Handheld SERS coupled with QuEChERs for the sensitive analysis of multiple pesticides in basmati rice

Rice is a staple for over 60% of the global population, but production faces constraints from water scarcity, land-use change, climate change, pests and fungal diseases. High pesticide usage to maintain yields leads to residues in food and the environment, posing health risks. EU Rapid Alert System data show frequent MRL exceedances for acephate, carbendazim, thiamethoxam and tricyclazole in Basmati rice. Conventional LC/GC–MS methods are accurate but slow, costly and not portable. The study aims to develop a rapid, portable, handheld SERS screening platform using Au nanoparticles coupled with QuEChERs extraction for sensitive detection of multiple pesticide residues in Basmati rice, targeting sub-10 ppb detection in matrix and enabling on-site analysis.

SERS has been applied to detect food and environmental contaminants (mycotoxins, antibiotics, metals, alkaloids). Prior pesticide SERS works include: AuNP detection of carbendazim in tea; Au@Ag shell-thickness dependent enhancement for pesticide detection on fruit peels; GNR arrays for thiabendazole in apple; SERS imaging for thiophanate-methyl and carbendazim in peppers; and gecko-inspired nanotentacle substrates for simultaneous pesticide detection in fruit/vegetables. However, these are largely lab-based with benchtop systems. Existing on-site screening relies on enzymatic/colorimetric tests with limited sensitivity and multiplexing. For extraction, QuEChERs methods (original and buffered/acetate variants) are widely used in fruits/vegetables and have been adapted for grains/rice mostly for LC/GC–MS workflows, with limited integration into SERS. This work integrates handheld SERS with a QuEChERs acetate protocol to enhance sensitivity in rice matrix.

Substrate synthesis and characterization: AuNPs were synthesized via citrate reduction (Frens method) by varying citrate-to-gold ratios to tune particle size and SPR λmax (519, 528, 536, 587 nm; sizes ~25–82 nm by DLS). Citrate-stabilized AuNP surfaces facilitate ligand exchange and analyte adsorption. UV–vis assessed stability and aggregation; DLS measured size. SERS benchmarking with rhodamine 6G (R6G) showed best enhancement with AuNPs λmax=519 nm; analytical enhancement factor ~8.04×10^3 at 1362 cm−1. For pesticides, AuNPs λmax=528 nm yielded strongest SERS in presence of HCl (2 M), optimizing interparticle ‘hot spots’. Handheld/benchtop instrumentation: Handheld HRS-30 (Ocean Insight) 785 nm, ~400 mW (80%), 5 s integration with fiber probe; benchtop DXR2 Raman microscope (785 nm, 24 mW, 5 s, 10× objective). Spectra (400–1600 cm−1) averaged (n=3), smoothed (Savitzky–Golay) and analyzed in OriginPro 8.5. Handheld-SERS assay conditions (standards and extracts): In a vial, mix 1 mL AuNPs (λmax=528 nm, OD528=3.0) + 50 µL analyte + 5 µL HCl (2 M); invert; incubate 2 min at RT; acquire spectrum. Same protocol adapted to 96-well plates for microscope comparison. Pesticide standards: Stocks in ethanol or water; serial dilutions 0.001–10 ppm in water. Calibrations established from characteristic peaks: ACE (682 cm−1), CBM (1006 cm−1), THI (638 cm−1), TRI (1371 cm−1). LOD/LOQ by IUPAC (3S/M, 10S/M) using blank SD and slope. Rice spiking and extraction optimization: Indian Basmati rice cleaned, dried. Spiking: 5 g rice spiked with pesticide solutions, rolled 30 min, dried overnight. Four extraction approaches evaluated: E1 swab in ethanol; E2 solvent extraction (EtOH) of 1 g; E3 QuEChERs acetate (MeCN+1% acetic acid; salts 7 g MgSO4 + 1.8 g NaOAc); E4 original QuEChERs (MeCN; salts 4 g MgSO4 + 1 g NaCl). Due to low rice moisture, samples hydrated with 5 mL water; then 15 mL extraction solvent; vortex 1 min; salt addition; vortex 1 min; centrifuge 5 min at 1970×g; collect 10 mL supernatant. Sample size (5 g optimal) and hydration time (0–20 min tested; incubation not beneficial) were assessed. E3 selected for multi-residue extraction as it recovered ACE, CBM, THI and TRI in <15 min and buffered pH ~5.0–5.5 improved stability for labile analytes. Matrix analysis: Spiked rice ranges: ACE 0.5–10 ppm; CBM, THI, TRI 1 ppb–10 ppm. Background peaks in unspiked extracts (e.g., 760, 990, 1180, 1260, 1535 cm−1) guided selection of non-overlapping analytical peaks; CBM quantified at 633 cm−1 (instead of 1006 cm−1) to avoid 990 cm−1 background. Linearity, LOD, LOQ computed; handheld results compared with benchtop microscope. Validation and recovery: Recovery and RSD determined by spiking at three levels per analyte (e.g., ACE 1–10 ppm; CBM/THI/TRI 1 ppb–10 ppm). Release factor (RF) to assess matrix effects computed from SERS intensity in matrix vs solvent; RF 55–79%. REaccuracy and REprecision evaluated; handheld vs microscope agreement assessed. Multiplexing: Equal-concentration mixtures of ACE, CBM, THI, TRI analyzed in solvent and after QuEChERs acetate extraction from spiked rice. Key peaks: ACE ~680 cm−1; CBM ~730–735 cm−1; THI ~638–640 cm−1; TRI ~1318, 1372 cm−1. Concentrations studied down to 0.25 ppm (solvent) and 2.5 ppm (rice).

- Substrate selection: AuNPs with λmax=528 nm provided strongest SERS for pesticides when aggregated with HCl (2 M). UV–vis showed redshift/broadening (~650 nm) upon aggregation with pesticides; handheld SERS detected characteristic fingerprints only in presence of AuNPs + HCl. - Sensitivity in solvent: Linear ranges: ACE, CBM, THI from 0–1 ppm; TRI 0–0.1 ppm (R²=0.993–0.997). LODs: ACE 62 ppb; CBM 47 ppb; THI 75 ppb; TRI 5 ppb. Benchtop microscope LODs: 0.3–5 ppb (R²=0.963–0.997). Only TRI met EU MRL (10 ppb) in solvent on handheld. - Extraction optimization: Swab (E1) and simple solvent (E2) failed to recover TRI; QuEChERs methods succeeded. Original QuEChERs favored TRI; QuEChERs acetate favored ACE, CBM, THI. For multi-residue capability, QuEChERs acetate adopted (pH ~5–5.5; high ionic strength) enabling broad extraction in <15 min. Optimal sample size 5 g; extra hydration incubation not beneficial. - Sensitivity in rice matrix (handheld): Lowest visually detectable peaks: ACE 1 ppm; CBM 1 ppb; THI 1 ppb; TRI 1 ppb. Calibrations in extracts showed LOD range 0.6–800 ppb (R²=0.918–0.988); microscope gave similar 0.3–800 ppb (R²=0.937–0.996). CBM, THI, TRI detected below EU MRL (10 ppb); ACE not improved beyond ~800 ppb. - Mechanism of matrix sensitivity gain: QuEChERs acetate’s high ionic strength (~44% w/v salts) and low pH (~5.0) destabilize citrate-capped AuNPs, accelerating aggregation and hot-spot formation that trap trace pesticides, boosting SERS signals. - Recoveries/precision: Average recoveries 83.4–115.0%; RSDs 3.6–23.8%. REaccuracy −17% to 4.5%. Handheld vs microscope REprecision −0.14 to 0.33, indicating no significant difference. - Multiplexing: In 4-analyte mixtures, characteristic peaks visible at 0.25 ppm (solvent) and 2.5 ppm (rice). TRI dominated spectral region 1000–1600 cm−1 due to strong Au–S affinity, suppressing signals of ACE, CBM, THI; quantitative multiplexing limited by spectral overlap and competitive adsorption.

The study demonstrates that combining tailored AuNP substrates (λmax=528 nm) with controlled acid-induced aggregation and a QuEChERs acetate extraction enables handheld SERS to detect multiple pesticide residues in Basmati rice at low ppb levels. This addresses the need for rapid, portable, on-site screening tools where LC/GC–MS are impractical. In matrix, sensitivity improved versus solvent due to extraction medium–induced nanoparticle aggregation that enhances hot-spot density. The platform detected three of four target pesticides (carbendazim, thiamethoxam, tricyclazole) below the stringent EU MRL of 10 ppb and produced recoveries and precision suitable for screening. Agreement with benchtop Raman validates the handheld approach. However, selectivity and competitive adsorption affect analyte response, especially in mixtures where TRI dominates, highlighting the need for substrate engineering and data-driven spectral unmixing to realize robust multiplex quantitation.

A portable SERS platform coupled with QuEChERs acetate extraction enables sensitive, rapid detection of acephate, carbendazim, thiamethoxam and tricyclazole in Basmati rice. Key contributions include: (i) optimization of AuNP substrates and acid-induced aggregation for handheld operation; (ii) integration of QuEChERs acetate to both recover analytes and enhance SERS sensitivity in matrix; (iii) validation against a benchtop Raman system with comparable performance; and (iv) demonstration of preliminary multiplex capability. Future work should focus on substrate functionalization to enhance specificity and mitigate matrix interferences, strategies to stabilize or derivatize labile analytes like acephate, and application of machine learning (e.g., SMA, CNNs) to resolve spectral overlap and improve multiplex quantification. The approach shows strong potential as a tier-one on-site screening tool across diverse foods and environmental samples.

- Acephate sensitivity remained insufficient in matrix (not below 10 ppb; detectable only down to ~800 ppb), likely due to instability/degradation and extraction challenges. - Selectivity/competitive adsorption: Analyte–substrate interactions vary; sulfur-containing TRI binds strongly to Au, dominating signals and suppressing others in mixtures. - Spectral overlap and matrix background peaks complicate quantification (e.g., CBM peak near 990 cm−1 background required alternate peak selection). - Matrix effects and incomplete release: Release factor 55–79% indicates 21–45% of spiked analyte may not be released or detected due to matrix interferences. - Bare AuNPs are sensitive to ionic strength and pH; while leveraged here for enhancement, this may reduce robustness across varying matrices without careful control. - Multiplex quantitation is limited; further chemistry (substrate specificity) and computational unmixing are needed.