As the CL reagent, luminol concentration was an important factor which affected the CL intensity. The effect of luminol concentration on the CL intensity was examined in the range of 1.0 × 10^-5 mol/L-1.0 × 10^-3 mol/L, with 3 replicates. ΔI (the increasing amount of CL intensity) reached the maximum value when the concentration of luminol solution was 5.0 × 10^-4mol/L. Therefore, for the experiments, the concentration of luminol was maintained at the optimal value of 5.0 × 10^-4mol/L.

The effect of H₂O₂ concentration on the CL intensity was tested in the range of 0.05 mol/L-1 mol/L, with 3 replicates. The result (Figure 1) showed that ΔI reached the maximum value when the concentration of H₂O₂ was 0.05 mol/L. Therefore 0.05 mol/L was chosen as the optimal H₂O₂ concentration for further experiment.

The pH of the reaction of quinalphos with luminol-H₂O₂ in alkaline medium is an important parameter. The effects of a medium with pH in the range 12.5-13.5 were investigated, with 3 replicates. ΔI reached the maximum value when the pH was 13. Therefore pH 13 of luminol solution was selected as optimum for consequent research [²²].

Our preliminary experiment showed that the CL reaction of luminol-H₂O₂- organophosphorus could be sensitized by several surfactants and inorganic salts. Surfactants can form micelles which can change the physical characteristics of molecular guests and their reactivity in aqueous solutions. Thus, two non-ionic surfactants (Tween-80, Polythylene Glucol-400(PEG-400)), two anionic surfactants (sodium dodecyl sulplhate (SDS), sodium dodecyl benzene sulplhate(SDBS)), one cationic surfactants (Cetyltrimethylammonium bromide (CTMAB)) and five inorganic salts (KBr, NaBr, NaCl, KI, KCl) were tested. However, the results (Table 1) showed that the effect of enhancers were not obvious under strong alkaline condition and the CL intensity was even inhibited by some surfactants. It indicated that quinalphos can apparently enhance the CL intensity of the luminol-H₂O₂ system without sensitizers.

The effect of some common inorganic ions and organic compounds on the CL reaction was tested by analyzing a standard solution of quinalphos (1.0 × 10^-7 g/mL) to which increasing amounts of foreign ions were added. The tolerance limit was taken as the amount which caused a relative error of ± 5% in the peak height. Because some metal ions such as Cu (II), Co (II) may interfere the CL system in the luminol-hydrogen peroxide method. The tolerable ratio for foreign species was 1000-fold for glucose, CO₃^2-, NO₃^-, Cl^-, Na⁺; 100-fold for Mg²⁺; 1-fold for Pb²⁺; 0.01-fold for Co²⁺, Fe³⁺, Cu²⁺[^23-26]. It can be seen that obvious interference was caused by Co²⁺, Fe³⁺ and Cu²⁺. Thus, 1.0 × 10^-3 mol/L of EDTA was added to the luminol solution to restrain the interference from metal ions.

Under the optimal conditions, the calibration curve of ΔI against quinalphos concentration was linear in the range of 2 × 10^-8g/mL-1 × 10^-6g/mL and the calibration curves was y = 1081 × -2650.4 (where × is the concentration of quinalphos, 10^-8g/mL), with the correlation coefficient of 0.9921. The detection limit of quinalphos was 0.0055 μg/mL, calculated from the International Union of Pure and Applied Chemistry (IUPAC) recommendations (3σ). The relative standard deviations (RSD) for 7 injections with 1 × 10^-7g/mL quinalphos was 5.6%.

The proposed method was applied to the assay of quinalphos in vegetable sample. The eluent of quinalphos which was sprayed on the surface of vegetables was analyzed. According to the experimental data, the average recoveries for quinalphos in cherry tomato and green pepper 97.20% and 90.13% (Table 2). The relative standard deviations (RSD) for recoveries of quinalphos on the surface of cherry tomato and green pepper were 5.6% and 4.9%.

In this experiment, the kinetic characteristics of the proposed CL reaction were studied. The response curve of luminol (5 × 10⁻⁴ mol/L) CL reaction in the present quinalphos (1 × 10⁻⁷ g/mL) was recorded to study the kinetic characteristic of the CL reaction. Figure 2 demonstrated that the CL intensity peak appeared within 1 s of the injection of the quinalphos. The CL signals would decrease to a very low level within 20 s. The kinetic curve indicated the CL method was rapid and sensitive enough to be suitable for the detection of quinalphos.

Many research works have focused on the CL reaction mechanism of the luminol system. The research results have confirmed that the 3-aminophthalate anion was the emitter irrespective of medium and oxidant used. The maximum emission wavelength of luminol was about 420 nm. The resulting fluorescence spectrum (Figure 3) indicated that the maximum emssion wavelength of three systems (Luminol, Lumino-H₂O₂ and quinalphos-Lumino-H₂O₂) was also near 420 nm and the signal of fluorescence emission spectrum of quinalphos-lumino-H₂O₂ system is higher than lumino-H₂O₂ system. This means that the emitter of quinalphos-lumino-H₂O₂ system was also 3-aminophthalate anion.

Based on the analysis of the fluorescence spectrum combined with the kinetic curve, the possible mechanism of the present reaction was proposed as following: [¹] Quinalphos was oxidized by H₂O₂ to peroxophosphonate; [²] Peroxophosphonate oxidized luminol to produce an excited state 3-aminophthalate anion; [³] The excited state 3-aminophthalate anion relaxed to the ground state, producing the observed emission [²⁷,²⁸] (Figure 4).

All reagents were analytical reagent grade and distilled water was used throughout these experiments. Luminol (5-amino-2,3-dihydro-1, 4-phthalazinedione, 99%) was purchased from the Corporation of Sigma(USA). Quinalphos standard solution (0.1 mg/mL in acetone) was purchased from the Corporation of Standard Material (Beijing, China). Tween-80(Amresco), PEG-400(Guangdong, China), SDS(Guangdong, China), SDBS(Shanghai, China), CTMAB(Wuhan, China), KBr(Guangdong, China), NaBr(Tianjin, China), NaCl(Jiaozuo, China), KI(Wuhan, China), KCl(Wuhan, China) were analytical reagent grade.

The 0.01 mol/L luminol stock solution was prepared by dissolving 0.1772 g luminol with 2 mL 1 mol/L NaOH solution and diluted with distilled water to 100 ml. The luminol solution was stable for at least 1 week when stored in refrigerator at 4°C. Working standard solutions of luminol were freshly prepared from the stock solution by appropriate dilutions before use and adjusted its pH with 0.1 mol/L NaOH. The 0.01 mg/mL quinalphos stock solution was prepared by diluting 0.1 mg/mL standard solution of quinalphos with 1 mL acetone and then diluting to 10 mL with distilled water. The obtained stock solution was stored in refrigerator at 4°C [^29-31].

The CL signal was measured by a BPCL ultra-weak luminescence analyzer (Institute of Biophysics, Chinese Academy of Science, Beijing, China) as shown in Figure 5. The CL intensity, amplified by a sensitive photomultiplier tube (PMT) operated at -400 V, was measured with a detector under the control of a computer. The CL intensity ΔI was calculated by ΔI = I_s −I₀, where I_s and I₀ are the CL signals in presence and absence of quinalphos respectively. The determination method was based on the relationship between ΔI and corresponding concentration of quinalphos. The fluorescence spectrum was obtained by RF-5301PC fluorescence recording spectrophotometer (Shimadzu, Japan).

The proposed method was utilized for the determination of quinalphos in vegetable sample. Cherry tomato and green pepper were bought from the market in our campus. The weight of cherry tomato was about 15 g. The weight of green pepper was about 10 g. They were cleaned with distilled water. In order to perform the recovery test, a known amount of quinalphos standard solution was added to the sample and then washed with distilled water. The washings were diluted to suitable concentration with distilled water for analysis.