Surface-Enhanced Raman Scattering Imaging Assisted by Machine Learning Analysis: Unveiling Pesticide Molecule Permeation in Crop Tissues

Abstract

Surface-enhanced Raman scattering (SERS) imaging technology faces significant technical bottlenecks in ensuring balanced spatial resolution, preventing image bias induced by substrate heterogeneity, accurate quantitative analysis, and substrate preparation that enhances Raman signal strength on a global scale. To systematically solve these problems, artificial intelligence techniques are applied to analyze the signals of pesticides based on 3D and dynamic SERS imaging. Utilizing perovskite/silver nanoparticles composites (CaTiO₃/Ag@BONPs) as enhanced substrates, enabling it not only to cleanse pesticide residues from the surface to pulp of fruits and vegetables, but also to investigate the penetration dynamics of an array of pesticides (chlorpyrifos, thiabendazole, thiram, and acetamiprid). The findings challenge existing paradigms, unveiling a previously unnoticed weakening process during pesticide invasion and revealing the surprising permeability of non-systemic pesticides. Of particular note is easy to overlook that the combined application of pesticides can inadvertently intensify their invasive capacity due to pesticide interactions. The innovative study delves into the realm of pesticide penetration, propelling a paradigm shift in the understanding of food safety. Meanwhile, this strategy provides strong support for the cutting-edge application of SERS imaging technology and also brings valuable reference and enlightenment for researchers in related fields.

Keywords: 3D and dynamic SERS imaging; artificial intelligence; perovskite/silver nanoparticles composites; pesticide penetration; surface‐enhanced Raman scattering imaging.

Figure 1

Schematic diagram illustrating the research on pesticide penetration behavior. A) The distribution of penetrated pesticides and sprayed CaTiO₃/Ag@BONPs on plant profiles. B) The schematic diagram of 3D SERS imaging. C) Verification of the ability of CaTiO₃/Ag@BONPs to detect, adsorb and remove pesticides, and interpretation of the SERS imaging process for pesticides in fruits and vegetables. D) The principle and process of artificial intelligence analysis methods, including vertex component analysis (VCA), multivariate curve resolution‐alternating least squares (MCR‐ALS), Euclidean distance (ED), and spatial machine learning (SML).

Figure 2

Characterization of substrate. A) The preparation process of CaTiO₃/Ag@BONPs. B) The electromagnetic field simulation of the Finite Difference Time Domain (FDTD) of thiram, the transmission electron microscope (TEM), and scanning electron microscope (SEM) by Ag@BONPs and CaTiO₃/Ag@BONPs. C) The dynamic light scattering (DLS) of Ag@BONPs and CaTiO₃/Ag@BONPs. D) SERS spectra obtained by 100 randomly selected groups of thiram using the current method. E) Stability verification of thiram spectra and intensity change histogram of 1380cm^‐1 characteristic peak within 30 days.

Figure 3

Demonstration of the ability of CaTiO₃/Ag@BONPs to detect and adsorb pesticides. A) Schematic representation of pericarp and pulp detection. B) SERS signals of chlorpyrifos (red line), thiabendazole (blue line), thiram (green line), and acetamiprid (yellow line) in pulp compared with SERS signals (grey line) on the pericarp. SERS signals of chlorpyrifos, thiabendazole, thiram, and acetamiprid were obtained on pericarp C) and in the pulp D) after washing with water (pink line) compared with CaTiO₃/Ag@BONPs (purple line). Red and blue colors indicate the lipophilic group and hydrophilic group, respectively. SERS imaging (black background) and corresponding VCA analysis (blue background) before and after CaTiO₃/Ag@BONPs washed chlorpyrifos (red), thiabendazole (blue), thiram (green), and acetamiprid (yellow) on pericarp E) and in the pulp F).

Figure 4

Pesticide penetration behavior and content analysis of pesticides in apples. A) Schematic representation of SERS imaging and data processing principles. B–E) Pesticide penetration SERS imaging (black background) detection and corresponding VCA analysis (blue background) of chlorpyrifos (red), thiabendazole (blue), thiram (green), and acetamiprid (yellow) at 2, 4, 6, 8, 10, 12, 24, 48, 72, and 96 h. F–I) The heat maps of penetration content of chlorpyrifos, thiabendazole, thiram, and acetamiprid were analyzed by MCR‐ALS at 2, 4, 6, 8, 10, 12, 24, 48, 72, and 96 h. J) Representative SERS spectra. The center lines represent a mean of ten spectra from one apple pulp sample to show general signal patterns. The light‐colored area represents the standard deviation range. K) The change of penetration content of chlorpyrifos, thiabendazole, thiram, and acetamiprid was obtained at 2, 4, 6, 8, 10, 12, 24, 48, 72, and 96 h. Colored lines represent mean changes obtained from a B‐spline curve fit.

Figure 5

Pesticide penetration SERS imaging and related VCA and MCR‐ALS analyses results in various fruits and vegetables. A) Schematic representation of SERS imaging and data processing principles. SERS imaging (black background) detection and corresponding VCA analysis (blue background) of acetamiprid B) and chlorpyrifos C) at 2, 4, 6, 8, 10, 12, 24, 48, 72, and 96 h. D) Comparison of penetration content of non‐systemic pesticides and systemic pesticides in different fruits and vegetables (From top to bottom are oranges, pears, tomatoes, cucumbers, and grapes). Color lines represent mean changes obtained from a B‐spline curve fit.

Figure 6

Spatial machine learning for SERS imaging of pesticide penetration. A) The computational flow of machine learning. B) ROC curves of the training set of acetamiprid and chlorpyrifos. C) Confusion matrix for acetamiprid and chlorpyrifos test samples. Application of the established SML method. D) Trends in scores of different fruits and vegetables. E) Confusion matrix for acetamiprid and chlorpyrifos.

Figure 7

The penetration behavior of mixed pesticides in apples. A) Illustrates the visualization process of mixed pesticide infiltration. B) Displays pesticide penetration SERS imaging of chlorpyrifos (highlighted in red) and acetamiprid (highlighted in yellow) within 200 µm at 2, 4, 6, 8, 10, 12, 24, 48, 72, and 96 h. C) Shows the results of Euclidean distance (ED) analysis for the corresponding pesticides.