Silver nitrate (AgNO₃) was purchased from Sigma-Aldrich Chemical Co., Ltd. Glyphosate, sodium citrate (Na₃C₆H₅O₇· 2H₂O), sodium molybdate (Na₂MoO₄) and ninhydrin were obtained from Aladdin Industrial Corporation. Tap water was collected without further clean-up for real sample determination.

UV-Vis extinction spectra were measured with a lambda1050+ spectrophotometer (Perkinelmer).

Ag NPs were prepared as follows, 36.0 mg AgNO₃ was added into 200.0 mL H₂O first, and using sodium citrate (1% ω /V, 4 mL) as a reducing agent. A grey-green colloid formed and was naturally cooled at room temperature, after heating at 85 ℃ for 40 min.

Ninhydrin working reagent: Nninhydrin solution (5%, ω /V)+water+acetate buffer (0.4 mol· L^-1, pH 5.5) (2:1:1, V/V/V). Mixed working solutions: Ninhydrin working reagent+5% Na₂MoO₄+sample/glyphosate solution (1:1:1, V/V/V) were heated in boiling water for 30 min; and mixed with Ag NPs (1:1, V/V).

Water samples were only clean-up by a 0.45 μ m microporous ultrafiltration membrane, and then added different concentrations of glyphosate.

All geometries were optimized using the B3LYP exchange-correlation function. The 6-311++G(d, p) basis set was used for the H, C, O, N, P atoms, and LanL2DZ level for Ag NPs. All calculations were carried out using the Gaussian 09 software program.

The Ag NPs used in this study have a maximum absorption of around 440 nm (Fig.1a). The PD product displays a maximum absorption of around 570 nm (Fig.1d). When the PD absorbed on Ag NPs, there are slight shifts from 570 nm in the UV-Vis spectrum to 568 nm with a stronger peak (Fig.1b). This should originate from the interaction between the Ag NPs and the PD product molecules.

Fig.1

	Figure Option ViewDownloadNew Window
	Fig.1 UV-Vis spectrum of Ag NPs (a), the PD product+Ag NPs (b), ninhydrin+Na2MoO4+Ag NPs (c), PD product (d), and ninhydrin+Na2MoO4 (e) (glyphosate concentration of 1.0× 10-4 mol· L-1)

The optimized geometry of the PD product [Fig.2(a)], its Ag [Fig.2(b)] and Ag3 [Fig.2(c)] complexes were geometry-optimized.

Fig.2

	Figure Option ViewDownloadNew Window
	Fig.2 Molecular structures and electrostatic potential (MEP) mapping of purple color dye (PD) product and its Ag complexes (a): PD product; (b): Ag-PD product; (c): Ag3-PD product; (d): MEP mapping

The PD product includes two moieties, a ninhydrin molecule and a glyphosate molecule, connected by a C— N group [Fig.2(a)]. All the atoms in the ninhydrin part lie in the same plane and include the N16 atom with a large π -conjugated system. Thus, the PD product displays a maximum absorption peak at 570 nm due to this chromophore and auxochrome. Negative charges (electrophilic regions) are represented in red, while positive charges (nucleophilic regions) are represented in green; the MEP increases in the order of red< orange< yellow< green< blue [Fig.2(d)]. The electron density of the ninhydrin C=O group is higher than that of the other atoms, the color is green. These results suggested that this atom preferentially interacts with Ag and, thus, that the PD product would attach to Ag via an Ag— O bond. Significantly, the color of the C— N group is blue, which suggests that C— N coupled with the C=O group are the chromophore in the PD product.

Representative concentration-dependent UV-Vis spectra of the PD products were shown in Fig.3(a). The absorbance intensity at 568 nm shows a good linear relationship with the concentrations of PD product in the range of 1.00× 10^-9~1.00× 10^-4 mol· L^-1 (y=1.000 3+0.053 23× lgc, R²=0.939 5) [Fig.3(b)]. The LOD is thus calculated to be 2.017 4× 10^-11 mol· L^-1.

Fig.3

	Figure Option ViewDownloadNew Window
	Fig.3 UV-Vis spectra of the PD products at different glyphosate concentrations (a) and representative concentration-dependentof the PD products (b) (A, a): 1.00× 10-4; (B, b): 1.00× 10-5; (C, c): 1.00× 10-6; (D, d): 1.00× 10-7; (E, e): 1.00× 10-8; (F, f): 1.00× 10-9 mol· L-1 PD product; (G): AgNPs+Ninhydrin+Na2MoO4; (g): Ninhydrin+Na2MoO4

Recovery experiments were performed in each batch analysis to validate the method by spiking samples in five replicates at 1.00× 10^-8, 1.00× 10^-7, 1.00× 10^-6 and 1.00× 10^-5 mol· L^-1. The recoveries of water samples ranged from 71.26%± 1.63% to 114.68%± 11.30% without further clean-up. After further clean-up process, the recoveries ranged from 88.76%± 4.81% to 98.51%± 4.16%. It should be some unknown molecules in tap water that interfered with the adsorption of the target molecules on Ag NPs.

Table 1

Table 1

Table 1 Recoveries of glyphosate spiked into tap water Treatment Glyphosate taken /(mol· L-1) Recovery rate /% Water samples without further clean-up 1.00× 10-8 71.26± 1.63 1.00× 10-7 113.36± 6.27 1.00× 10-6 88.24± 5.31 1.00× 10-5 114.68± 11.30 Water samples with clean-up 1.00× 10-8 88.76± 4.81 1.00× 10-7 93.25± 3.47 1.00× 10-6 95.13± 3.82 1.00× 10-5 98.51± 4.16

Table 1 Recoveries of glyphosate spiked into tap water

Influence of common cations and anions, such as Na⁺, K⁺, Mg²⁺, Ca²⁺, Cl^-, S O42-, and N O3-in pure water samples containing 1.00× 10^-6 mol· L^-1 glyphosate were investigated (Table 2). In cation, the tolerance limit of Ca²⁺ or Mg²⁺ was 10.00× 10^-6 mol· L^-1, much lower than that of Na⁺. The tolerance limit of S O42-was 20.00× 10^-6 mol· L^-1 coupled with Na⁺ or K⁺. However, all the tolerance limits are not 10 times lower than glyphosate concentration (1.00× 10^-6 mol· L^-1). The feasibility of the method in the application to water samples is discussed.

Table 2

Table 2

Table 2 Influence of coexisting substances on glyphosate detection in water samples Coexisting substances Tolerance limit/(10-6 mol· L-1) Ca2+, Cl- 10.00 Na+, Cl- 100.00 Na+, S O42- 20.00 Na+, N O3- 100.00 K+, S O42- 20.00 Mg2+, S O42- 10.00 K+, N O3- 50.00

Table 2 Influence of coexisting substances on glyphosate detection in water samples