Introduction

Food is the basic necessity for people’s survival. With the rapid development of the economy, food and environmental pollution are major issues related to national health and safety. Nowadays, more and more harmful chemicals are entering the environment of human living (Ye et al. 2017; Ji et al. 2023). As a contact fungicide, dithiocarbamate compound (DTC) is effective against pathogens and widely used for crops, such as vegetables and fruits. However, with the introduction of such pesticides into the environment, their physical, chemical, and biological processes can be transformed into harmful substances and cause disease (Wang et al. 2020). As a common dithiocarbamate insecticide, the metabolite of mancozeb is ethylene thiourea (ETU), which has been classified as a category III carcinogen by the International Agency for Research on Cancer. Long-term consumption of ETU can cause adverse effects such as embryonic malformations and thyroid abnormalities (Dall’Agnol et al. 2021). Therefore, it should be of great significance to develop a fast, sensitive, accurate, and effective method to analyze and detect low-concentration fungicides contained in the environment and food.

In recent years, it has witnessed the great progress in pesticide residue detection technology which leads to the proposal of various technical approaches for detecting pesticide residue in fruits and vegetables (Łozowicka et al. 2017). These methods include gas chromatography-mass spectrometry (Wang et al. 2022), liquid chromatography-mass spectrometry (Yang et al. 2022), immunoassay (Xu et al. 2022a, b), and SERS. SERS is a rapid detection technique that has been utilized for food identification, quantitative analysis, and classification of food pollutants. SERS is well known for its sensitivity and ability to quickly recognize fingerprints, making it a valuable analytical tool for both qualitative and quantitative detection of food samples (Weng et al. 2019; Han et al. 2021; Zhou et al. 2021; Yang et al. 2023). SERS was first discovered in 1974, its principle is to amplify the Raman signals of molecules close to the surface of metal nanostructures as SERS substrate through the electromagnetic enhancement effect and chemical enhancement mechanism of noble metal nanostructures. In other words, the SERS substrate relies on plasmonic nanostructures, particularly Au and Ag NPs with strong surface plasmon resonance (SPR) effect (Atanasov et al 2019a, b). The performance of SERS can be adjusted by changing the size, morphology, and composition of the nanomaterials (Dick et al. 2016; Bi et al. 2023; Ye et al. 2020; Van Vu et al. 2023; Ying et al. 2023). Because of the development of nanotechnology, the SERS detection technique has entered a new stage of rapid progress. For instance, Wu et al.developed a SERS active substrate based on Au nanorods, which is applied to lakes and various fruits contaminated with methyl parathion (Wu et al. 2019). Deng demonstrated a SERS sensor covered by Ag nanowire network. The wrinkled biomimetic Ag as SERS substrate, with high density of hotspots, exhibited strong SERS effect and enabled the detection of pesticide molecules (Deng et al. 2023). Single plasmonic metal as SERS substrate with finite property cannot completely meet the requirements of analysis, such as weak affinity between analyte and substrate and low sensitivity of SERS. Versatile substrate with exceptional performance is crucial for practical applications. To address this, it is necessary to develop a SERS composite substrate that combines capture ability, rapid enrichment ability, ease of separation, and plasma resonance effect. Compared with ordinary nanoparticles, magnetic nanocomposites can achieve efficient enrichment for the target material and rapid separation by an external magnetic field (Zheng et al. 2021; Tiryaki et al. 2024). Notably, the combination of magnetic-plasmonic nanocomposites can achieve significant enhancements in the SERS performance.

In recent years, graphene oxide (GO) has gained attention in various fields due to its unique physical and chemical properties, especially the large specific surface area. Its large specific surface area and conjugated structure contribute to its fast adsorption rate for aromatic ring organic pollutants. In addition, GO and other semiconductor nanomaterials, such as TiO2, ZnO, MoS2, CuO, Fe2O3, Cu2O, Cu2S, Cu2Se, CuTe, CdS, PbS, AgS, CdTe, and NiO, demonstrate high SERS activity, and they contribute to the chemical enhancements (Sakano et al 2008; Shukla et al 2023; Avadaiappan et al 2022; Liu et al 2024; Fu et al 2024; Zheng et al 2023). The introduction of GO not only overcomes some inherent defects of SERS, but also has an excellent synergistic effect with traditional precious metal NPs as substrate (Weis et al. 2018, Suzuki 2019, Meenakshi et al. 2023). These characteristics of GO not only provide advantages for the preparation of composite substrate materials, but also expand the application field of SERS to a certain extent.

In order to achieve rapid and sensitive detection of pesticide residues in food, a three-layer multi-functional SERS composite substrate (Fe3O4/GO/Au) was designed and constructed, as depicted in Scheme 1. Combining the advantage of each layer, the prepared composite substrate for SERS analysis can achieve rapid separation and enrichment of target molecules without requiring excessive pre-treatment time, and the substrate provides abundant hotspots and amplifies the signal through synergy. Due to the excellent properties of Fe3O4/GO/Au above, mancozeb was directly detected. To verify the feasibility of the method, mancozeb in pakchoi was tested, and successful detection with high sensitivity and recovery rate was achieved. The novel composite nanomaterial substrate paves the way for the detection of pesticide residues and opens a new path for analysis in the food, environment, and biomedicine field.

Scheme 1
scheme 1

Schematic illustration of the fabrication of Fe3O4/GO/Au NPs

Full size image

Material and methods

Chemicals and materials

Chloroauric acid, ferric chloride, ferrous sulfate, graphite powder (industrial grade), potassium permanganate, trisodium citrate, and ascorbic acid were purchased from Aladdin Reagent (Shanghai) Co., Ltd. (Shanghai, China). Concentrated sulfuric acid, isopropanol, polyvinylimide, sodium nitrate, and sodium chloride were obtained from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). All glassware and stirring bars were soaked in aqua regia for 24 h and rinsed several times with ultrapure water before use. All chemicals were used without further purification. Milli-Q ultrapure water was used for all solution preparations in the study.

Preparation of the materials for SERS substrate

Preparation of Fe3O4 particles

Co-precipitation method was employed to synthesize Fe3O4 NPs. Initially, 80 mL of deionized water was added to 250 mL of three-neck flask and stirred mechanically. To remove oxygen, nitrogen flowed through the solution. Subsequently, 0.8641 g of ferrous sulfate and 0.4448 g of ferric chloride were added to the flask and completely dissolved quickly. The solution exhibited an orange-yellow. The temperature of the solution was then raised to 60 °C, and the pH of the solution was adjusted to 9.0–10.0 by ammonia. The solution rapidly changed from orange to black. After 30 min, the temperature was increased to 85 °C, and the mixture was aged for 30 min under the protection of nitrogen. Then, the three-neck flask was removed and allowed to cool to room temperature. The Fe3O4 solution was transferred to the beaker, and the Fe3O4 NPs were separated from the solution by the external magnetic field. The separated Fe3O4 NPs were washed alternatively with deionized water and absolute ethanol several times. Finally, they were dried at 50 °C in the oven to obtain magnetic Fe3O4 NPs (Jinzi and Chunyan 2020).

Preparation of graphene oxide

GO was obtained using a modified Hummers method (Lin et al. 2021). Initially, 40.0 mL of concentrated sulfuric acid was measured and mixed with 1.0 g of graphite powder and 0.5 g of sodium nitrate. The mixture was mechanically stirred for 30 min in an ice-water bath. Subsequently, 5.0 g of potassium permanganate was added gradually in ten small portions, and the reaction was continued for 60 min. The resulting mixture was then transferred to the water bath, kept temperature at 35 °C, and stirred continuously for 60 min. Then, 50 mL of deionized water was slowly added, and the temperature was adjusted to 80 °C; the reaction was carried out for 30 min. Finally, 100 mL of deionized water and 30% hydrogen peroxide were rapidly added until no bubbles were generated, resulting in an earthy yellow product. Once the reaction is complete, the obtained product is dried in a vacuum drying oven at 35 °C and labeled for future use.

Preparation of Fe3O4/GO

A total of 0.5 g of dried Fe3O4 NPs was added into 25 mL of isopropyl alcohol and dissolved by ultrasonic waves for 40 min. A total of 0.15 mL of (3-aminopropyl)triethoxysilane (APTES) was added gradually drop by drop, followed by ultrasonic treatment for 10 min (Akamine et al. 2021). Then, the mixture was transferred to the flask and kept at 85 °C with reflux for 3 h. After cooling to room temperature, the product was washed alternately with ethanol and deionized water and separated the amino-modified Fe3O4 by using a magnet. The obtained product was added to 35 mL of deionized water and dispersed completely. Similarly, 0.15 g of graphene oxide was dissolved in 35 mL of deionized water to obtain graphene oxide solution. The obtained solution was mixed and sonicated for 30 min. Stir the mixture overnight at room temperature, then magnetically separate it, and wash it several times with ethanol and deionized water alternately. Finally, dry it under vacuum at 60 °C, label it, and set it aside.

Preparation of Au NPS

To prepare the Au seeds, first, 0.5 mL of HAuCl4 (1%) was added to 30 mL of deionized water and stirred at room temperature for 10 min. Then, 0.2 mL of sodium citrate (1%) solution was added and continued for about 5 min with stirring. Followed by the addition of 1 mL of NaBH4 (0.375%) solution, and stirring for 10 min, orange-red Au seeds appeared. Next, 50 mL of deionized water in a conical flask was accurately measured. 0.35 mL of HAuCl4 (1%), and 0.35 mL of Au seeds were added and stirred for 5 min to create a homogeneous mixture at room temperature. Then, 0.1mL of sodium citrate (1%) solution and 0.1mL of polyvinylpyrrolidone (1%wt) were added, followed by stirring for 3 min. The temperature was then raised to 45 °C. After stirring for an additional 5 min, 0.5 mL of ascorbic acid (0.3 M) solution was added dropwise, and blue-violet Au sol was obtained (Wu et al. 2019).

Preparation of Fe3O4/GO/Au

Initially, 200 mg of Fe3O4/GO was added to 100 mL of deionized water, and 1 mL of polyethyleneimine in 49 mL of deionized water was added. The solution was mixed under ultrasonication for 40 min at room temperature to ensure complete dispersion, followed by mechanical stirring. After 24 h, magnetic separation was performed, followed by alternating washes with deionized water and ethanol for 3–4 times. The product was then dried. Next, 100 mg of amino-functionalized Fe3O4/GO was accurately weighed and dispersed in 25 mL of deionized water with ultrasonication. Then, 20 mL of Au sol was added, followed by ultrasonication for 15 min. The obtained Fe3O4/GO/Au was dried in the oven at 60 °C (Ghanei and Hosseinifar 2020).

Characterization

The X-ray diffraction (XRD) data was recorded on the D8ADVANCE diffractometer, with Cu as the target source, and scanned in the 2θ range of 10 ~ 80°. The UV–visible absorption peak and the infrared characteristic peak of the prepared material were measured, respectively. The structure and morphology of synthesized Fe3O4/GO/Au nanoparticles were studied by transmission electron microscopy (TEM). Raman spectra were obtained using a confocal microscopy Raman spectrometer system (Horiba, LabRAMHR) with an excitation wavelength of 785 nm. The laser power and integration time are set to 20 mW and 12 s, respectively. Twenty microliters of the Fe3O4/GO/Au aqueous suspension mixed with 20 μL of R6G, respectively, and sonicated for 30 min. R6G was detected by using 785 nm laser.

Application of active substrates

Detection of mancozeb with standard solution

A stock solution of mancozeb was prepared in pyridine at a concentration of 100 ppm. This stock solution was then diluted with deionized water to obtain a series of mancozeb concentrations (0.5 ppm, 1 ppm, 10 ppm, 25 ppm, and 50 ppm) as the working standard solution. Prior to SERS measurement, 1 mL of the mancozeb standard solution with different concentrations was added to 5 mL of centrifuge tube containing 3 mL of Fe3O4/GO/Au. One hundred microliters of sodium chloride solution was added and sonicated. Thirty microliters of a portion of the resulting solution was analyzed using laser confocal microscopy and Raman spectroscopy (Atanasov et al. 2019a, b). For SERS measurements, a 785 nm laser was employed as an excitation source, with an acquisition time of 12 s and a laser power of 20 mW. Spectra were collected in the range of 400–1600 cm−1.

Detection of mancozeb in real sample

Fresh pakchoi was purchased from the local supermarket and was used as tested samples. The QuEchERS (quick, easy, cheap, effective, rugged, and safe) method was followed to extract mancozeb from fresh pakchoi (Pan et al. 2021). Specifically, the purchased pakchoi was washed and dried, and 10 g of the sample was weighed into a 50 mL centrifuge tube. Then, 5 mL of acetonitrile, 1g of sodium chloride, and 0.2 g of anhydrous sodium acetate were added sequentially. The tube was shaken well and vortexed, followed by centrifugation at 4200 rpm for 5 min. The supernatant of the pakchoi appeared green. Two milliliters of the supernatant was transferred to a 10 mL centrifuge tube containing an appropriate amount of anhydrous magnesium sulfate and graphitized carbon black which were used to primarily remove the influence of chlorophyll and vitamins in the pakchoi. The solution was centrifuged for 5 min to obtain a colorless supernatant. The suspension was filtered and used for SERS measurements. In order to illustrate the feasibility of the method, a spiked recovery experiment was performed. The spiked solution with different concentrations was added into the unprocessed pakchoi. The QuEchERS method was used to extract mancozeb from pakchoi for SERS measurement. The detection procedure was the same as for the standard solution determination.

The test of activity, stability, and repeatability of Fe3O4/GO/Au

In order to quantify the SERS ability, the analytical enhancement factor (EF) was estimated. The formulation (LOD = 3δ/k) was used to determine the limit of detection (LOD). To examine the stability of Fe3O4/GO/Au, time-dependent SERS measurements were performed. To assess the repeatability of Fe3O4/GO/Au, three samples were obtained following the same conditions, at 11 random locations on the substrate.

Results and discussion

Characterization of Fe3O4/GO/Au nanocomposites

The crystal structures of Fe3O4, GO, and Fe3O4/GO/Au NPs were characterized by XRD, and the results were displayed in Fig. 1. The XRD spectrum of Fe3O4 showed strong diffraction peaks with 2θ at 30.2°, 35.6°, 43.2°, 57.1°, and 62.7°, corresponding to 220, 311, 400, 511, and 440, respectively, consistent with the spinel structure of Fe3O4 (JCPDS No.19–0629) (He et al. 2022). It was indicated that the prepared sample is pure Fe3O4. A new sharp peak appears at 11.3°, which belongs to the peak of GO (JCPDS No.47–1743) (Supriyanto et al. 2018). In addition, there is a clear diffraction peak at 26.6°, which may be due to insufficient oxidation time resulting in incomplete oxidation of graphite. In the Fe3O4/GO/Au spectrum, there are two main diffraction peaks with 2θ at 38.0° and 44.6°, which correspond to the (111) and (200) crystal planes of Au (JCPDS No. 04–0784) (Wu et al. 2023), respectively. Except for the diffraction peaks of Fe3O4, Au, and GO, no other diffraction peaks were found. This may be due to the destruction of oxygen-containing functional groups on the surface of oxidized graphene. In a word, three-layer multifunctional Fe3O4/GO/Au composite materials were successfully fabricated.

Fig. 1
figure 1

XRD of the GO, Fe3O4, and Fe3O4/GO/Au

Full size image

The Fe3O4/GO/Au composite materials were identified by infrared spectra, ranging from 500 to 4000 cm−1, as shown in Fig. 2. The IR spectra of GO showed peaks at 1721, 1632, 1395, and 1042 cm−1, indicating the presence of hydrophilic functional groups such as C = O and C–O–C on the surface or edge of the prepared GO. It not only facilitates the stable existence of the composite material in water, but also enriches the analyte. In the IR spectrum of Fe3O4/GO/Au composite material, the stretching vibration of −C = O in the amide bond (−CONH) is at 1628 cm−1, and the stretching vibration peak of C-N bond in the amide bond (−CONH) is at 1403 cm−1. In addition, the absorption peak at 596cm−1 can be attributed to the stretching vibration peak of Fe–O, indicating the successful loading of Fe3O4 nanoparticles on GO (Khan et al. 2022).

Fig. 2
figure 2

Infrared spectra of the GO, Fe3O4, and Fe3O4/GO/Au

Full size image

Au, GO, and Fe3O4/GO/Au NPs were confirmed by UV. From Fig. 3, it can be seen that GO exhibits a distinct characteristic absorption peak at 230 nm, corresponding to the π → π transition of conjugated C = C on GO, indicating the successful preparation of graphene oxide using the Hummers method. The Au sol exhibits a significant absorption peak at around 523 nm (Jing et al. 2016). Compared with the obtained GO, the absorption peak of GO in Fe3O4/GO/Au showed a slight red shift. Maybe aggregation has occurred after interaction for the whole substrate. The Au of UV in Fe3O4/GO/Au was a little blue shift. The blue shift of 523 can be attributed to the localization of Au NPs on GO net (substrate), preventing Au NPs aggregation. This aggregation of the whole substrate may generate more SERS hotspots, which are crucial for the electromagnetic enhancement of SERS signals. In one word, the three-layer structure of the composite material has been successfully designed.

Fig. 3
figure 3

Ultraviolet–visible absorption spectra of the GO, Fe3O4, and Fe3O4/GO/Au

Full size image

In order to better analyze the composite of GO and Au NPs, the comparison of the Raman spectrum for GO and Fe3O4/GO/Au was shown in Fig. 4. The peak intensity of Fe3O4/GO/Au is obviously higher than GO. It demonstrated that Au NPs have a high SERS enhancement effect for GO. The two groups of Raman spectra can clearly display the characteristic peaks of GO, that is, the D peak (close to 1350cm−1) corresponds to the defective carbon bond structure, which is related to the vibration of sp3 carbon atoms. The G peak located at 1583cm−1 corresponds to the E2g vibrational mode structure of graphite SP2 carbon atom phonons (Tene et al. 2023). The Raman signal of Fe3O4/GO/Au nanocomposites is stronger than that of graphene oxide materials, indicating that the prepared Au NPs have a strong surface-enhanced Raman spectroscopic effect based on SPR.

Fig.4
figure 4

Raman spectra of GO and Fe3O4/GO/Au

Full size image

TEM image of Fe3O4/GO/Au NPs is presented in Fig. 5a. As seen in Fig. 5, the prepared Fe3O4 NPs are dark gray with agglomeration. The black dots are Au with a particle size of less than 50 nm and uniform distribution. The slightly wrinkled thin layer is GO. It is shown that Au and Fe3O4 NPs are effectively adsorbed on the GO. It is confirmed that Fe3O4/GO/Au is successfully prepared (Liu et al. 2022a, b). The EDSX of the Fe3O4/GO/Au NPs is presented in Fig. 5b. The Fe, O, and Au elements were accurately identified.

Fig. 5
figure 5

TEM imaging of Fe3O4/GO/Au microspheres (a) and EDX (b)

Full size image

SERS performance

10−6 M of R6G was selected as a Raman probe. From Fig. 6, it can be seen that the Raman signal of R6G with Au NPs as substrate was weak. However, SERS spectra of R6G with a high signal-to-noise ratio were obtained with Fe3O4/GO/Au nanocomposite material as the substrate. Among them, the C = C double bond stretching vibration characteristic spectra (1312, 1345, and 1566 cm−1) and in-plane and out-of-plane deformation vibration characteristic spectra (610, 774, and 1182 cm−1) related to the benzene ring of R6G molecule were significantly enhanced (Seo and Ha 2018). The above results demonstrate that Fe3O4/GO/Au composite materials enhance the SERS of R6G. It may be attributed to the adsorbability of graphene and the SPR of Au NPs so that it reduced the distance between the substrate and the probe molecules. As a result, the SERS signal of R6G improved remarkably. Therefore, the prepared Fe3O4/GO/Au composite material can serve as an active SERS substrate for rapid detection.

Fig. 6
figure 6

The Raman spectra of R6G from Au and Fe3O4/GO/Au

Full size image

SERS detection of mancozeb

The obtained Fe3O4/GO/Au composite material was used as SERS active substrate, Raman spectra of different concentrations of mancozeb in pyridine solution were collected, and the spectra of pyridine were given as control. Figure 7 shows the average spectrum of mancozeb in the range of 0.125–25 ppm (0.125 ppm, 0.25 ppm, 1.25 ppm, 2.5 ppm, 12.5 ppm, and 25 ppm), and all spectra are smoothed. The strong band at 665 cm−1 can be attributed to the interaction between Zn and Mn and C-S–S groups, the strong band at 953 cm−1 may correspond to the stretching vibration of C-S. The weak band at 849 cm−1 can be attributed to the bending vibration of N–H. The vibration at 1185 cm−1 can be attributed to the coupling of CH3 rotational vibration and C-N stretching vibration, and the detected spectral band at 1320 cm−1 may be caused by the coupling of N–H and C–C stretching vibration (Tsen et al. 2021).

Fig. 7
figure 7

Raman spectra of different concentrations of mancozeb

Full size image

In order to quantitatively analyze mancozeb, we plotted a working curve using the intensity of mancozeb at 1320 cm−1with different concentrations. It can be seen that the intensity of SERS increases with the increase of mancozeb concentration. As shown in Fig. 8, the standard curve showed perfect linearity in the range of 0.25 ~ 25 ppm (0.25 ppm, 1.25 ppm, 2.5 ppm, 12.5 ppm, and 25 ppm), with the correlation coefficient R2 of 0.9927, and the optimal fitting equation was Y = 1057.7X + 7954. These results indicate that Fe3O4/GO/Au composite materials as SERS active substrates can meet the requirements of quantitative detection.

Fig. 8
figure 8

The corresponding calibration plot of Raman intensity versus the concentrations of mancozeb

Full size image

Determination of mancozeb in pakchoi

In order to evaluate the feasibility of the developed method for application, the real sample as a representative, pakchoi purchased from the supermarket was detected. To minimize the influence of the pakchoi matrix, the QuEchERS method was employed for extraction. The results proved that the mancozeb was undetected in pakchoi. Subsequently, three different concentrations of mancozeb were artificially added to the unprocessed pakchoi for recovery. As shown in Table 1, the developed SERS method exhibited recovery ranging from 82.0 to 102% for detecting mancozeb in pakchoi and water. The recovery of water is better than pakchoi. Maybe it is attributed to the loss of mancozeb in the process of extraction.

Table 1 Results of recovery experiments for detecting mancozeb
Full size table

Activity, stability, and repeatability of Fe3O4/GO/Au

In order to quantify the SERS ability, the analytical enhancement factor (EF) was estimated, using the following equation: EF = ISERS × C0/I0 × CSERS, where ISERS and I0 are the intensity of SERS and the normal Raman (Bell et al 2020). The CSERS and C0 are the concentrations of mancozeb absorbed at Fe3O4/GO/Au (0.125 ppm) and on bare mancozeb (25 ppm). The EF for the mancozeb at 1320 cm−1 was 3.28 × 104. This may be attributed to the contribution of two different enhancement mechanisms, namely, electromagnetic enhancement and chemical enhancement. The formulation (LOD = 3δ/k) was used to determine the limit of detection (LOD), where δ represents the standard deviation of intensities from the blank platform and k is the slope of the calibration curve. At peak 1320 cm−1, the calculated LOD was 0.077 ppm. The aforementioned SERS experiment confirms that Fe3O4/GO/Au has a greater capacity to identify mancozeb. The comparison of different materials for mancozeb detection is listed in Table 2.

Table 2 The comparison of different materials for mancozeb detection
Full size table

Stability is a critical feature for any outstanding SERS-active substrate in practical application. To examine the stability of Fe3O4/GO/Au, time-dependent SERS measurements were performed, as shown in Fig. 9. It showed that the intensity of the SERS signal did not obviously change over 25 days. It indicated that the Fe3O4/GO/Au substrate is stable during preservation.

Fig. 9
figure 9

SERS spectra of 25 ppm mancozeb on storage samples with various periods

Full size image

In addition, the repeatability of Raman signals is an important factor for the performance of SERS substrate. To assess the repeatability of Fe3O4/GO/Au, three samples were obtained following the same conditions, at 11 random locations on the substrate. The results clearly show that the Fe3O4/GO/Au as SERS substrate has high repeatability, with a low RSD of approximately 4.4% (Fig. 10).

Fig. 10
figure 10

a SERS spectra of mancozeb from 11 randomly selected substrates. b The Raman intensity distribution of 1320 cm−1 peak from the obtained 11 spectra

Full size image

Mechanism exploration

Generally, it contains an electromagnetic enhancement mechanism (EM) and chemical reinforcement mechanism (CM) for SERS(Jensen et al 2008; Cialla et al 2012; Valley et al 2010; Morton, et al 2009). For our research, the Au in the Fe3O4/GO/Au substrate plays a significant role in enhancement because of LSPR and the coordination between Au and mancozeb molecular with sulfhydryl and amino group. Due to the coordination interaction, the distance between the substrate and target molecule gets shorter, in the region of the hot spot or junction. In addition, because of the adsorption property of GO, maybe it adsorbs target molecules and shortens the distance between them. Furthermore, maybe there is a coupling effect between transition metal (Fe) and noble metal nanostructures (Au) which improve the reflectivity of incident light. As a result, the local electromagnetic field intensity near the noble metal nanostructure strengthened. Furthermore, the magnetic properties of the Fe3O4 create a high density of interparticle hot spots by magnetism-induced aggregation which endow the Fe3O4 with enrichment ability. Due to the abovementioned reason, the substrate and target molecule get closer and closer. As a result, the electromagnetic field will be enhanced in the region of the hot spot, and the Raman signal will be amplified.

Chemical reinforcement mechanism mainly includes ground state chemical enhancement mechanism, resonance Raman enhancement mechanism, and charge transfer (CT) enhancement mechanism (Zhang, et al 2022). The mechanism of CT enhancement means that when the target molecules are adsorbed on the surface of precious metal nanoparticles, chemical bonds are formed. When incident light irradiates on the surface of noble metal nanoparticles and the energy of incident light match that of the charge transfer, also if the Fermi energy level of the metal surface is between the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) of the adsorbed molecule, resonance Raman scattering occurs. The free electrons on the surface of the metal will resonate and transit to the adsorbed molecule. The process of resonant electron transition will increase the polarizability of adsorbed molecules. Thus, the Raman signal of adsorbed molecules is enhanced. Meanwhile, due to the tunable band gap, semiconductors have attracted researchers’ attention in the SERS application. Because of the distinctive charge transfer capability, semiconductors have the possibility to match the energy levels of target molecules with large SERS enhancement (Zhang et al 2021). The CT route was found to be the excitation of electrons from the conduction bands of the semiconductor to the LUMO of the molecule. For our research, maybe the free electrons resonate and transit from the conduction band of GO and excited states of Au NPs to LUMO mancozeb, as shown in Fig. 11. Initially, the energy gap between the HOMO and LUMO can witness the direct resonant Raman scattering by laser. The excess of electrons can be contributed from GO and Au NPs, which endow the Raman signals further enhancement by the EM and CM (Ye et al 2012; Mandavkar et al 2022; Rajput et al. 2022).

Fig. 11
figure 11

Schematic of CM and EM enhancements in the SERS process

Full size image

Conclusion

In the study, a three-layer multifunctional Fe3O4/GO/Au composite nanomaterial with a high hot spot was designed and prepared as a reinforced matrix. This composite material enabled the rapid, sensitive, and quantitative detection of mancozeb using the SERS method. The magnetic precious metal composite nanomaterials were characterized and analyzed. The Fe3O4/GO/Au and Au NPs were experimentally investigated using R6G as the probing molecule. The results demonstrated that the prepared composite nanomaterials exhibited excellent performance than single Au NPs for SERS assay. Moreover, the Fe3O4/GO/Au composite was successfully applied to detect mancozeb, showing linear correlations ranging from 0.25 to 25 ppm at 1320 cm−1. The linear relationship was represented by the equation Y = 1057.7X + 7954, with a coefficient of determination (R2) of 0.9927. When applied to actual samples, the method exhibited acceptable recoveries, indicating excellent accuracy and applicability for detecting mancozeb in pakchoi. The proposed method showed well sensitivity to mancozeb with a limit of detection (0.077 ppm) and high EF, stability, and repeatability. In addition, the mechanism of SERS enhancement with EM and CM was discussed in detail, which provides a new perspective in understanding SERS mechanisms. The main advantages of this method are short sample pretreatment time with multi-function and strong signal amplification. It is expected to achieve quantitative detection of pesticide residues in the fields of food, agriculture, and the environment.