Nanogold sol plasmon discattering assay for trace carbendazim in tea coupled aptamer with Au3+-glyoxal-carbon dot nanocatalytic reaction

Carbendazim (CBZ) is a broad-spectrum fungicide, which is toxic to mammals. Therefore, it is very necessary to establish a sensitive detection for food safety. An experiment found that CDFe exhibited excellent catalysis for the nano-indicator reaction of HAuCl4-glyoxal to produce gold nanoparticles (AuNPs) and that the generated AuNPs have a very strong surface-enhanced Raman scattering (SERS) effect at 1613 cm−1 in the presence of Victoria blue B molecular probes, and resonance Rayleigh scattering (RRS) signals at 370 nm. The aptamer (Apt) suppressed the catalysis of CDFe to cause the SERS and RRS signals decreasing. With the addition of CBZ, the specific Apt reaction occurred to restore the catalysis of CDFe, and resulting in a linear increase in the signals of RRS and SERS. As a result, this new nanocatalytic amplification indicator reaction was coupled with a specific Apt reaction of carbendazim (CBZ), to construct a new CDFe catalytic amplification-aptamer SERS/RRS discattering assay for ultratrace CBZ, which was used to analyze CBZ in tea samples with satisfactory results. In addition, this biosensoring platform can be also used to assay profenofos.

proposed a solvothermal method using corn stalks as raw materials to prepare a CDs-based nanohybrid dual-emission system and used it as a fluorescence sensor to detect Hg 2+ , with a concentration range of 0-40 µM Hg 2+ and a detection limit (DL) of 9.0 nM. Srinivasan et al. (6) prepared a MoS 2 nanosheets-DNA-CD complex and used them as fluorescent probes to establish a fluorescence method for the detection of Hg 2+ , with a DL of 1.02 nM. Wang et al. (7) coupled CD with a glyphosate antibody to construct antibody-labeled CD (IgG-CD). The prepared IgG-CD could specifically recognize glyphosate in a concentration range of 0.01-80 µg/ml. Recently, it has been found that the quantum yield of CDs can be improved by doping negative electron heteroatoms such as N, S, and P (8-10). Konar et al. (11) developed nitrogen-doped CDs as a new fluorescent probe for a quantitative determination of l-cysteine and found a linear relationship in the range of 10-210 µM and a DL of 75.6 nM. As a new nanomaterial, iron-doped carbon dots (CD Fe ) have superior optical stability, cheap raw materials, high-contrast efficiency, and good adaptability. It is expected that nanomaterials will be better used in the application of biomarkers, bioimaging, and drugs. For example, Huang et al. (12) used glutathione (GSH), ethylenediaminetetraacetic acid disodium salt (EDTA), and FeSO 4 as raw materials and synthesized Fe-CQDs by hydrothermal method, which was suitable for T1-weighted MR/FL bimodal bioimaging and had biocompatibility. Fluorescent CD Fe was prepared by hydrothermal carbonization of methyl thymol blue sodium salt (MTB) and FeCl 3 as precursors, a fluorescence sensing system based on CD Fe ratio of H 2 O 2 and glucose was designed by Zhu et al. (13), which quantifies H 2 O 2 and glucose in the range of 0-133 and 0-300 µm with a DL of 0.47 and 2.5 µM, respectively. To the best of our knowledge, there are few reports on the environmentally friendly preparation and high catalytic activity of CD Fe , which was used for nanocatalytic amplification-aptamer surface (Apt)-enhanced Raman scattering (SERS) quantitative analysis of trace carbendazim (CBZ).
Surface-enhanced Raman scattering can sensitively detect the concentration of molecules adsorbed on the nanosurface and provide a wealth of molecular structure information (14). Qi et al.
(15) developed a multi-cytosine SERS nanoprobe, which could detect Hg 2+ quickly, sensitively, and highly selectively and shows a good SERS response in the concentration range of 0.1-1,000 nM. Similarly, a non-pretreatment SERS detection was completed by using a plasma platform functionalized (16). To improve the selectivity and detection range, some selective bioreactions such as Apt and peptide were used in SERS (17)(18)(19). Yang et al. (20) established an Apt-SERS sensor for the highly selective detection of thrombin, combining the nanocatalyzed SERS reaction of gold nanoparticles (AuNPs) on HAuCl 4 -H 2 O 2 with the Apt reaction, which could effectively detect the concentration range of 0.13-53.33 nmol/L Pb 2+ . A stable and cleavable enzyme-coated reductive graphene oxide (rGO)-pdAu probe was prepared by Xu et al.
(21) as a signal probe and proposed the visual colorimetric and electrochemiluminescence dual-mode sensing of Pb 2+ . Zong et al. (22) also proposed a colorimetric and SERS dual-mode telomerase activity detection method. On the contrary, resonance Rayleigh scattering (RRS) has the advantages of strong signal, simple operation, good reproducibility, and no introduction of molecular probe reagents (23). Therefore, coupling of SERS and RRS two spectral techniques is of great significance for providing two choices and being of two methods and vantages.
Carbendazim is a broad-spectrum fungicide, which has a wide range of applications and can control many crop diseases caused by fungi. However, its residues can cause liver disease and chromosomal aberration, which are toxic to mammals. Therefore, it is very necessary to establish a sensitive detection method for trace amounts of CBZ in food. At present, the determination methods of CBZ residues in fruits include liquid chromatography, ion chromatography, and liquid chromatography-mass spectrometry/mass spectrometry (24). The most commonly used method was liquid chromatography, since it has strong selectivity. However, these methods have some disadvantages such as expensive equipment, time-consuming, and the pretreatment method requiring complex processes such as liquid-liquid extraction and solid-phase extraction. With the development of new technologies and materials, some new nanoanalytical methods for the determination of CBZ have also been reported (25)(26)(27)(28). Although these assays have their advantages, their shortcomings cannot be ignored. However, there are few reports on the quantitative determination of CBZ by Apt-SERS method. Therefore, in this article, we avoided the use of organic solvents and developed a hydrothermal procedure to prepare CD Fe with high activity and stability, and the new nanocatalytic indicator reaction of CD Fe -HAuCl 4 -glyoxal (GX) was coupled with Apt to establish a new SERS/RRS dimode assay for the detection of CBZ.

. . Reagents
Aptamer of CBZ (Apt CBZ ) with a sequence of GGG CAC ACA ACA ACC GAT GGT CCA GCC ACC CGA ATG ACC AGC CCA CCC GCC ACC CCG CG; aptamer of oxytetracycline . . Preparation of Fe carbon quantum dots (CD Fe ) We first took an accurately weighed 0.0180 g of ferrocene, then added 0.0180 g of sodium citrate and 30 ml of water, ultrasound for 15 min, transferred it into the autoclave, sealed it, and placed it in the muffle furnace, set the program to 30 min, raised it to 180 • C, and kept warm for 2 h. After the reaction was over, the solution was cooled at room temperature with the ice water and diluted to 30 ml with water to obtain a 0.6 mg/ml CD Fe that was calculated as Fer. The results of the catalytic experiment showed that CD Fe had a stronger catalytic effect without dialysis, thus, CD Fe without dialysis was selected as the catalyst in the experiment.
Next, we took an accurately weighed 0.0180 g of Hemin and then added 0.0180 g of sodium citrate and 30 ml water to the digestion tank. A concentration of 0.6 mg/ml of ferrocene carbon dots (CD HM) was prepared by microwave digestion at 180 • C for a reaction time of 2 h.

. . Experimental method
In a 5.0 ml stoppered test tube, a certain concentration of CBZ solution and 180 µl 10 nmol/L Apt were sequentially added, then 120 µl of 0.6 mg/L CD Fe solution was added and bathed in 80 • C water for 5 min. Subsequently, 80 µl of 1 mg/ml HAuCl 4 , 180 µl of 0.01 mol/L HCl, and 60 µl of 0.087 mol/L GX were added and diluted to 1.5 ml. After 35 min in a water bath at 80 • C, the reaction was terminated with ice water, and the solution was taken in a quartz dish, at volt = 350 V, excited slit = emission slit = 5 nm, and scanned with a fluorescence spectrophotometer to obtain an RRS spectrum. The RRS value at 370 nm was measured and recorded as I 370nm , and the RRS value without adding CBZ was recorded as a blank value ( I 370 nm)0 . For SERS detection, 100 µl of 10 µmol/L VBB was added and mixed well, the solution was scanned in the Raman spectrum under the conditions of a Raman spectrometer light source power of 1.0 mW and a slit of 25.0 µm and measured the SERS intensity at 1,613 cm −1 as I 1,613cm−1 , without CBZ as a blank value (I 1,613cm−1 ) 0 , and the value of I 1,613cm−1 =I 1,613cm−1 -(I 1,613cm−1 ) 0 was calculated.
. Results and discussion . . Analysis principle HAuCl 4 and GX did not react at room temperature, and it was extremely difficult to react even with underwater bath heating. However, due to its small particle size, when CD was added to the reaction system, the CD Fe could strongly catalyze the reaction of HAuCl 4 by GX to form AuNPs. The AuNPs generated by the reaction had a strong RRS signal. After adding the VBB probe molecule, AuNPs showed a strong SERS signal. After the addition of Apt, it could be adsorbed on CD Fe to reduce its catalytic activity. The system generated fewer AuNPs, which reduced the SERS/RRS signal of the system. Next, when the target molecule CBZ was added, it would bind specifically to the Apt and then CD Fe would be released so that its catalytic activity would gradually recover. The higher the concentration of the CBZ added, the higher the concentration of CD Fe released and the more the AuNPs generated, and the signal of SERS/RRS of the system was linearly enhanced. Therefore, a new SERS/RRS analysis method for the determination of trace CBZ could be established ( Figure 1).

. . . Selection of CD Fe preparation conditions
The preparation of CD Fe in the muffle furnace was tested. The conditions for the preparation of CD Fe were tested by the single-factor transformation method. The effects of three factors, namely, Fer dosage, reaction time, and reaction temperature, were investigated, and the prepared CD Fe was used to catalyze the GX-HAuCl 4 reaction ( Supplementary Figures 1-3). The slope k of the linear relationship between the catalyst concentration and the I 370nm value of the system was used to evaluate the catalytic effect of the catalyst. That is, the higher the slope, the stronger the catalytic effect. Supplementary Figure 4A shows that as the reaction time increases, the catalytic effect first increases and then decreases. When the reaction time is 120 min, the catalytic activity is the strongest, so the CD Fe reaction time is selected as 120 min. As can be seen from Supplementary Figure 4B, the effect of the catalyst first increases and then decreases with the increase in reaction temperature. As can be seen from Supplementary Figure 4C, with the increasing amount of ferrocene, the catalytic effect first increases and then decreases. When the ferrocene used is 0.018 g, the catalytic activity is the strongest. Therefore, the amount of Fer added is 0.018 g. In summary, the optimal conditions for the preparation of CD Fe by the Fer hydrothermal method are as follows: Fer addition of 0.018 g, adding 30 ml of water, and heating at 180 • C for 120 min.
. . . CD Fe fluorescence/RRS/Abs spectrum The fluorescence spectrum of CD Fe (volt = 500 V, excited slit = emission slit = 10 nm) shows that as the excitation wavelength (λ EX ) shifts from 300 to 380 nm, there is still no fluorescence intensity (Supplementary Figure 5A). Based on the reported data .
/fnut. . of carbon quantum dots, it is found that some of them will have fluorescence behavior at the excitation wavelength of 320 nm, thus, when the excitation wavelength is 320 nm, CD Fe shows weak fluorescence behavior (Supplementary Figure 5B). In addition, when the excitation wavelength is 320 nm, the fluorescence peak is roughly distributed at 420 nm. The excitation wavelength of 320 nm was selected to obtain the fluorescence spectra of CD Fe with different concentrations, and the results showed that the fluorescence intensity of CD Fe at 420 nm has a good linear relationship with its concentration within the range of 0-0.128 g/L CD Fe concentration (Supplementary Figure 5B). The fluorescence peak of CD HM is roughly distributed at 445 nm, and there is a good linear relationship between the fluorescence intensity and the concentration in the concentration range of 0-0.096 g/L (Supplementary Figure 5C). The RRS spectrum obtained by the synchronous scanning technique of fluorimeter is a simple and sensitive technique for studying the scattering characteristics of nanoparticles. For the RRS spectrum of CD Fe and CD HM (volt = 350 V, excited slit = emission slit = 10 nm), the RRS intensity at 370 nm showed a linear correlation with the concentration (Supplementary Figures 5D, E), indicating that the particles with larger molecular weight than Fer were formed during the preparation of CD Fe . For the absorption spectrum, CD Fe has a wide absorption band at 200-600 nm, and its absorbance gradually increases with an increase in CD Fe concentration, and an absorption peak appears at 250 nm (Supplementary Figure 5F). The absorption band of CD HM is relatively narrower than that of CD Fe , but its absorbance also becomes stronger with an increase in CD HM concentration, and its absorption peak appears at 300 nm (Supplementary Figure 5G).

. . . CD Fe infrared and SERS
Infrared spectroscopy (FTIR) shows that there are 7 characteristic peaks of CD Fe (Figure 2A): 485.9 cm −1 is the characteristic peak of Fe-O, 783.0 cm −1 was caused by the C-H out-of-plane bending vibration, 1,040.3 and 1,144.1 cm −1 are attributed to the C-H in-plane deformation of the benzene compound, and 1,418.5 cm −1 is due to the tensile vibration of the C-O. The 1,636.1 cm −1 is caused by the vibration of the framework of the substituted benzene C = C, and the 3,363.7 cm −1 is caused by the tensile vibration of C-OH, which may also be caused by the crystal water in the test sample. In the figure, the peak of ferrocene at 462.85 cm −1 is caused by the stretching vibration of the cyclolocene, and Fe, 815.39 cm −1 is attributed to πC-H, 1,000.47 cm −1 is formed by the C-H deformation vibration on the cyclolocene, 1,105.61 cm −1 is caused by the reverse stretching vibration of the ring, and 1,406.75 cm −1 is attributed to the C = C stretching vibration on the locene ring. The Fe-O peak was formed while CD Fe was formed by Fer hydrothermal method, which indicated that its structure has changed after calcination at high temperature. Moreover, CD Fe has oxygen-containing groups, which indicates that it has superior hydrophilicity and stability. At the same time, the infrared spectra of CD HM and its precursor Hemin are also given ( Figure 2B). It can be seen from the figure that CD HM peaks are at 2,361.3, 1,138.2, and 1,038.9 cm −1 .
/fnut. . compared to Hemin. We studied the SERS spectrum of CD Fe with AgNPs as the SERS substrate, and the results showed that there was no obvious SERS peak ( Figure 2C). Finally, the SERS spectra of CD HM were also studied using AgNPs as the substrate. Similar to CD Fe , there was no obvious SERS signal ( Figure 2D).
. . . Transmission electron microscope and x-ray di raction of CD Fe Figure 3A is the TEM image taken at 50 nm resolution after the sample was diluted with ethanol after centrifugation. Figures 3B, C are the TEM images taken directly at 20 and 10 nm resolution on CD Fe , respectively, with the particle size ranging from 1.5 to 3.5 nm. Figure 3D is the HAADF-STEM image of CD Fe . Figure 3E shows that Fe is uniformly distributed on the carbon quantum dots. CD Fe also contains C and O elements, but the content of Fe is relatively small. The results of the combined XRD analysis show that Fe can bind not only to cyclopentadiene but also to O, that is, Fe is anchored on the surface of carbon quantum dots in the form of Fe-O and Fer-Fe bonds. Figure 3F is the EDS of CD Fe . The XRD of CD Fe was studied. The XRD analysis results show ( Figure 3G) that the CD Fe produced by the Fer hydrothermal method has a characteristic peak at about 25 • and 40-45 • , respectively, which are the peaks of graphitic carbon (002) and (101), indicating that CD Fe has the amorphous carbon structure. At the same time, the phases of iron and iron compounds are not seen in XRD, which indicates that there are no obvious Fe particles in CD Fe . Fe may be combined with cyclopentadiene and O. Similarly, in the XRD study of CD HM and Hemin, CD HM has the same amorphous carbon structure as CD Fe , but there is a peak around 10 • ( Figure 3H). Figure 3I is the full XPS spectrum of CD Fe . The characteristic peaks at 284.7, 531.8, and 711.3 eV correspond to C1s, O1s, and Fe2p, respectively. It can be seen from the figure that CD Fe contains three elements: carbon, oxygen, and iron. Their relative atomic percentage contents are 69.42, 28.31, and 2.27%, respectively. Figure 3J is the high-resolution energy spectrum of C1s of CD Fe . Three element characteristic peaks are fitted at 284.7, 286.5, and 288.8 eV, corresponding to the C-C/C-H, C-O, and C=O bonds, respectively. Figure 3K is the highresolution energy spectrum of O1s of CD Fe , the characteristic peaks of the elements at 531.7, 531.8, and 532.1 eV correspond to Fe-O, -OH, and C-O bonds, respectively; Figure 3L is the high-resolution energy spectrum of Fe2p of CD Fe , the peak at 711.3 eV is Fe2p 3/2 , and the peak at 724.7 eV corresponds to Fe2p 1/2 . The characteristic peaks at these two peaks can indicate that Fe in CD Fe exists in the chemical state of +3 valence Frontiers in Nutrition frontiersin.org . /fnut. . and in the oxidized state. According to the infrared spectrum and XPS analysis of CD Fe , the CDs were successfully doped with iron.

. . . Stability of nanosol
The prepared CD Fe and CD HM were stored in a refrigerator at 4 • C. The stability of CD Fe was examined by fluorescence  (FL), RRS, and UV. It was found that the signals of CD Fe changed a little within 15 days, with RSDs of 3.2, 5.2, and 2.9% ( Figures 4A, B). The test stated that the prepared CD Fe had good stability over time. In addition, the stability of CD Fe in different concentrations of NaCl solution was investigated. The results showed that the FL, RRS, and Abs signals of CD Fe were stable in different concentrations of NaCl solution (Figures 4C, D). The experimental results show that the preparation method has good stability. Under naked eye observation, there will be no visible precipitation within 10 days, and its stability is good. Although CD HM can make molecular spectra, after reacting at 180 • C for 2 h, cooling and standing for half an hour will produce obvious precipitation. It is suspected that it may be carbonized for too long, then try to burn at 30, 60, and 90 min, but all of them were Frontiers in Nutrition frontiersin.org . /fnut. . precipitated, thus, the experiment was continued without using this catalyst.

. . SERS spectra of nanocatalysis analysis system
With AuNPs as the SERS substrate, adding VBB molecules can generate strong SERS signals. When CD Fe /CD HM nanosol is used as a catalyst, GX-HAuCl 4 reacts to generate a large number of AuNPs. After adding VBB molecular probes, the system generated three SERS peaks at 1,202, 1,396, and 1,613 cm −1 . The SERS peaks at 1,202 and 1,396 cm −1 are caused by the C-H bond deformation in the saturated CH = CH bond. The SERS peak at 1,613 cm −1 is caused by the strong stretching vibration of the C = C bond in the dicyclopentadiene framework. The SERS signal at 1,613 cm −1 is the strongest, and the SERS intensity gradually increases with the CD Fe /CD HM concentration (Figures 5A, B). As can be seen from the slope of SERS spectrum of CD Fe /CD HM catalytic system, the catalytic effect of CD Fe is significantly stronger than that of CD HM system. Afterward, CD Fe was used as a catalyst to catalyze the GX-HAuCl 4 system at 85 • C in a water bath. The results showed that Au 3+ could be reduced to AuNPs. After adding Apt to the GX-HAuCl 4 catalytic system, the SERS signal at 1,613 cm −1 of the system gradually decreased with the increase in Apt concentration ( Figures 5C, D). For the CD Fe -Apt-HAuCl 4 -GX reaction, when the target molecule CBZ/PF was added, they would specifically bind to Apt, thereby releasing CD Fe to restore its catalytic activity and releasing more CD Fe when the concentration of the target molecules was increased, more AuNPs were generated. In addition, the signal intensity of the SERS spectrum increased linearly with the concentration of the target (Figures 5E, F) in a certain linear range when VBB was added.

. . RRS spectra
The HAuCl 4 -GX-HCl-CD Fe catalytic system has three spectral peaks at 280, 370, and 550 nm. As the concentration of CD Fe increases, the concentration of AuNPs increases gradually and the RRS peak at 370 nm is the largest ( Figure 6A). Figure 6B shows the RRS spectrum of CD HM as a catalyst. It was found in the experiment that the RRS signal decreased with the addition of Apt (Figures 7C, D). For the CD Fe -Apt-HAuCl 4 -GX reaction, when the   target molecule CBZ/PF was added, the ligand-specific binding would occur and the CD Fe would be released so that the reaction system restored its catalytic activity. Within a certain linear range, as the number of target molecules increased, the signal strength of RRS would also increase ( Figures 6E, F). The results showed that the sensitivity of the reaction system to CBZ was very high and the linearity was relatively good.

. . TEM of the analytical system
The reaction solution of the system was obtained according to the experimental method of Section Experimental method, and 10 µl solution was dropped on the surface of the ultra-thin carbon film three times to obtain TEM of the system. Figure 7A shows that the catalytic activity of CD Fe is inhibited when CBZ is not added, and the system produces very few AuNPs, with a particle size between 20 and 100 nm and good dispersibility. After adding CBZ, CD Fe was released, its catalytic activity was restored, and the number of AuNPs generated increased, with a particle size between 25 and 150 nm and aggregation ( Figure 7B). There are very strong Au-M (Au-Mα, Au-Mβ, Au-Mγ) characteristic peaks signal around 2.131 keV in the energy spectrum and a relatively strong Au-Lα peak around 9.712 keV. There are also signals of Au-Lβ (Au-Lβ 1 , Au-Lβ 2 , Au-Lβ 3 , Au-Lβ 4 ) characteristic peaks around 11.440 and 11.583 keV, which proved that AuNPs were generated in this system.

. . CD Fe catalysis and aptamer inhibition
Under experimental conditions, both CD Fe and CD HM could catalyze the reduction of HAuCl 4 by GX to produce AuNPs. In a certain concentration range, with the increase in catalyst concentration, the number of AuNPs was generated, and the SERS and RRS signals of the system had a linear relationship with .

. Selection of analysis conditions
With the increase in the amount of CD Fe , the number of AuNPs generated by the reaction increased, and the RRS signal of the system was enhanced continuously. When the amount of CD Fe was 0.048 mg/L, the I of the system reached the maximum. Thus, the dosage of CD Fe was 0.048 mg/L (Supplementary Figure 6A). Apt could interact with CD Fe to form CD Fe -Apt on its surface, thereby weakening its catalytic effect. When CBZ was added, Apt combined with CBZ, causing CD Fe to be exposed again and restore its catalytic activity, and the catalytic reaction proceeded smoothly. In excess, Apt wrapped on the surface of CD Fe and still inhibited the catalytic reaction. In addition, the results showed that when the dosage of Apt was 4.6 nmol/L (Supplementary Figure 6B), the system had the best response effect. When the amount of GX was continuously increased, due to the rapid generation of AuNPs, the aggregation of AuNPs would lead to a decrease in I in the system. The concentration of GX was 3.48 mmol/L (Supplementary Figure 6C). When the dosage of HAuCl 4 was continuously increased, the accumulation of AuNPs occurred due to the rapid formation of AuNPs. At the same time, the blank also reacted, and the I of the system decreases. Thus, when the dosage of HAuCl 4 was 0.053 mg/ml, the I of the system reached the maximum (Supplementary Figure 6D). When there was no HCl in the system, the rate of AuNP generation was too fast, the AuNPs were aggregated, and the I of the system was low. As the concentration of HCl increased, the rate of AuNPs generation gradually slowed down, the amount of AuNPs produced by the reaction also increased, and the RRS signal of the system continued to increase. When the amount of HCl was continuously increased, the production of AuNP was inhibited, and the I of the system decreases instead. Therefore, the amount of HCl used in this experiment was 1.2 mmol/L (Supplementary Figure 6E). The temperature had a great influence on the reaction system. The higher the temperature, the faster the reaction speed of the system. At 60 • C, the reaction did not occur, but when the temperature was too high, the reaction speed was too fast and the system was unstable. When the reaction temperature was 80 • C, the I of the system reached the maximum. Therefore, the reaction temperature was selected as 80 • C in the experiment (Supplementary Figure 6F). At the reaction temperature of 80 • C, the reaction of the system gradually accelerated with time and reached the highest value at 35 min. After 35 min, the reaction occurred completely, but the blank also began to react. Therefore, the reaction time of this experiment was selected as 35 min (Supplementary Figure 6G).

. . Working curve
Under the selected optimal test conditions, the linear regression equations of all the methods for detecting the analytes presented in this article are shown in Table 1 with the concentration of the analyte and the corresponding RRS value I 370nm and the SERS value I 1,613cm−1 . It could be seen from the table that the SERS method was better and its sensitivity was higher. The RRS method was simpler than the SERS without molecular probes. Compared with the reported analytical methods for the determination of CBZ (Table 2), this new SERS method was one of the most sensitive methods.

. . Analysis application
The tea sample weighing 5.0 g was put into a 100-ml conical flask, then 50 ml methylene chloride was added. After vortex oscillation for 3 min, it was pumped and filtered in a sand core funnel paved with Celite 545, evaporated and concentrated at 45 • C, and blown dry with nitrogen. A 0.2 g C 18 was added and fixed to volume with 1 ml ethanol, whirled for 2 min, and filtrated by 0.22 µM membrane filtration to obtain the sample solution. The CBZ was measured according to the experimental method. The measurement results are shown in Table 3, and the SERS results are in agreement with that of the HPLS method. The recovery rate was 92.0-109%, with a relative standard deviation (RSD) of 5.4-8.4%. The results indicated the SERS assay was accurate and reliable. According to the National Food Safety Standard maximum residue limits of pesticides in foods (GB 2763-2021), the maximum content of carbendazim in apples is 5 mg/kg. The CBZ content in industrial samples did not exceed the standard.

. Conclusion
A new CD Fe was prepared and characterized in detail. A new SERS/RRS dimode indicator reaction of CD Fe -Au (III)-GX was constructed. It was coupled with the Apt reaction to construct a dimode Apt biosensor analysis platform for detecting CBZ. In addition, this new strategy can be used to detect PF. There are some limitations to this study, including limited Apt species and complex combination of catalyst and Apt, and further studies on looking for nanoprobes that can be used as both catalysts and specific recognition functions to expand the application of this platform.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.