\u003eDetection of AFB1 standards. (A) Surface-enhanced Raman spectra of different concentrations of AFB1 standard solutions based on NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs substrates (B) Correlation curve between the concentration of AFB1 standard solution and the intensity of surface-enhanced raman signal. Determination of AFB1 in Pu-erh tea infusion. (C) Surface-enhanced Raman spectra of different concentrations of AFB1 solutions in Pu-erh tea based on NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs substrates (D) Correlation curves between the concentration of AFB1 and the intensity of surface-enhanced raman signals in Pu-erh tea infusion.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/ad24961e552ea7657bfd6e45.png"},{"id":93744573,"identity":"1e72e860-0e5b-4add-9b10-78a13aaa4a9a","added_by":"auto","created_at":"2025-10-17 06:17:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2594448,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/1c834cba-76be-46c3-b14b-a879c3446d28.pdf"},{"id":87802290,"identity":"c49d944e-7459-4c72-9a73-0a47fbce2fa3","added_by":"auto","created_at":"2025-07-29 08:01:47","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":20469,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/e989c94e91eb5cf0ca3e1208.docx"},{"id":87804528,"identity":"85202678-dc09-4975-aae3-e7bfe27052c2","added_by":"auto","created_at":"2025-07-29 08:17:47","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":382025,"visible":true,"origin":"","legend":"\u003cp\u003escheme 1\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7144899/v1/b07a16fee2d416be0c03f21f.png"}],"financialInterests":"No competing interests reported.","formattedTitle":"NH2-PEG-AuNR-Based Surface-Enhanced Raman Spectroscopy for Rapid Detection of AFB1: Dark Tea Safety Assessment","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eDark tea is widely consumed globally because of its distinctive flavour and numerous health-promoting components \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. The pile fermentation process of dark tea is strongly influenced by microorganisms, which play a crucial role in determining the quality and flavour profile of dark tea. \u003cem\u003eAspergillus\u003c/em\u003e spp. and \u003cem\u003ePenicillium\u003c/em\u003e spp are the main microorganisms involved in this fermentation process\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Microbial toxin contamination can severely affect the quality and safety of dark tea. The risk of such contamination arises from mycotoxins produced during pile fermentation, spoilage-like substances produced during mould decomposition, and microbial toxins generated through interactions between environmental microorganisms and those present in the tea that occur when tea is improperly stored\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eAflatoxin B1 (AFB1) accounts for 75% of the mycotoxin contamination that occurs in food and feed\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Aflatoxins, which are primarily produced by \u003cem\u003eAspergillus\u003c/em\u003e fungi, particularly \u003cem\u003eAspergillus flavus\u003c/em\u003e and \u003cem\u003eAspergillus parasiticus\u003c/em\u003e, can be classified into four major subtypes: AFB1, AFB2, AFG1, and AFG2. Among these, AFB1 is the most common aflatoxin monomer and has the highest toxicity\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. AFB1 is physicochemically stable, has poor water solubility, and is readily soluble in organic solvents such as methanol, ethanol, and acetonitrile. Regarding its structure, it contains a double-furan ring with a double bond that undergoes epoxidation, which facilitates its interaction with nucleic acids, proteins, and other biological macromolecules, leading to toxic effects. Additionally, AFB1 contains an oxonaphthalene\u0026ndash;octanone (coumarin) ring that contributes to AFB1\u0026rsquo;s carcinogenicity. Because of its potent toxicity and carcinogenicity, AFB1 has been classified as a class I carcinogen by the International Agency for Research on Cancer\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Therefore, a sensitive and accurate method for the rapid detection of AFB1 in dark tea must be developed to ensure product safety and support the sustainable development of the tea industry.\u003c/p\u003e\u003cp\u003eNumerous analytical techniques have been developed for detecting AFB1, including high performance liquid chromatography (HPLC)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, enzyme-linked immunosorbent assay\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e and liquid chromatography-mass spectrometry (LC-MS)\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Although these assays provide highly accurate quantitative results, they are often complex, time-consuming, and costly, and therefore, they cannot be used for rapid, on-site detection. In consideration of these limitations, a growing number of studies has focused on developing nanomaterial-based sensors for AFB1 detection. For instance, Wang developed a high performance electrochemiluminescence sensor based on ultra-stable perovskite quantum dots@ZIF-8 composites for the detection of AFB1 in maize\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Additionally, Zhang developed a composite gold nanophotoelectrochemical aptamer\u0026ndash;based sensor for AFB1 detection and successfully applied it in peanuts and wheat\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Although electrochemical techniques offer advantages such as high sensitivity and rapid response times, electrochemical sensors are highly susceptible to interference from complex sample matrices, which can lead to cross-reactivity and reduced detection accuracy. Chen identified AFB1 in wheat by using a composite nanocolourimetric sensor array\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, whereas Lu developed a colorimetric sensor based on gold nanoparticles (AuNPs) and a smartphone for detection of AFB1 in beans\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. This nanosensor-based colorimetric method relied on evaluation of colour changes before and after a reaction. Although they achieved visualisation, the colorimetric results were highly influenced by ambient lighting conditions and instrument variability, resulting in low sensitivity. Dou developed a fluorescence method by using a ratiometric NMOFs-Aptasensor based on fluorescence resonance energy transfer (FRET) for detecting AFB1 in maize\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Similarly, Lu developed a FRET-based fluorescence method by using CdZnTe quantum dots and AuNPs for detecting AFB1 and successfully achieved such detection in peanuts\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Fluorescence methods enable multichannel detection of AFB1 through fluorescence signals at different wavelengths, which can enhance accuracy and reliability. However, these methods typically require complex sample pretreatment, which limits their practical application. In consideration of these constraints, a nanosensor must be developed for trace detection of AFB1 that is both simple to operate and resistant to interference from sample matrices.\u003c/p\u003e\u003cp\u003eSurface-enhanced Raman scattering (SERS) is an advanced spectroscopic technique that provides a unique molecular \u0026lsquo;fingerprint\u0026rsquo; for an analyte. SERS offers the benefits of accurate trace detection, high resolution, and minimal sample consumption. Among the various SERS substrates that have been proposed, nanomaterials composed of gold and silver are the most efficient hotspot sensors and are widely employed for the trace detection of AFB1 in food\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. AFB1 detection using precious metal nanomaterials is generally conducted using one of two methods. The first involves the direct detection of AFB1 by using nanomaterials; however, this approach is often associated with signal instability and poor reproducibility. The second method involves modifying nanomaterials to facilitate the detection of target molecules through physical or chemical adsorption on the substrate surface. This approach has gradually replaced direct detection. Notably, aptamer-modified nanomaterial sensors are highly sensitive receptors for specific target detection. Wu detected AFB1 in peanuts by using SH-aptamer-modified AuNP dimers as SERS aptamer sensors\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Additionally, He employed SERS aptamers comprising SH-cDNA-modified Fe3O4@Au NFs and SH-Apt-modified Au-4MBA@Ag NSs to detect AFB1 in peanut oil\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. In addition to aptamer-based approaches, immunosensors have been widely employed in SERS-based detection; these sensors use antibody\u0026ndash;antigen interactions to identify targets. Li developed a DSNB-modified AuNP SERS immunosensor for the detection of mycotoxins in maize, rice, and wheat\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, and Zhang employed DTNB-labelled SRES immunosensors for the detection of mycotoxins in maize, rice, and wheat\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Although these sensors have considerably improved detection accuracy, antibody\u0026ndash;antigen- and aptamer-based methods are complex. Furthermore, research focused on the trace detection of AFB1 in tea matrices is scarce. Such detection is difficult because of the composition of tea extracts, which form a complex background for detection. Therefore, a simple and easy-to-use method must be developed that enables rapid and accurate quantitative analysis of AFB1 in tea matrices.\u003c/p\u003e\u003cp\u003eIn consideration of this, the present study developed a simple and rapid method for trace detection of AFB1 in dark tea extracts by using SERS. This approach involved the preparation of highly sensitive, reproducible, and stable SERS substrates, namely, amino-terminally modified polyethylene glycol gold nanorods (NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs). This method enables detection through the use of the hydrogen bonding interactions between NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs and AFB1 (Scheme 1). By integrating Raman spectroscopy with emerging nanotechnology, this study established a simple, sensitive, and accurate method for detecting AFB1 to ensure tea safety. The study findings can be applied to enhance food safety measures and advance the tea industry.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"2. Results and Discussion","content":"\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.1 Characterisation of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe morphology and homogeneity of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs were analysed through TEM and scanning electron microscopy. As presented in Figure 1AB,\u0026nbsp;the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs exhibited a uniform size distribution, with an average size of approximately 20 nm \u0026times; 80 nm. The elemental composition of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was determined through energy-dispersive X-ray spectroscopy, as indicated in\u0026nbsp;Figure 1C. The analysis confirmed that the samples were mainly composed of Au (red), C (green), and O (blue). A UV-vis absorption spectrum of the NH₂-PEG-AuNRs (Figure 1D) revealed transverse plasmon resonance absorption peaks corresponding to the transverse surface plasmon resonance (TSPR) and the longitudinal plasmon resonance (LSPR). The TSPR peak was observed at approximately 508 nm, whereas the LSPR peak occurred at approximately 808 nm. Additionally, the surface chargeability of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was characterised using zeta potential maps. As presented in Figure 1E, the zeta potential of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was 34.97 mV.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.2 SERS Activity of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SERS sensitivity of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNR substrates was evaluated using R6G at various concentrations (10\u003csup\u003e\u0026minus;3\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;4\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;5\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;6\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;7\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;8\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;9\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;10\u003c/sup\u003e, and 10\u003csup\u003e\u0026minus;11\u003c/sup\u003e mol/L). The corresponding SERS spectra for different R6G concentrations are presented in Figure 2A, which reveals characteristic peaks of R6G at 612, 772, 1184, 1311, 1361, 1506, and 1647 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e. The peaks at 612, 1311, 1361, 1506, and 1647 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e were attributable to C\u0026ndash;C ring stretching vibrations, whereas those at 772 and 1184 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e were associated with C\u0026ndash;H bending vibrations\u003csup\u003e20\u003c/sup\u003e. When the concentration of R6G was greater than 10\u003csup\u003e\u0026minus;11\u003c/sup\u003e mol/L, the characteristic peaks of R6G molecules at 1361 and 1506 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e were still clearly visible. As indicated in Figure 2B, a linear relationship was noted between the R6G characteristic peak intensity at 1506 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e and the concentration of R6G, yielding the equation y = \u0026minus;310.20 x + 3501.76, with a high correlation coefficient of R\u003csup\u003e2\u003c/sup\u003e = 0.9960. This strong linear correlation confirms the high sensitivity of the NH2-PEG-AuNRs prepared in this study.\u003c/p\u003e\n\u003cp\u003eThe reproducibility of the NH2-PEG-AuNR substrates was investigated using R6G (10\u003csup\u003e\u0026minus;7\u003c/sup\u003e mol/L) as a model analyte. As presented in Figure 2C, the Raman signal intensity mapping plot of 121 evenly distributed sites within a 100 \u0026mu;m \u0026times; 100 \u0026mu;m square matrix at 1361 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e revealed a uniform colour distribution. Figure 2D presents the SERS spectra of 20 randomly selected sites on the substrate. The RSD of the Raman signal intensity at 1506 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e was calculated as 4.08%. These findings confirm that the NH2-PEG-AuNR substrates exhibit high reproducibility.\u003c/p\u003e\n\u003cp\u003eThe stability of the NH2-PEG-AuNR substrate was evaluated by monitoring the SERS signals of R6G (10\u003csup\u003e\u0026minus;7\u003c/sup\u003e mol/L) every 15 days for 4 months. As presented in Figure 2E, the characteristic peaks of the SERS spectra remained consistent across all time points, indicating that the NH2-PEG-AuNRs are structurally stable and meet the requirements for practical applications. Variations in signal intensity with time over 120 days were recorded, and the RSD was calculated to be 9.75%. This relatively low variation indicates that the NH2-PEG-AuNR substrate prepared in this study has excellent stability.\u003c/p\u003e\n\u003cp\u003eThe SERS and Raman spectra of R6G, presented in\u0026nbsp;Figure 2F, illustrate the substantial signal enhancement effect of the NH2-PEG-AuNR substrates on R6G. The EF of the NH2-PEG-AuNRs was calculated using eq. (1) as 4.51 \u0026times; 10\u003csup\u003e7\u003c/sup\u003e. This result confirms the excellent performance of the NH₂-PEG-AuNRs in SERS detection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.3 Optimisation of the NH2-PEG-AuNR Substrate Mixing Ratio for AFB1 Detection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the potential for enhanced Raman signal detection of AFB1 by using NH2-PEG-AuNR substrates, the Raman spectra of AFB1 solid powder and SERS spectra of AFB1 ethanol solution were obtained (Figure 3A). A comparison of these spectra revealed distinct AFB1 characteristic peaks located at 524, 775, 1075, 1144, 1267, 1496, and 1550 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e, with particularly strong response signals at 1144, 1267, and 1496 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e. To investigate the feasibility of detecting AFB1 in tea matrices, the SERS spectra of AFB1 in ethanol solution, the SERS spectra of AFB1 in tea extracts (at equivalent concentrations), the SERS spectra of tea extracts, and the SERS spectra of the NH2-PEG-AuNR substrate itself were compared, and the results are presented in Figure 3B. The SERS signal intensity of AFB1 in the tea extract was significantly weaker than that of AFB1 in ethanol solution. However, the tea extract itself exhibited no distinct characteristic peaks, and the NH2-PEG-AuNR substrate itself had no effect on the detection results. This suggests that the tea extract matrix may influence the quantity of AFB1 captured by the NH2-PEG-AuNR substrate, although it does not appear to alter the SERS characteristic peaks of AFB1. Further investigation into the influence of different dark tea extracts spiked with AFB1 was conducted, with the corresponding SERS spectra displayed in Figure 3C. Notably, the AFB1 characteristic peak at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e remained clearly visible across all samples. These findings indicate that variations in tea extract matrices did not exert any obvious effect on AFB1 detection.\u003c/p\u003e\n\u003cp\u003eTo assess the effect of different mixing ratios of AFB1 and NH2-PEG-AuNR substrates on Raman signal intensity, a systematic investigation was conducted, as illustrated in Figure 3D. The results indicate that the Raman signal intensity of AFB1 was the highest at a volume ratio of 1:1. However, when the volume ratio either increased or decreased, the signal intensity declined. This may be attributable to the optimal formation of hotspots at the suitable volume ratio, where the interaction between the substrate and the target molecule is most effective. Therefore, a mixing ratio of 1:1 was selected as the optimal condition for subsequent detection experiments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.4 Establishment of AFB1 Standard Curve\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.4.1 Establishment of the AFB1 Standard Curve in Ethanol Solutions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SERS spectra of AFB1 ethanol solution\u0026nbsp;were recorded across\u0026nbsp;17 concentrations ranging from 10\u003csup\u003e\u0026minus;1\u003c/sup\u003e to 10\u003csup\u003e\u0026minus;17\u003c/sup\u003e ng/mL, with the lowest instrumental limit of detection (LOD) being 10\u003csup\u003e\u0026minus;17\u003c/sup\u003e ng/mL, as indicated in Figure 4A. Repeated experiments confirmed that the intensity of the spectral peak at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e was stable, and a linear relationship was observed in the specific concentration range.\u003c/p\u003e\n\u003cp\u003eAs presented in Figure 4B, the Raman signal intensity at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e gradually weakened as the concentration of AFB1 was reduced. A linear correlation was observed between the peak intensity at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e and the AFB1 concentration in the range of 1 \u0026times; 10\u003csup\u003e\u0026minus;1\u003c/sup\u003e to 1 \u0026times; 10\u003csup\u003e\u0026minus;17\u003c/sup\u003e ng/mL, following the equation: y = \u0026minus;57.69 x + 1009.87, R\u003csup\u003e2\u003c/sup\u003e = 0.9971.\u003c/p\u003e\n\u003cp\u003eIn the field of rapid detection of toxins, studies regarding SERS-based toxin detection have generally struggled to achieve direct detection using nanomaterials\u003csup\u003e18\u003c/sup\u003e. In the current study, NH2-PEG-AuNRs were employed to bind AFB1. This method offers advantages such as a broad detection range and low detection limit for AFB1 quantification. The experimental results of this study confirmed the high sensitivity of the method for AFB1 detection in standard solutions, providing a strong foundation for its application in the quantitative detection of AFB1 in real samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.4.2 Establishment of Standard Curve for AFB1 in Pu-erh Tea Extracts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe NH2-PEG-AuNR substrates prepared in this study could effectively detect the SERS spectra of AFB1 specimens.\u0026nbsp;Conventional methods for detecting\u0026nbsp;AFB1 in tea are often subject to interference from the complex matrix of tea infusions\u003csup\u003e3\u003c/sup\u003e.\u0026nbsp;However, NH2-PEG-AuNRs can selectively adsorb the AFB1 molecules in tea extracts onto the substrate, which improves SERS detection. Figure 4C presents the SERS spectra of Pu-erh tea extract spiked with AFB1 at concentrations ranging from 1 \u0026times; 10\u003csup\u003e\u0026minus;1\u003c/sup\u003e to 1 \u0026times; 10\u003csup\u003e\u0026minus;9\u003c/sup\u003e ng/mL, with a total of 9 concentrations, with an instrumental LOD of 1 \u0026times; 10\u003csup\u003e\u0026minus;9\u003c/sup\u003e ng/mL. Figure 4D presents the correlation curve between the AFB1 concentration in Pu-erh tea extracts and the intensity of characteristic SERS peak at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e. Figure 4D presents the correlation curve between the AFB1 concentration in Pu-erh tea extract and the intensity of the characteristic SERS peak at 1267 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e. The established linear regression equation is given by y = \u0026minus;20.41x + 214.52. The correlation coefficient was R\u003csup\u003e2\u003c/sup\u003e = 0.9953, which indicates that this method can effectively detect different concentrations of AFB1 in Pu-erh tea extract. Moreover, the detection limit and sensitivity achieved in this assay comply with regulatory standards established by China and the European Union\u003csup\u003e4\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.5 Development of a Quantitative Prediction Model for AFB1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.5.1 Development of a Quantitative Prediction Model for AFB1 in Standard Solutions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, a PLS regression model was developed for the quantitative prediction of AFB1 concentrations in standard solutions. The model\u0026rsquo;s performance metrics, presented in Table 1,\u0026nbsp;reveal an Rc value of 0.9924, Rp value of 0.9936, and an RPD of 8.93, as calculated using eq. (4).\u0026nbsp;These high statistical values indicate that the PLS model exhibits excellent predictive accuracy, rendering it suitable for the accurate prediction of AFB1 in standard solutions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.5.2 Development of a Quantitative Prediction Model for AFB1 in Real Samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA PLS quantitative prediction model was developed for the quantification of AFB1 in Pu-erh tea extracts. The model\u0026rsquo;s performance metrics, presented in Table 1,\u0026nbsp;indicate the Rc value of 0.9750, Rp value of 0.9743, and RPD of 4.40, as calculated using eq. (4). These results confirm the high predictive accuracy and reliability of the model for the quantitative prediction of AFB1 in Pu-erh tea extracts.\u003c/p\u003e\n\u003cp\u003eCompared with traditional detection methods\u003csup\u003e21,22\u003c/sup\u003e, the SERS-based approach using NH2-PEG-AuNRs developed in the present study offers the advantages of simplicity, speed, and efficiency for rapid detection of AFB1 in Pu-erh tea extracts. This study established a novel method for rapid quantitative prediction of AFB1 in Pu-erh tea extracts.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003cstrong\u003e.5.3 External Validation of the Quantitative Prediction Model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe developed PLS quantitative prediction model was applied to spectral data collected from four types of dark tea extracts. The results, summarised in Table 1, indicate that the RPD value for all tested samples exceeded 1.9, confirming the broad applicability of the assay for various dark tea matrices.\u003c/p\u003e\n\u003cp\u003eThe recovery rates were calculated as the ratio of the AFB1 concentration detected in the tea extract to the actual spiked concentration, and the results are presented in Table 2. The spiked recovery rates ranged from 98.54% \u0026plusmn; 5.55% to 102.54% \u0026plusmn; 3.71%, with an RSD of \u0026lt;8.06%. These results indicate the high accuracy and precision of the developed assay, highlighting its robustness and reliability for the quantitative determination of AFB1 in real tea samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e PLS quantitative forecasting model results.\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"652\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 151px;\"\u003e\n \u003cp\u003eSamples\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 84px;\"\u003e\n \u003cp\u003eModel parameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 171px;\"\u003e\n \u003cp\u003eTraining sets\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 194px;\"\u003e\n \u003cp\u003ePrediction sets\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 52px;\"\u003e\n \u003cp\u003eRPD\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003eRMSECV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003eRc\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003eRMSEP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003eRp\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eAFB1 ethanol solution\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"6\" style=\"width: 84px;\"\u003e\n \u003cp\u003eLVs=7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9924\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e8.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eLabelling of Pu-erh tea\u0026nbsp;infusion\u0026nbsp;with AFB1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9743\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e4.40\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eQianliang tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9431\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e2.39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eFuzhuan tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9528\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e2.38\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eKangzhaun tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9165\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e2.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 151px;\"\u003e\n \u003cp\u003eLiubao tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 88px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 108px;\"\u003e\n \u003cp\u003e0.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.9013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 52px;\"\u003e\n \u003cp\u003e1.91\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2.\u003c/strong\u003e Determination and recovery of AFB1 concentration in different dark teas (n=15).\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"601\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003eSamples\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003eAdditive concentration\u003c/p\u003e\n \u003cp\u003e（-lg C ng/mL）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003eDetected concentration\u003c/p\u003e\n \u003cp\u003e（-lg C\u0026nbsp;ng/mL）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003eRecovery rate\u003c/p\u003e\n \u003cp\u003e（%）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003eRSD\u003c/p\u003e\n \u003cp\u003e（%）\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 120px;\"\u003e\n \u003cp\u003eQianliang tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e4.99\u0026plusmn;0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e99.84\u0026plusmn;5.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.19\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.15\u0026plusmn;0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e102.54\u0026plusmn;3.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.62\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.08\u0026plusmn;0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e101.08\u0026plusmn;3.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.20\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 120px;\"\u003e\n \u003cp\u003eFuzhuan tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.13\u0026plusmn;0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e102.63\u0026plusmn;6.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.14\u0026plusmn;0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e102.49\u0026plusmn;3.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.82\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.03\u0026plusmn;0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e100.44\u0026plusmn;3.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.23\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 120px;\"\u003e\n \u003cp\u003eKangzhaun tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e4.96\u0026plusmn;0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e99.14\u0026plusmn;7.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e8.06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.19\u0026plusmn;0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e103.23\u0026plusmn;3.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.07\u0026plusmn;0.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e101.05\u0026plusmn;2.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e2.91\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 120px;\"\u003e\n \u003cp\u003eLiubao tea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e4.93\u0026plusmn;0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e98.54\u0026plusmn;5.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e5.63\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e6.11\u0026plusmn;0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e101.91\u0026plusmn;4.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e4.63\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e7.02\u0026plusmn;0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e100.32\u0026plusmn;3.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 120px;\"\u003e\n \u003cp\u003e3.22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003ch2\u003e2.6\u0026nbsp;Discussion\u003c/h2\u003e\n\u003cp\u003eAFB1, a contaminant produced by \u003cem\u003eAspergillus\u003c/em\u003e species, is among the most toxic mycotoxins and is a potential contaminant in the processing, transportation, and storage of Pu-erh tea. In this study, a SERS-based approach to detecting AFB1 was developed that uses the hydrogen bonding interactions between NH2-PEG-AuNRs and AFB1. This method enables rapid and accurate detection of AFB1 in complex tea matrices.\u003c/p\u003e\n\u003cp\u003eMost SERS sensors detect AFB1 in samples through physical or chemical adsorption.\u0026nbsp;However, sensors that rely on chemical adsorption often\u0026nbsp;have the disadvantages of complex preparation processes and limited adsorption efficiency. As indicated in Table S1, the method proposed in the current study offers a broader detection range, lower detection limit, and higher efficiency, addressing a critical gap in the literature with respect to\u0026nbsp;AFB1 detection\u0026nbsp;in tea matrices. Au@AgNPs have previously been employed as a signal-enhancing substrate, with melamine used as an adsorbent; they facilitate AFB1\u0026nbsp;adsorption in tea oil through hydrogen bonding with\u0026nbsp;a linear range of 10\u003csup\u003e\u0026minus;4\u003c/sup\u003e to 10\u003csup\u003e\u0026minus;7\u003c/sup\u003e mol/L and a detection limit 10\u003csup\u003e\u0026minus;8\u003c/sup\u003e mol/L\u003csup\u003e23\u003c/sup\u003e. However, tea leaf extracts have a more complex background\u0026nbsp;than tea oil does, and therefore, AFB1 detection is more challenging in teas. Compared with Fe3O4@AuNFs-cDNA and Au-4MBA@AgNSs-Apt\u003csup\u003e18\u003c/sup\u003e, NH2-DNA1-CS-Fe3O4 and SH-DNA2-ADANR\u003csup\u003e24\u003c/sup\u003e, NH2-Rh-Au@Ag CSNPs, and AuNP dimers-MXenes assembly\u0026nbsp;aptamer sensors\u003csup\u003e17\u003c/sup\u003e,\u0026nbsp;the NH2-PEG-AuNR-enhanced substrate proposed in this study involves a simpler preparation process and faster AFB1\u0026nbsp;binding kinetics. Moreover, compared with a previously proposed DSNS-AuNP immunosensor\u003csup\u003e19\u003c/sup\u003e, the NH2-PEG-AuNR substrate prepared in this study is more stable and has a 4-month validity.\u0026nbsp;The method proposed in this study yielded satisfactory results in the detection of AFB1 in tea samples. However, in the real world, microbial toxins are often complex and diverse, and this study did not explore simultaneous detection of multiple microbial toxins. However, in theory, the substrate prepared in this study can adsorb other toxins and enhance the Raman fingerprints of other toxins through hydrogen bonding interactions. Nevertheless, the binding energy between the substrate and mycotoxins and the potential for cross-reactivity among different mycotoxins remain uninvestigated. In the future, we will calculate the binding energies between the substrate and different mycotoxins by using\u0026nbsp;density functional theory\u0026nbsp;and investigate the ability of the substrate to simultaneously detect different mycotoxins in the presence of multiple mycotoxins.\u003c/p\u003e\n\u003cp\u003eThe present study\u0026nbsp;successfully developed\u0026nbsp;a quantitative method for detecting AFB1 by using SERS with NH2-PEG-AuNRs and established a rapid and effective approach to AFB1 analysis of dark tea. The proposed method involves a simple pretreatment process; overcomes\u0026nbsp;key challenges associated with\u0026nbsp;aptamer sensors, which have limited binding efficiency for AFB1; and overcomes key challenges associated with immunosensors, which exhibit instability.\u0026nbsp;Additionally, this method involves\u0026nbsp;a broader detection range and a lower detection limit than\u0026nbsp;conventional approaches do.\u0026nbsp;By enhancing the precision and reliability of AFB1 detection, this study\u0026nbsp;makes a notable contribution that can improve the quality control and safety assessment of tea products. This method has potential for broader application in the field of food safety; it can be used to detect other toxins in\u0026nbsp;complex food matrices.\u003c/p\u003e\n\u003cp\u003eThis study successfully developed a SERS technique using NH2-PEG-AuNRs that enables rapid and sensitive detection of AFB1 in tea. The NH2-PEG-AuNRs demonstrated excellent SERS activity, leading to a broad linear detection range of 1 \u0026times; 10\u003csup\u003e\u0026minus;1\u003c/sup\u003e to 1 \u0026times; 10\u003csup\u003e\u0026minus;17\u003c/sup\u003e ng/mL, with R\u003csup\u003e2\u003c/sup\u003e = 0.9971 and LOD = 1 \u0026times; 10\u003csup\u003e\u0026minus;17\u003c/sup\u003e ng/mL for AFB1 in standard solutions. Additionally, a PLS quantitative prediction model was established that had a prediction accuracy of 99.36% and an RPD value of 8.93. For the detection of AFB1 in tea extracts, the proposed method yielded a linear range of 1 \u0026times; 10\u003csup\u003e\u0026minus;1\u003c/sup\u003e to 1 \u0026times; 10\u003csup\u003e\u0026minus;9\u003c/sup\u003e ng/mL, an R\u003csup\u003e2\u003c/sup\u003e value of 0.9953, and an LOD of 1 \u0026times; 10\u003csup\u003e\u0026minus;9\u003c/sup\u003e ng/mL. A separate PLS quantitative prediction model for AFB1 in Pu-erh tea extract was developed that had a prediction accuracy of 97.43% and an RPD value of 4.40. External validation experiments further confirmed the robustness of the model, with prediction accuracies exceeding 90.00%, RPD values above 1.9, and recovery rates between 98.54% \u0026plusmn; 5.55% and 102.54% \u0026plusmn; 3.71%. Notably, the entire detection process can be completed in 20 min, with this including sample\u0026ndash;substrate mixing and drying time, and online detection can be completed in less than 1 min. The method\u0026rsquo;s broad detection range, low detection limit, and rapid analysis time address a key gap in AFB1 detection in tea by offering a means of rapid detection. This approach holds substantial potential for application in tea processing.\u003c/p\u003e"},{"header":"3. Methods","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Materials and Reagents\u003c/h2\u003e\u003cp\u003eThe dark tea used in this study was Pu-erh ripe tea (Menghai Tea Factory, Dayi Pu-erh Ripe Tea, 2008), Qianliang tea (Baishaxi Tea Factory Limited Liability Company, Hunan Province, 2007), porcupine brick tea (Anhua, Yiyang, Hunan Province, China Tea Hunan Anhua Tea Factory Limited Company, 2014), Kang brick tea (Ya\u0026rsquo;an, Sichuan Province, 2012), and Liu Baotong tea (Wuzhou, Guangxi Province Maosheng Tea Co. Ltd., 2006). AFB1 (99.5%) was purchased from the Yunmo Quality Inspection Standard Material Centre. NH\u003csub\u003e2\u003c/sub\u003e-PEG3400-NH\u003csub\u003e2\u003c/sub\u003e (97%) and R6G (95%) were purchased from Shanghai Yi En Chemical Technology Co. Ethanol was purchased from Shanghai Zhan Yun Chemical Co. The following chemical reagents, all of which were of analytical grade, were obtained from Sinopharm Chemical Reagent Co. Ltd.: CTAB, HAuCl\u003csub\u003e4\u003c/sub\u003e-4H\u003csub\u003e2\u003c/sub\u003eO, NaBH\u003csub\u003e4\u003c/sub\u003e, NaOL, AgNO\u003csub\u003e3\u003c/sub\u003e, ascorbic acid, HCl, H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e, NaOH.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e3.2. Preparation and Characterisation of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs\u003c/h2\u003e\u003cp\u003eThe method used to synthesise NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was adapted from a previously reported method, with some modifications\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. First, the synthesised gold nanorods were dispersed in an amino-polyethylene glycol-amino (NH\u003csub\u003e2\u003c/sub\u003e-PEG-NH\u003csub\u003e2\u003c/sub\u003e) solution and allowed to react for 12 h to achieve surface modification. Following functionalisation, the solution was centrifuged at 6000 rpm for 20 min, and the supernatant was discarded. The remaining lower layer was redispersed in ultrapure water and centrifuged again under identical conditions. This purification process was repeated three times to ensure the removal of excess reagents and byproducts. The final NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs were stored at 4\u0026deg;C in the dark.\u003c/p\u003e\u003cp\u003eThe ultraviolet-visible (UV-vis) spectra of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs were measured using an enzyme labeller (SpectraMax M2, Molecular Devices, USA) in the detection range of 300\u0026ndash;1000 nm, with a step size of 1 nm. The morphological structure of the NH₂-PEG-AuNRs was examined using transmission electron microscopy (TEM) at an accelerating voltage of 180.0 KV (HT7700, Hitachi, Japan). The elemental composition of the synthesised nanorods was analysed using an ultra-high resolution field emission scanning electron microscope (SU8600, Hitachi) equipped with an energy-dispersive spectrometer. The zeta potential of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was measured using a nanoparticle sizer (Zetasizer Lab, Malvern Panalytical, UK). Each measurement was performed in triplicate, and the instrument automatically determined average values.\u003c/p\u003e\u003cp\u003eTo evaluate the applicability and accuracy of the newly developed method for the rapid quantitative detection of AFB1 in Pu-erh tea extracts, four representative dark teas were selected for comparison: Qianliang tea, Kangzhi tea, Poria tea, and Liubao tea. Each dark tea extract was spiked with AFB1 at three different concentrations: 1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e, 1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e, and 1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e ng/mL. The spiked tea extracts were then mixed at a 1:1 ratio with the NH2-PEG-AuNR substrate for SERS detection.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003e3.3. Validation of SERS Performance of NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs\u003c/h2\u003e\u003cp\u003eThe sensitivity, reproducibility, stability, and enhancement factor (EF) of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs as a SERS substrate were evaluated using Raman spectrometry (Lab RAM HR Evolution, Horiba Jobin Yvon, France). R6G was used as a reference analyte, with a 785-nm laser used in the detection range of 400\u0026ndash;1800 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. The detection time was 8 s, the accumulation number was 1, and the power was 25%. To determine the sensitivity of the NH₂-PEG-AuNRs, a series of R6G standard solutions with concentration gradients of 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;9\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;10\u003c/sup\u003e, and 10\u003csup\u003e\u0026minus;\u0026thinsp;11\u003c/sup\u003e mol/L were prepared. Each solution was mixed in a 1:1 ratio with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs, and 15 spectra were recorded per concentration. The correlation coefficient between the Raman signal intensity and R6G concentration was analysed to confirm the sensitivity of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs. For reproducibility testing, 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e mol/L R6G was mixed with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs in a 1:1 ratio, and 20 randomly selected measurement points were analysed. The reproducibility of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs was confirmed by calculating the RSD values of the signal intensities across these measurement points. To assess the stability of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs, a Raman assay was performed every 15 days over a 4-month period by mixing 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e mol/L R6G with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs in a 1:1 ratio, and 15 valid spectra were recorded at each time point. The RSD values of the signal intensities across the time points were calculated. The EF of the NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs as a SERS substrate was calculated using the following Eq.\u0026nbsp;2\u003csup\u003e7\u003c/sup\u003e:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:\\text{E}\\text{F}=\\left({\\text{I}}_{\\text{S}\\text{E}\\text{R}\\text{S}}/{\\text{C}}_{\\text{S}\\text{E}\\text{R}\\text{S}}\\right)/\\left({\\text{I}}_{\\text{R}\\text{S}}/{\\text{C}}_{\\text{R}\\text{S}}\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere I\u003csub\u003eSERS\u003c/sub\u003e and C\u003csub\u003eSERS\u003c/sub\u003e denote the Raman signal intensity and concentration of R6G bound to the SERS substrate, respectively, and I\u003csub\u003eRS\u003c/sub\u003e and C\u003csub\u003eRS\u003c/sub\u003e denote the signal intensity of the Raman spectra and concentration of R6G in its free form, respectively.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e3.4 SERS Acquisition of AFB1\u003c/h2\u003e\u003cdiv id=\"Sec19\" class=\"Section3\"\u003e\u003ch2\u003e3.4.1 Optimisation of Detection Conditions\u003c/h2\u003e\u003cp\u003eTo prepare an AFB1 standard solution, AFB1 powder was completely dissolved in ethanol; this was followed by serial 10-fold dilutions from 10\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 10\u003csup\u003e\u0026minus;\u0026thinsp;17\u003c/sup\u003e ng/mL. A total of 17 samples were prepared and stored in sealed containers at \u0026minus;\u0026thinsp;20\u0026deg;C. To determine the optimal mixing ratio between AFB1 and NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs, a 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e ng/mL AFB1 standard solution was mixed with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs at different ratios (3:1, 2:1, 1:1, 1:2, and 1:3) in centrifuge tubes. The mixtures were then ultrasonicated for 10 min, and 5 \u0026micro;L of the mixture was aspirated and dripped onto gold-plated slides, which were dried in an oven at 55\u0026deg;C. After, Raman spectra were collected using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France) by using a 785-nm laser with an acquisition range of 400\u0026ndash;1800 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, an acquisition time of 1 s, and an accumulation number of 1. A total of 15 spectra were collected for each set of experiments and averaged. The optimal mixing ratio was determined by comparing the signal intensity across the average spectra.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section3\"\u003e\u003ch2\u003e3.4.2 SERS Detection of AFB1 in Standard Solutions\u003c/h2\u003e\u003cp\u003eOn the basis of the optimal mixing ratio, AFB1 standard solutions of varying concentrations were thoroughly mixed with NH2-PEG-AuNRs in centrifuge tubes. The mixtures were ultrasonicated for 10 min, and 5 \u0026micro;L of the mixture was aspirated and dropped on a gold-plated slide, which was dried in an oven at 55\u0026deg;C. Raman spectra were acquired using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section3\"\u003e\u003ch2\u003e3.4.3 SERS Detection of AFB1 in Real Samples\u003c/h2\u003e\u003cp\u003eTo prepare tea extracts spiked with AFB1, 5 g of each type of tea (i.e., Pu-erh tea, Liu Bao tea, Qian Liang tea, Kang brick tea, and Por brick tea) was weighed using an electronic balance and transferred to a 500-mL conical flask. Subsequently, 250 mL of boiling water (95\u0026ndash;100\u0026deg;C) was added, and the mixture was steeped for 5 min. The resulting tea infusions were filtered, collected into clean conical flasks, and rapidly cooled to room temperature.\u003c/p\u003e\u003cp\u003eAFB1 standard solutions were added to the Pu-erh tea extract to achieve a series of concentrations ranging from 10\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 10\u003csup\u003e\u0026minus;\u0026thinsp;9\u003c/sup\u003e ng/mL in 10-fold decrements, with a total of nine samples obtained with different concentrations. The AFB1 standard solution was similarly added to the Liubao tea, Qianliang tea, Kang brick tea, and Por brick tea extracts to obtain final concentrations of 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e, 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e, and 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e ng/mL, respectively.\u003c/p\u003e\u003cp\u003eOn the basis of the optimal mixing ratio, AFB1 standard solutions of varying concentrations were thoroughly mixed with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNR substrates in centrifuge tubes. The mixtures were ultrasonicated for 10 min, and 5 \u0026micro;L of the mixture was aspirated and dripped onto gold-plated slides, which were dried in an oven at 55\u0026deg;C. Raman spectra were acquired using a confocal Raman spectrometer (Lab RAM HR Evolution, Horiba Jobin Yvon, France) by using a 785-nm laser at a detection time of 8 s, a cumulative number of 2, and a detection range of 400\u0026ndash;1800 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. In total, 15 spectra were collected for each sample concentration.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003e3.5 Data Processing and Analysis\u003c/h2\u003e\u003cdiv id=\"Sec23\" class=\"Section3\"\u003e\u003ch2\u003e3.5.1 Establishment of the AFB1 Standard Curve\u003c/h2\u003e\u003cp\u003eTo quantify AFB1, a correlation curve was constructed using the Raman signal intensity at 1267 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e as the y-axis and the AFB1 concentration as the x-axis. For the AFB1 standard solutions, a total of 17 samples with varying concentrations were analysed, with 15 valid spectral data points per sample, for a total of 255 spectral data points. For the Pu-erh tea extracts spiked with AFB1, a total of 9 samples with different concentrations were tested, and 15 valid spectral data points were obtained for each sample, for a total of 135 data points. The standard curves of AFB1 in the standard solution and Pu-erh tea extract were established using this method.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec24\" class=\"Section3\"\u003e\u003ch2\u003e3.5.2 Development and External Validation of the AFB1 Quantitative Prediction Model\u003c/h2\u003e\u003cp\u003ePartial least-squares (PLS) regression models were developed to quantitatively predict AFB1 concentrations in both standard solutions and spiked Pu-erh tea extracts by using SERS with NH\u003csub\u003e2\u003c/sub\u003e-PEG-AuNRs as the substrate. To construct these models, the spectral data were first divided into calibration and prediction sets in a 2:1 ratio, with Raman spectral data serving as input variables and AFB1 concentrations as output variables.\u003c/p\u003e\u003cp\u003eFor external validation, the PLS prediction model was applied to four types of dark tea extracts, namely Qianliang tea, Kang brick tea, Por brick tea, and Liu Bao tea. Each extract was spiked with 3 concentrations of AFB1, and 15 valid Raman spectra were collected per concentration. This resulted in a dataset of 45 spectral measurements per tea extract. The developed PLS model was then applied to these datasets to assess its predictive accuracy.\u003c/p\u003e\u003cp\u003ePLS regression is a widely used technique in quantitative predictive modelling. It establishes a quantitative forecasting model by predicting one or more dependent variables (output variables) from known independent variables (input variables). External validation of a PLS quantitative forecasting model involves applying an established PLS model to a new dataset that has not been used during model development. This dataset is referred to as a test or validation set\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The metrics used to evaluate the performance of this quantitative prediction model are the coefficient of determination (Rc) for the training set and its root mean square error of calibration, the coefficient of determination (Rp) for the prediction set and its root mean square error of prediction (RMSEP), and the residual prediction deviation (RPD). RPD is defined as the ratio of the standard deviation of the samples in the prediction set to the RMSEP. In general, RPD\u0026thinsp;\u0026lt;\u0026thinsp;1.4 indicates that the developed model has poor prediction performance and the model is unreliable; 1.4\u0026thinsp;\u0026lt;\u0026thinsp;RPD\u0026thinsp;\u0026lt;\u0026thinsp;1.8 indicates that the developed model can be used for rough prediction; 1.8\u0026thinsp;\u0026lt;\u0026thinsp;RPD\u0026thinsp;\u0026lt;\u0026thinsp;2.0 indicates that the model has favourable prediction performance and is suitable for quantitative analysis of a set of samples; and RPD\u0026thinsp;\u0026gt;\u0026thinsp;2 indicates that the model has excellent quantitative prediction ability. In this study, the model evaluation indices were calculated as follows\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:\\text{R}=\\sqrt{1-\\frac{\\sum\\:_{\\text{i}=1}^{\\text{n}}{({\\text{y}}_{\\text{i}}-{\\widehat{\\text{y}}}_{\\text{i}})}^{2}}{\\sum\\:_{\\text{i}=1}^{\\text{n}}{({\\text{y}}_{\\text{i}}-{\\stackrel{-}{\\text{y}}}_{\\text{m}})}^{2}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:\\text{R}\\text{M}\\text{S}\\text{E}=\\sqrt{\\frac{\\sum\\:_{\\text{i}=1}^{\\text{n}}{({\\text{y}}_{\\text{i}}-{\\widehat{\\text{y}}}_{\\text{i}})}^{2}}{\\text{n}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:\\text{R}\\text{P}\\text{D}=\\frac{\\text{S}\\text{D}}{\\text{R}\\text{M}\\text{S}\\text{E}\\text{P}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{y}}_{i}\\)\u003c/span\u003e\u003c/span\u003e are the true value and the model predicted value of the first sample in the training or prediction set, respectively; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\stackrel{-}{y}}_{m}\\)\u003c/span\u003e\u003c/span\u003e denotes the average of the true values of the samples in the training or prediction set; and n denotes the number of samples in the training or prediction set.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\u003ch2\u003e3.6 Software\u003c/h2\u003e\u003cp\u003eAll spectral data in this study were baseline corrected and smoothed for noise reduction by using Labsix Sigma software (Horiba Scientific, NJ, USA) with uniform parameters. All maps were plotted in Origin 2023b (OriginLab Corp, MA, USA), and the PLS model was developed and implemented in MATLAB 2014a (Mathworks, Natick, USA).\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eAll authors declare no financial or non-financial competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eS.C., Y.Z. designed experiments and data analyses; S.C., Y.Z., Q.Y., andY.W. performed the experimental work; Q.L. provided technical support; S.C.,Y.Z. analyzed data and wrote the manuscript; J.N., Q.L. supervised theproject. All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e\u003cp\u003eThis study was financially supported by National Nature Science Foundation of China (32202543), National Key Research and Development Program (2021YFD1601102), National Key Research and Development Program (2023YFD1601300), Key Research and Development Program of Anhui Province (202104h04020023), the earmarked fund for CARS (CARS-19).\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eThe data that support the findings of this study are available on request from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eDing, Q. et al. Comparison of hypoglycemic effects of ripened pu-erh tea and raw pu-erh tea in streptozotocin-induced diabetic rats. Rsc Advances, 9(6), 2967-2977 (2019).\u003c/li\u003e\n\u003cli\u003eZhang, W. et al. Multiplex SERS-based lateral flow immunosensor for the detection of major mycotoxins in maize utilizing dual raman labels and triple test lines. Journal of Hazardous Materials, 393, 122-348 (2020). \u003c/li\u003e\n\u003cli\u003eZhang, X. et al. Fungal flora and mycotoxin contamination in tea: Current status, detection methods and dietary risk assessment-A comprehensive review. Trends in Food Science \u0026amp; Technology, 127, 207-220 (2022). \u003c/li\u003e\n\u003cli\u003eFan, L. et al. Dummy molecularly imprinted solid-phase extraction-SERS determination of AFB1 in peanut. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 288, 122-130 (2023).\u003c/li\u003e\n\u003cli\u003eOstry, V., Malir, F., Toman, J., \u0026amp; Grosse, Y.. Mycotoxins as human carcinogens-the IARC monographs classification. Mycotoxin Research, 33(1), 65-73(2017). \u003c/li\u003e\n\u003cli\u003eHassan, F. et al. Aflatoxin B1 in Rice: Effects of storage duration, grain type and size, production site, and season. Journal of Food Protection, 85(6), 938-944 (2022). \u003c/li\u003e\n\u003cli\u003eGholizadeh, S. et al. Ultrasound-assisted solvent extraction combined with magnetic ionic liquid based-dispersive liquid-liquid microextraction for the extraction of mycotoxins from tea samples. Journal of Food Composition and Analysis, 114, 104-831 (2022). \u003c/li\u003e\n\u003cli\u003eGazioglu, I., \u0026amp; Kolak, U. Ochratoxin a levels in food and beverage samples from Turkey. Acta Alimentaria, 47(2), 189-194. https://doi.org/10.1556/066.2018.47.2.7\u003c/li\u003e\n\u003cli\u003eCina, M.et al. Martinez, L. D., \u0026amp; Cerutti, S. Development of a novel UHPLC-MS/MS method for the determination of ochratoxin A in tea. Heliyon, 7(4), e06-663 (2021). \u003c/li\u003e\n\u003cli\u003eWang, Q. et al. Li, J., Deng, Q., Zhang, X., Wang, S., \u0026amp; Chen, M. High-performance electrochemiluminescence sensors based on ultra-stable perovskite quantum dots@ZIF-8 composites for aflatoxin B1 monitoring in corn samples. Food Chemistry, 410, 135-325 (2023). \u003c/li\u003e\n\u003cli\u003eZhang, B. et al. Simple \u0026quot;signal-on\u0026quot; photoelectrochemical aptasensor for ultrasensitive detecting AFB1 based on electrochemically reduced graphene oxide/poly(5-formylindole)/Au nanocomposites. Biosensors \u0026amp; Bioelectronics, 134, 42-48(2019). \u003c/li\u003e\n\u003cli\u003eChen, Z. et al. Discrimination of toxigenic and non-toxigenic Aspergillus flavus in wheat based on nanocomposite colorimetric sensor array. Food Chemistry, 430, 137-048(2024).\u003c/li\u003e\n\u003cli\u003eLu, L., Yu, R., \u0026amp; Zhang, L. AFB1 colorimetric aptamer sensor for the detection of AFB1 in ten different kinds of miscellaneous beans based on gold nanoparticles and smartphone imaging. Food Chemistry, 421, 136-205 (2023).\u003c/li\u003e\n\u003cli\u003eDou, X. et al. Construction of a nanoscale metal-organic framework aptasensor for fluorescence ratiometric sensing of AFB1 in real samples. Food Chemistry, 416, 135-805(2023).\u003c/li\u003e\n\u003cli\u003eLu, X. et al. Target-driven switch-on fluorescence aptasensor for trace aflatoxin B1 determination based on highly fluorescent ternary CdZnTe quantum dots. Analytica Chimica Acta, 1047, 163-171 (2019).\u003c/li\u003e\n\u003cli\u003eDing, S. et al. Electromagnetic theories of surface-enhanced raman spectroscopy. Chemical Society Reviews, 46(13), 4042-4076(2017).\u003c/li\u003e\n\u003cli\u003eWu, Z.et al.. Ti3C2Tx MXenes loaded with Au nanoparticle dimers as a surface-enhanced raman scattering aptasensor for AFB1 detection. Food Chemistry, 372, 131-293(2022). \u003c/li\u003e\n\u003cli\u003eHe, H.et al. Bridging Fe3O4@Au nanoflowers and Au@Ag nanospheres with aptamer for ultrasensitive SERS detection of aflatoxin B1. Food Chemistry, 324, 126-832(2020). \u003c/li\u003e\n\u003cli\u003eLi, Y. et al. Microarray surface enhanced raman scattering based immunosensor for multiplexing detection of mycotoxin in foodstuff. Sensors and Actuators B-Chemical, 266, 115-123(2018).\u003c/li\u003e\n\u003cli\u003eDuy Vu, T. et al. A facile paper-based chromatography coupled Au nanodendrite on nickel foam for application in separation and SERS measurement. Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy, 313, 124-137(2024).\u003c/li\u003e\n\u003cli\u003eMedina, Det al. Current role of modern chromatography and mass spectrometry in the analysis of mycotoxins in food. Trac-Trends in Analytical Chemistry, 135, 116-156(2021).\u003c/li\u003e\n\u003cli\u003eZhou, H. et al. Mycotoxins in tea ((Camellia sinensis (L.) Kuntze)): Contamination and dietary exposure profiling in the Chinese population. Toxins, 14(7), 452 (2022).\u003c/li\u003e\n\u003cli\u003eChen, W. et al. Optimum synthesis of Au@Ag nanoparticle as plasma amplifier to detect trace concentration of AFB1 via object-binder-metal SERS method. Journal of Food and Drug Analysis, 30(4), 603-613(2022).\u003c/li\u003e\n\u003cli\u003eChen, Q. et al. A large raman scattering cross-section molecular embedded SERS aptasensor for ultrasensitive Aflatoxin B1 detection using CS-Fe3O4 for signal enrichment. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 189, 147-153(2018).\u003c/li\u003e\n\u003cli\u003eJiao, T. et al.. Aggregation triggered aflatoxin B1 determination in foodstuff employing 5-aminotetramethylrhodamine decorated gold-silver core-shell nanoparticles in surface enhanced raman scattering. Sensors and Actuators B-Chemical, 331, 129-424(2021).\u003c/li\u003e\n\u003cli\u003eDeng, Y. et al. Trimetallic Au@PtPd mesoporous nanorods as efficient electrocatalysts for the oxygen reduction reaction. Acs Applied Energy Materials, 1(9), 4891-4898(2018). \u003c/li\u003e\n\u003cli\u003eLi, L. et al. Rapid detection of multiple colorant adulteration in Keemun black tea based on hemp spherical AgNPs-SERS. Food Chemistry, 398, 133-841(2023).\u003c/li\u003e\n\u003cli\u003eWang, Y. et al. Evaluating taste-related attributes of black tea by micro-NIRS. Journal of Food Engineering, 290, 110-181(2021).\u003c/li\u003e\n\u003cli\u003eViegas, T. et al. Determination of quality attributes in wax jambu fruit using NIRS and PLS. Food Chemistry, 190, 1-4(2016). \u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Scheme ","content":"\u003cp\u003eScheme 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Aflatoxin B1, Dark tea, Amino-terminally modified polyethylene glycol gold nanorods, Surface-enhanced Raman scattering","lastPublishedDoi":"10.21203/rs.3.rs-7144899/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7144899/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThere is a risk of aflatoxin B1 (AFB1) contamination during dark tea processing, transportation, and storage. Given the potent carcinogenicity of AFB1, there is an urgent need to develop rapid and sensitive detection methods to address the requirements of tea safety monitoring. A rapid and trace detection method for AFB1 based on surface-enhanced Raman spectroscopy (SERS) using NH2-PEG-AuNRs was established and successfully applied to dark tea. The established method exhibited a linear dynamic range (LDR) for the AFB1 standard solution from 10⁻\u0026sup1; to 10⁻\u0026sup1;⁷ ng/mL, with a limit of detection (LOD) of 10⁻\u0026sup1;⁷ ng/mL. The partial least-squares (PLS) prediction model achieved an accuracy of 99.36%. For dark tea infusions, the LDR extended from 10⁻\u0026sup1; to 10⁻⁹ ng/mL, and the LOD of 10⁻⁹ ng/mL. The PLS model achieved an accuracy of 97.43%, and recovery rates ranged from 98.54\u0026ndash;102.54%. This establishes a robust methodology for quantifying AFB1 in dark tea.\u003c/p\u003e","manuscriptTitle":"NH2-PEG-AuNR-Based Surface-Enhanced Raman Spectroscopy for Rapid Detection of AFB1: Dark Tea Safety Assessment","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-29 08:01:42","doi":"10.21203/rs.3.rs-7144899/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"00a796c6-5db1-4617-bd91-f7d7c7eab506","owner":[],"postedDate":"July 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":52246185,"name":"Biological sciences/Biochemistry"},{"id":52246186,"name":"Biological sciences/Biological techniques"},{"id":52246187,"name":"Biological sciences/Biotechnology"},{"id":52246188,"name":"Physical sciences/Chemistry"},{"id":52246189,"name":"Earth and environmental sciences/Environmental sciences"}],"tags":[],"updatedAt":"2025-10-17T06:09:06+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-29 08:01:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7144899","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7144899","identity":"rs-7144899","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}