Sodium hypochlorite (NaClO), hydrogen tetrachloroaurate hydrate (HAuCl₄·xH₂O), and sodium hydroxide (NaOH) were bought from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Calcium chloride (CaCl₂), SA, glutathione (GSH), ATCh, and cobalt (II) chloride (CoCl₂) were bought from Aladdin Reagent Co. Ltd. (Shanghai, China). AChE was bought from Ryon Biological Technology Co. Ltd. (Shanghai, China). Tris-HCl buffer solution was obtained from Sigma Aldrich Reagent Co. Ltd. (St. Louis, MO, United States). Deionized water (DI water) with good resistance of 18 MΩ was used in this work. The detailed microstructures of the samples were analyzed with a JEM-7500 scanning electron microscope (SEM) (JEOL, Japan) and a JEM-2100 transmission electron microscope (TEM) (JEOL, Japan). The UV-Vis absorbance results are acquired by using a UV-2550 spectrometer (Shimadzu).

CoOOH NFs were synthesized in accordance with the previous work (Su et al., 2021): 1 ml of CoCl₂ (10 mmol L⁻¹) solution and 0.3 ml of NaOH (1.0 mol L⁻¹) were mixed together first. The mixture was sonicated for 1 min. Then, 50 μl of NaClO (0.9 mol L⁻¹) was introduced into the mixed solution and sonicated for 10 min. After that, 650 μl of deionized water was introduced. The CoOOH NFs were collected after centrifuging for 10 min at 10,000 rpm. Then, the products were washed three times with DI water.

AuNCs were prepared on the basis of the previously reported approach (Luo et al., 2012). The aqueous solutions of GSH (0.3 ml, 100 mmol/L) and 8.7 ml DI water (resistivity >18 MΩ cm) were successively added into HAuCl₄ (1.0 ml, 20 mmol/L). The mixture solutions were stirred for 5 min at 25°C. After that, the temperature was heated to 70°C and maintained for 24 h. The red-emission Au clusters were acquired. Then, the AuNC solutions were purified using a 1,000 Da dialysis bag for 24 h. The obtained AuNCs were placed at 4°C.

In a 2ml centrifuge tube, 1.565 ml of the SA (3.83 mg ml⁻¹) and 40 μl of AuNCs were mixed. When adding CaCl₂ (20 μl, 12.5 mg L⁻¹) to the aforementioned solution, the hydrogels were generated immediately.

Different concentrations of chlorpyrifos (25 µl) and 25 µl of AChE (0.5 U ml⁻¹) were mixed at 37°C for 20 min. Then, 100 µl Tris-HCl (pH 8.0, 10 mmol L⁻¹) and 50 µl ATCh (2 mmol L⁻¹) were added and reacted at 37°C for 25 min. Subsequently, 75 µl of CoOOH NFs (0.25 mg ml⁻¹) and 100 μl deionized water were added at room temperature for 10 min. After that, the mixed solution was added to the prepared AuNC-based hydrogel and equilibrated for 15 s. Finally, the cuvette was interposed into a portable device to catch the fluorescence image under the excitation of a 365-nm laser. The image of the hydrogel was analyzed by ImageJ software.

Real samples such as lake water, apple juice, and pear juice were prepared as analysis samples to confirm the practical applications of the fluorescent hydrogel sensing platform. The lake water sample (100 ml) was directly used. Apple juice and pear juice were obtained from apples and pears by using a juicer. The supernatant of the apple juice and pear juice was collected after being centrifuged twice at 10,000 rpm for 10 min. Then, the obtained supernatant was diluted 100-fold to measure the chlorpyrifos according to the mentioned process. For the recovery study, certain amounts of chlorpyrifos standard (0.625, 1.25, and 12.5 ng ml⁻¹) were spiked into the samples and then evaluated by the described procedure.

Red-emission AuNCs are simply synthesized by means of a classical reduction approach (Li et al., 2019), using GSH as a reductant and stabilizer. The morphology of AuNCs was characterized using TEM, demonstrating that the mean size is about 3.69 ± 0.75 nm (Figure 1A). The CoOOH NFs as a nanoquencher are prepared by using NaClO as an oxidizer under alkaline conditions, which is composed of CoO₆ octahedrals and possessed thickness at the nanoscale. As shown in the TEM image, two-dimensional CoOOH NFs possess hexagonal morphology with an average diameter of 78 ± 14.3 nm, which provided a large surface area for loading AuNCs (Figure 1B). The X-ray diffraction (XRD) pattern of the obtained CoOOH NFs was consistent with the hexagonal rhomb-centered phase (JCPDS NO. 07-0169), confirming the successful preparation of CoOOH NFs (Supplementary Figure S1). The fabricated CoOOH NFs were also confirmed by Förster transform infrared (FT-IR) spectra. Supplementary Figure S2 shows an absorption peak at 3,300, 1,652, and 540 cm⁻¹, relating to the −OH group, double bond (Co-O), and Co-O^2- complex vibration of CoOOH NFs, respectively. After introducing the AuNCs (−32.3 mV), the zeta potential of CoOOH NFs (16.8 mV) was changed to be −21.3 mV, forming a forceful electrostatic attraction (Figure 1C). Compared with CoOOH NFs (407 nm), the absorption peak at 393 nm of the CoOOH-AuNCs displayed a 14-nm blue shift, confirming the formation of a CoOOH-AuNC composite (Figure 1D). These results in accordance with the TEM image showed the modification of AuNCs on the CoOOH NF surface (Figures 1E, F). The energy dispersive spectroscopy (EDS) patterns further displayed the elemental composition of Co, O, Au, and S, revealing the combination of CoOOH NFs and AuNCs (Figure 1G). For comparison, the fluorescence emissions of AuNCs (black line) were prominently overlapped by the absorption band of CoOOH NFs (red line) in the range of 500–800 nm (Figure 1H). Furthermore, after adding CoOOH NFs, the fluorescence intensity of AuNCs was obviously quenched (Supplementary Figure S3), and the lifetime of AuNCs changed from 13.31 to 11.3 μs (Figure 1I and Supplementary Table S1). Based on the aforementioned experiments, it is supposed that the quenched principle is ascribed to the FRET effect.

Combining the fluorescence characteristics of AuNCs with AChE-controlled acetylthiocholine (ATCh) hydrolysis, the fluorometric platform was designed to detect OPs (Figure 2A). The enzyme hydrolysis products are TCh and CH3COOH. The testing system could generate 50 μmol L⁻¹ of CH3COOH if ATCh completed catalyzed hydrolysis by AChE. As shown in Supplementary Figure S4, the fluorescence intensity of AuNCs decreased by 2.74% with CH3COOH (50 μmol L⁻¹). During the detection process, we add 100 µl Tris-HCl (pH 8.0, 10 mmol L⁻¹) buffer to ensure the detection system is in a relatively stable pH environment. TCh could decompose the CoOOH NFs, controlling the quenching efficiency of CoOOH NFs toward AuNCs. Chlorpyrifos inhibits the catalytic activity of AChE, further causing fluorescence response in the system. As shown in Figure 2B, CoOOH NFs could easily quench the fluorescence signal of AuNCs (from the red line to the purple line). When the CoOOH NFs were reduced to Co²⁺ by the enzyme hydrolysis products (TCh), the fluorescence intensity of AuNCs was increased along with the decomposition of CoOOH NFs (black line). Furthermore, the fluorescence intensity of the system could be regulated on account of AChE-induced ATCh hydrolysis (blue line). Owing to the specific inhibition of chlorpyrifos and effective catalyzation toward the substrate, AChE is viewed as an antenna in the sensing system. Therefore, we successfully established the AuNC/CoOOH/AChE/ATCh sensing system to be applied in chlorpyrifos analysis. With the increase of chlorpyrifos, the fluorescence intensity at 630 nm gradually decreased (Figure 2C). As expected, an acceptable linear relationship (R² = 0.96) is obtained with the chlorpyrifos concentrations ranging from 0.625 to 125 ng ml⁻¹ (Figure 2D).

Benefiting from the simple preparation, superb biocompatibility, and 3D networks, hydrogels are selected as a robust carrier in fabricating stimuli-responded POCT platforms. SA, a promising natural polysaccharide, is made up of a β-D-mannuronic acid (M) unit and α-l-guluronic acid (G) unit. After introducing Ca²⁺, the “egg-box” structure is formed because of α-l-guluronic acid (G) unit stacking; thus, the hydrophilic hydrogel is acquired (Figure 3A). The SEM images of the freeze-dried SA hydrogel distinctly demonstrated the layer morphology of porous structures (Figure 3B), possessing a huge capability for nanomaterial anchoring. Considering the aforementioned characteristics, we have easily immobilized AuNCs into the SA hydrogel and constructed a TCh-controlled target response hydrogel test kit for the fluorometric analysis of chlorpyrifos.

The conventional detection methods usually consisted of bulky equipment, skilled person, and expensive cost, which could hardly meet the needs of on-site analysis for pesticides. Owing to its excellent capabilities of the digital camera program, good computation, and quick communication, the smartphone has drawn increased attention for developing real-time and portable sample-to-answer detection devices (Celik et al., 2021; Martinez-Avino et al., 2021; Zhu et al., 2021). Furthermore, the 3D printing technology is extensively applied in broad fields because of its low cost and capability to create objects with any predefined shape (Li and Pumera 2021). Hence, we explore a miniaturized handheld detection device that not only takes advantage of the aforementioned superiorities but also immobilizes the relative locations of the light source, sample, and smartphone camera. As shown in Figures 3C–E, the proposed platform is composed of a smartphone, a laser device (6.0 × 2.0 cm) of 365 nm, a four-channel cuvette with 1.0 cm inside diameter, a long-pass filter (1.0 × 1.0 cm), and a 3D-printed accessory (16.0 × 7.8 × 3.8 cm). When catching images, the excitation light is supplied by a 50-mW diode laser. To lower background noise, the long-pass filter with a resisting wavelength at 600 nm is settled between the smartphone camera and the detection sample. Thus, the POCT device is developed by combing the created AuNC/CoOOH/AChE/ATCh target response hydrogel test kit with the miniaturized homemade accessory for quantitatively tracing chlorpyrifos under good specificity and low background interference. Notably, the cost of the handheld POCT device was about $20. Using this handheld detection device, the concentrations of chlorpyrifos could be directly detected by analyzing the captured images (Figure 4A). As expected, an acceptable linear relationship (R² = 0.998) is obtained with chlorpyrifos concentrations ranging from 0.625 to 125 ng ml⁻¹ (Figure 4B). The detection limit is calculated to be 0.59 ng ml⁻¹ by the equation LOD = (3σ/k), where σ is the standard deviation of blank signals and k is the slope of the calibration curve. It is worth pointing out that the limit of detection (LOD) is 0.59 ng ml⁻¹ (S/N = 3), meeting the detection requirements of chlorpyrifos in agricultural products (the maximum residue limit of chlorpyrifos is 20 ng ml⁻¹ in fruits). The POCT device provided a low demand for sample concentration (0.625 ng ml⁻¹) and a satisfied LOD compared with the previously reported strategies (Supplementary Table S2).

A meaningful index to assess the performance of AuNC-based biosensors is the selectivity toward the target. The common interfering substances such as Na⁺, Ca²⁺, BSA (bovine serum albumin), GOX (glucose oxidase), GSH, Phe (phenylalanine), Met (methionine), Ala (alanine), and other pesticides (pyrazole pesticides, neonicotinoid pesticides, and pyrethroid pesticides) were selected to investigate the selectivity. As shown in Figure 4C, after adding chlorpyrifos or the common interfering substances into the hydrogel system, only chlorpyrifos (62.5 ng ml⁻¹) generated a prominent decrease in the fluorescence color. By transforming the color information, a histogram further displayed the normalized intensity of the gray value, exhibiting that the intensity was barely changed by interference substances (125 ng ml⁻¹). The good selectivity of the hydrogel test kit for chlorpyrifos detection might ascribe to the excellent specificity of AChE that holds a unique catalyzation effect toward the substrate and specific recognition capability to chlorpyrifos. The outstanding properties for detecting chlorpyrifos might be attributed to the following particular characteristics: 1) The AuNCs as a signal indicator with red-emission could conquer the influence of inevitable autofluorescence of the biological matrix, efficiently blocking background interference and enhancing anti-interference capability. 2) The CoOOH NFs possess high quenching efficiency toward AuNCs on the base of the FRET effect, remarkably improving the detection sensitivity. 3) Compared with the strict pH requirements of acid/alkaline phosphatase and the reversible inhibition of tyrosinase, AChE is the most broadly used enzyme in pesticide detection, which performed specificity recognition toward ATCh and a robust response toward pesticides. 4) SA hydrogels allow small molecules to diffuse through the porous structure and also act as a barrier to avoid the infiltration of biomacromolecules, which are favorable to the stability of the platform.

The practical ability of a hydrogel test kit for pesticide detection is meaningful to environmental assessment and food safety monitoring. To investigate the practical applications of the test kit, we selected lake water, apple juice, and pear juice as test samples. The precision and accuracy of the test kit for chlorpyrifos are investigated using the standard addition method. The recovery rates of chlorpyrifos (0.625, 1.25, and 12.5 ng ml⁻¹) were obtained according to the hydrogel test kit. In Supplementary Figure S5, the 25 µl 100-diluted apple juice and pear juice rarely affected the fluorescent signal of AuNCs. As shown in Table 1, the recoveries were determined in the range of 87.47–112.64%, with relative standard derivations (RSDs) lower than 4.43%. Overall, the hydrogel test kit possessed high reliability for monitoring OPs in a complex sample.