Introduction

Contamination by heavy metal ions, particularly Pb2+, poses a serious threat to the environment and human health. If blood lead levels exceed the provided standard (0.48 μmol L−1), mental development could be impaired, indigestion could occur, and renal functions could be compromised [1, 2].

To date, several methods have been developed for the determination of Pb2+, including high-performance liquid chromatography [3], flame atomic absorption spectrometry (AAS) [4], inductively coupled plasma-mass spectrometry [5], turbidimetry [6], and inductively coupled plasma atomic emission spectrometry [7]. However, these methods are generally time-consuming and expensive, and require further improvements. Because these defects limit the practical application of the existing methods, it is necessary to develop a rapid, economical, and sensitive method for the detection of Pb2+.

For nearly three decades, quantum dots (QDs) have been rapidly developing because of their unique advantages. Compared with traditional organic dyes, QDs have excellent optical properties, such as higher fluorescence intensity, photochemical stability, a tunable emission spectrum, and a narrower spectral line-width. QDs have attracted wide interest due to their various potential applications in drug screening, cell imaging, and biomarking [8–12], among others. Recent advances in QDs show great promise in molecular detection, including compounds with Cu2+ [13, 14], Au3+ [15], Ag+ [16], Co2+ [17], Hg2+ [18], Fe2+, and Fe3+ [19]. The use of QDs in food analysis is also an emerging field [20]. Unfortunately, few reports have been published regarding the detection of Pb2+ with QDs. For example, Wu et al. [21] detected Pb2+ using thiol-capped CdTe QDs with a detection limit of 2.7 × 10−7 mol L−1.

In this paper, water-soluble thioglycolic acid-capped CdTe QDs are used as fluorescent probe to detect Pb2+. The relationship between the QD size and the detection sensitivity of Pb2+ is studied in detail, and the mechanism of interaction between the five colors of CdTe QDs and Pb2+ is systematically explored. In addition, the influence of six other coexisting ions on the fluorescence signal of Pb2+ is also determined. A novel platform for selective detection by utilizing CdTe QDs as fluorescence probes is developed. This platform is based on the quenching fluorescence of CdTe QDs. The smallest particle, CdTe QDs-I, is chosen to detect Pb2+ in spinach and citrus leaves. Compared with Ref. [21], the use of our method increases Pb2+sensitivity by almost 180-fold.

Experimental

Apparatus and chemicals

Fluorescence measurements were performed using an RF-5301 spectrofluorometer (Shimadzu, Japan). UV absorption spectra were recorded with a 3100 UV-Vis spectrophotometer (Shimadzu, Japan). Transmission electron micrography (TEM) was performed on a JEM-2100 system operating at an acceleration voltage of 200 kV (Japan). X-ray powder diffraction (XRD) spectra were taken on a XRD-6000 system (Shimadzu, Japan). Graphite furnace atomic absorption spectra were obtained with a Wa Li-An AA220 instrument (China).

Doubly deionized water was used throughout the experiment. Thioglycolic acid was purchased from Shanghai Lingfeng Reagent Company (China), CdCl2·2.5H2O was obtained from Shanghai Jinshanting New Chemical Reagent Company (China), and NaBH4 and Te powder were acquired from Guoyao Chemical Reagent Company (China). The Pb2+ solution was prepared by dissolving 1.6594 g of Pb(NO3)2 in 50 mL doubly deionized water. Dilutions were made with doubly deionized water whenever necessary. A phosphate buffer solution (Na2HPO4-KH2PO4 0.067 mol L−1, pH 7.4) was used in the experiments.

The synthesis of multicolor CdTe QDs

CdTe QDs were synthesized following a previous method [17]. Briefly, 50.0 mg NaBH4, 3 mL anhydrous ethanol, 47.85 mg Te, and 1 mL doubly deionized water powder reacted in a small flask at 40 °C. Sodium hydrogen telluride (NaHTe) was produced. CdTe precursors were prepared by adding freshly prepared NaHTe solution to a N2-saturated CdCl2 solution at pH 8.0 in the presence of TGA as a stabilizing agent. The molar ratio of Cd2+:Te2-:TGA was 1:0.5:2.4. Further nucleation and growth of the QDs proceeded with refluxing at 100 °C under open air conditions with an attached condenser. Five fractions (QDs-I, QDs-II, QDs-III, QDs-IV, and QDs-V) were extracted at different times (10 min, 1 h, 4 h, 8 h, and 20 h, respectively) during reflux.

Procedures for fluorescence detection of Pb2+

A phosphate buffer solution (pH = 7.4) of about 800 μL and approximately 80 μL CdTe QDs-I were mixed thoroughly in a 10-mL volumetric flask. Two milliliters of the above solution were placed in a quartz cuvette, followed by a series of varying concentrations of Pb2+. Fluorescence spectra were measured after the sample had reacted for 5 min at room temperature to obtain stable fluorescence intensities. Fluorescence spectra were obtained by an RF-5301 spectrofluorometer at an excitation wavelength of 325 nm. The slit width of excitation was 3 nm and the emission slit was 3 nm. Spectra of the other QDs (II–V) were also obtained based on similar procedure for the detection of Pb2+.

Interference fluorescence quenching measurements

Doubly deionized water was used to prepare individual solutions of 0.5 μmol L−1 Ca2+, Na+, Mg2+, K+, Al3+, Ba2+, Zn2+, and Fe3+, and 0.1 μmol L−1 Hg2+, Ag+, and Cu2+. Dual ionic interference studies were conducted by preparing a 10−7 mol L−1 solution mixture of Pb2+ and Ca2+, Na+, Mg2+, Cu2+, Al3+, Ba2+, Zn2+, Hg2+, Ag+, Fe3+, and K+. Pb2+ was first introduced to the QDs, and then mixed with the various concentrations of ions. Characteristic spectra were obtained by the spectrofluorometer.

Detection of Pb2+ in spinach and citrus leaves

Spinach and citrus leaves were decomposed by dry digestion in a muffle furnace [22]. Solid samples were powdered and passed through 100 mesh sieves. These were then incinerated in porcelain crucibles until no smoke further emerged and calcined at 520 °C for 7–8 h in a muffle furnace. The ashes were dissolved using a 0.5 mol L−1 HNO3 solution. The pH was adjusted to 7.4 with 0.1 mol L−1 NaOH solution, and 50 μL of the resulting solution was added to a CdTe QDs-I solution. The Pb2+ concentration of the final solution was then determined using the Stern-Volmer equation.

Results and discussion

Characterization of multicolor CdTe QDs

Figure 1 shows the fluorescence and absorption spectra of the QDs. The fluorescence emission peak appeared to red-shift from 520 to 595 nm after refluxing from 10 min to 20 h, indicating that the sizes of the CdTe QDs gradually increased. This was mainly due to the Ostwald ripening process, in which smaller particles dissolve and larger particles grow. The UV absorption spectra showed obvious first excitonic absorption peaks. This was due to the quantum confinement effect. The shapes of the emission wavelengths were symmetrical and smooth. The full width at half-maximum was less than 50 nm, indicating that the CdTe QDs had homogeneous sizes.

Fig. 1
figure 1

Normalized emission fluorescence and absorption spectra of five different-sized CdTe QDs (QDs-I–V), λex = 325 nm

Full size image

The morphology of CdTe QDs was studied by TEM. Figure 2a shows the surface characteristics of QDs-II, which was obtained by refluxing the CdTe QD solution for 1 h. The shapes of the nanoparticles were close to spherical and monodispersed. Their average particle diameter was about 2.5 nm, which coincided with the diameter calculated by the empirical formula in Ref. [23].

$$ D = \left( {9.8127 \times {{10}^{{ - 7}}}} \right){\lambda^3} - \left( {1.7147 \times {{10}^{{ - 3}}}} \right){\lambda^2} + 1.0064\lambda - 194.84 $$

where D (nm) is the diameter of the CdTe QDs, and λ (nm) is the wavelength of the first excitonic absorption peak. Using the formula above, the mean diameters of QDs-I, QDs-II, QDs-III, QDs-IV, and QDs-V were found to be 1.6, 2.4, 2.9, 3.1, and 3.3 nm, respectively. A typical electron diffraction ring image for QDs-II is shown in Fig. 2b. This image shows that the CdTe QDs possessed a good crystalline structure.

Fig. 2
figure 2

a TEM image of CdTe QDs-II. b Electron diffraction ring of CdTe QDs-II

Full size image

Figure 3 shows the powder XRD pattern for CdTe QDs-II. The diffraction pattern of CdTe QDs was found to be relatively close to that of bulk cubic CdTe. The three diffraction peaks of the CdTe QDs at 23.6°, 40.2°, and 46.8° were respectively indexed to the (111), (220), and (311) planes of the cubic CdTe lattice.

Fig. 3
figure 3

XRD pattern of CdTe QDs-II

Full size image

The quantum yields of CdTe QDs were measured according to a previous method [24]. Rhodamin 6G, which was dissolved in anhydrous alcohol, was chosen as the standard sample. Experimental results showed that the quantum yields of CdTe QDs-I, -II, -III, -IV, and -V were 30%, 60%, 51%, 44%, and 32%, respectively.

The detection sensitivities of five colors of CdTe QDs towards Pb2+

Figure 4 shows the fluorescence spectra of QDs-I with increasing Pb2+ concentrations. The fluorescence of QDs-I was gradually quenched with increasing Pb2+ concentration. While the fluorescence emission peak of QDs-I shifted by 6 nm to a longer wavelength, the fluorescence emission peaks of QDs-II–V did not apparently red-shift with the continuous addition of Pb2+.

Fig. 4
figure 4

Effect of Pb2+ concentrations on the luminescence of CdTe QDs-I: A 0, B 9.9 × 10−10 mol L−1, C 1.96 × 10−9 mol L−1, D 3.38 × 10−9 mol L−1, E 5.66 × 10−9 mol L−1, F 7.41 × 10−9 mol L−1, G 1.30 × 10−8 mol L−1, and H 2.59 × 10−8 mol L−1

Full size image

The surface of the QDs played an important role in their fundamental properties. For smaller particles, the surface atoms are chemically more active because they have fewer adjacent coordinate atoms and unsaturated sites or more dangling bonds. Therefore, the imperfection of the particle surface induces additional electronic states in the band gap, which act as electron or hole trap centers [17]. The obvious red-shift in the luminescence band of QDs-I in response to Pb2+ is attributed to effective electron transfers from thioglycolic acid to Pb2+ and the formation of a new radiative center [25].

The change in absorbance peak with increasing concentrations of Pb2+ solution is shown in Electronic Supplementary Material Fig. S1. The shape of QDs-I became asymmetrical after the addition of Pb2+ solution, which can be explained by static quenching [21].

QDs-I had a detection limit of 4.7 nmol L−1, a sensitivity about 60 times greater than that of QDs-IV. QDs-II–V had proximate sensitivities, linear ranges, and detection limits for detecting Pb2+. The related data are shown in Table 1.

Table 1 Analytical parameters for the detection of Pb2+ using CdTe QDs-I–V
Full size table

The quenching efficiencies of five colors of CdTe QDs towards Pb2+

Pb2+ quenched the fluorescence intensity of CdTe QDs in a concentration-dependent manner. This could be best described by the Stern-Volmer equation:

$$ {{{{F_0}}} \left/ {{F = {K_{{SV}}}\left[ Q \right]}} \right.} + 1, $$

where F0 and F are the fluorescence intensity of CdTe QDs in the absence and presence of Pb2+, respectively, KSV is the Stern-Volmer quenching constant, which is related to the quenching efficiency, and [Q] is the concentration of Pb2+.

A good linear relationship between F0/F and [Q] can be seen from Electronic Supplementary Material Fig. S2. The KSV of QDs-I was almost 40 times as large as those of QDs-II–V, and its quenching efficiency was the highest among the five colors. These findings were consistent with the report of Xia [26].

In CdTe QDs, the band gap, particle size, and surface curvature are related and can influence quenching processes. When the particle size reaches a certain measurement, the dimension may become a major factor affecting quenching efficiency [17]. In this case, the quenching radius is better viewed as the action radius for the quencher as it approaches the surface of the QDs. For smaller QDs (1.6 nm), a quencher can reach across most of the nanoparticles, whereas for larger QDs (3.3 nm), the quencher, given its quenching radius of only 0.45 nm, can only interact with excitons near a limited region of the surface. This is illustrated in Fig. 5. The large QD size-dependence of the quenching radius was a surprising discovery. We believe that such quenching could be partially controlled by the surface characteristics of the QD, notably its curvature [27]. QDs-I had the smallest-sized particles; thus, its quenching efficiency, which was 9.74 × 107 mol−1 L, was the highest (Table 1).

Fig. 5
figure 5

Schematic representation of a molecule (black circle) positioned on the surface of red (3.3 nm) and green (1.6 nm) quantum dots. The grey circle represents the Perrin radii, interpreted as a sphere of action for the quencher

Full size image

The selectivities of five colors of CdTe QDs towards Pb2+

Figure 6 shows that the fluorescence intensity of CdTe QDs was insensitive to Ba2+, Hg2+, Ag+, and other physiologically important cations even if their concentrations were 500 times higher than that of Pb2+. Only Pb2+ could effectively quench the fluorescence of CdTe QDs. The distinct discrimination between Pb2+ and other ions makes it possible for CdTe QDs to be used for the analysis of Pb2+ in the presence of other ions [28]. The results showed that QDs-I had a higher selectivity towards Pb2+ and that the influences of other metal ions, except Cu2+, were much weaker. The presence of Cu2+ significantly influenced Pb2+ determination likely because of Cu+ formation, during which non-radiative recombination became easier among excited electrons in the conduction band and holes in the valence band, resulting in more powerful QD fluorescence quenching [29]. The selectivity of larger QDs was relatively low. Thus, QDs-I, which has the smallest size, is the most suitable for use as a probe for the detection of Pb2+.

Fig. 6
figure 6

Effect of metal ions on the fluorescence intensity of QDs-I–V in a phosphate buffer solution (pH = 7.4) (the concentrations of Cu2+, Hg2+ and Ag+ were 0.1 μmol L−1, other ion concentrations were 0.5 μmol L−1)

Full size image

Detection of Pb2+ in spinach and citrus leaves

QDs-I was chosen to detect Pb2+ in spinach and citrus leaves. The results are shown in Table 2. The contents of Pb2+ in spinach and citrus were found to be 14.43 and 11.44 mg/kg, respectively. The QDs-I method showed an accuracy comparable with that of AAS, as no significant difference was found between the two methods when p = 95% (t = 0.103 < t0.05,4 = 2.776). In addition, coexisting ions, such as Ca2+, Na+, Mg2+, Cu2+, Al3+, Ba2+, Zn2+, K+, Hg2+, Fe3+, and Ag+, yielded no interference within the allowable error range. Pb2+ recovery by QDs-I ranged from 92.2% to 107%, as shown in Table 3. Therefore, the QDs-method was accurate, sensitive, and selective for determining Pb2+.

Table 2 Comparison of AAS and QDs-I methods for the detection of Pb2+ in spinach and citrus leaves
Full size table
Table 3 Pb2+ concentrations in agricultural products as determined by AAS and QDs-I
Full size table

Conclusions

Multicolor water-soluble CdTe QDs modified by thioglycolic acid were synthesized. These CdTe QDs showed good distribution and displayed high fluorescence quantum yields. The experiment showed that the size of QDs influenced the detection sensitivity of Pb2+. With increasing size of QDs, the sensitivity decreased. QDs-I, which was the smallest particle, had the highest quenching efficiency, selectivity, and sensitivity in the detection of Pb2+; its detection limit was 4.7 nmol L−1, which is the lowest in the current study. The larger QDs had wider linear ranges of 1.96 × 10−9–3.33 × 10−7 mol L−1. A new method using QDs-I for detecting Pb2+ in spinach and citrus leaves was established. Pb2+ amounts of 14.43 and 11.44 mg/kg were found in spinach and citrus, respectively. The results are in good agreement with those obtained with AAS. In summary, TGA-capped water-soluble CdTe QDs are excellent fluorescent probes for the detection of Pb2+.