Introduction

Copper is widely used in agriculture and industry, and also plays an important role in the environment and in human physiology. High concentrations of Cu2+ may be toxic to the human body [1, 2] and lead to liver or kidney diseases [3]. For this reason, Cu2+ is listed as a priority pollutant by the Environmental Protection Agency (EPA) [4]. Therefore, the detection of the concentration of Cu2+ in environmental water samples has important practical value in the water pollution monitoring and management. Many analytical methods have been developed and utilized to detect Cu2+ including electrochemical method [5], atomic emission spectrometry [6], atomic absorption spectrometry [7], colorimetry [8,9,10,11], fluorometry [12] and so on. Among these methods, fluorometry has been proved to be an ideal technique for the trace analysis due to its advantages of easy operation and high sensitivity.

Various fluorometric methods for the detection of Cu2+ have been designed by utilizing fluorescence dyes or fluorescence nanomaterials as fluorescent probes [13,14,15]. However, many of them still suffer from some drawbacks, such as complicated synthetic procedures, poor water solubility and toxicity, which limit their further applications [16, 17]. On the other hand, most methods are based on the interaction between Cu2+ and the functional groups of fluorescence probes. However, some metal ions may influence the selectivity due to their similar chemical properties and ionic radii. For instance, utilizing the affinity effect between the amino group at the surface of carbon quantum dots and the Cu2+, Chi et al. developed a fluorescence method to detect Cu2+ [18]. However, some cations, such as Ni2+, Fe2+ and Hg2+, seriously disturb the detection of Cu2+ due to the combinations of Ni2+, Fe2+ and Hg2+ with amino groups. Based on the chelation of Cu2+ and the “N” and “O” atoms, Sun et al. developed a polymer nanodots-based fluorescence method to detect Cu2+. Although satisfactory results were obtained, the selectivity results indicated that Fe3+ influence the selectivity [19]. Furthermore, similar phenomena are also found in some of the reported literatures [4, 19,20,21]. Therefore, smart design should be introduced for the construction of Cu2+ detection method with better selectivity.

It has been demonstrated that Cu2+ can not only catalyze the cysteine (Cys) to cystine and produce H2O2 under the O2 atmosphere [22, 23], but also catalyze decomposition of as-produced H2O2 into hydroxyl radicals through Fenton-like reaction [24, 25]. In the present work, using silicon nanoparticles (SiNPs) which with low cytotoxicity, high biocompatibility, good hydrophilicity and stable fluorescence as fluorophore [26], the authors developed a novel method to detect Cu2+ based on the strong oxidizing properties of hydroxyl radicals which were produced by the dual catalytical effects of Cu2+. The fluorescence of SiNPs was quenched significantly by hydroxyl radicals for oxidizing and destroying the surface state of SiNPs, which was proportional to the concentration of Cu2+. This novel method shows high selectivity and sensitivity toward Cu2+ due to the unique dual catalytical effects of Cu2+. Furthermore, the application of this method to detect Cu2+ in real water samples is also demonstrated successfully.

Experimental section

Reagents

(3-Aminopropyl)trimethoxysilane (APTMS) (97%) was purchased from Shanghai Energy Chemical Co., Ltd. (http://www.energy-chemical.com/). Other reagents were purchased from Sinopharm Chemical Reagent Co., Ltd. (http://www.reagent.com.cn:666/ScrcBackGroup/reagent/newindex.jsp), and used directly without further purification. HEPES (10 mM) buffer was prepared by dissolving 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid in ultrapure water and adjusting the pH value to 8.0 with 1 M NaOH. The ultrapure water from a Millipore system (resistance ≥18.2 MΩ cm) was used throughout the experiments.

Apparatus

The UV-vis absorption spectra were measured on a UV2450 spectrophotometer (Shimadzu CO., LTD, Japan). The fluorescence (FL) emission spectra were collected on an F-7000 spectrophotometer (Hitachi CO., LTD, Japan). The Fourier transform infrared (FT-IR) spectra were performed from Nicolet Nexus 670 FT-IR spectrometer (Nicolet Instrument CO., U.S.A.). The microstructures of the SiNPs were examined using a JEOL-1230 transmission electron microscope (JEOL CO., LTD, Japan).

Preparation of silicon nanoparticles and detection of Cu2+

SiNPs were prepared according to the literature [26] (The detailed synthesis process was presented in supplementary material). The detection of Cu2+ was performed at environment temperature (about 20 °C) in pH 8.0, 10 mM HEPES buffer. Firstly, 20 μL SiNPs (600 μg mL−1) solution and 10 μL Cys (50 mM) solution were mixed with 960 μL HEPES buffer. Then, 10 μL of varied concentrations of Cu2+ were added into the mixed solution. The above solutions were incubated for 30 min at environment temperature (about 20 °C). After that, the fluorescence emission spectra were recorded immediately under excitation wavelength at 350 nm. For comparison, other ions including Na+, K+, Ag+, Mg2+, Ni2+, Zn2+, Ca2+, Co2+, Cd2+, Hg2+, Pb2+, Mn2+, Fe2+, Fe3+ and Al3+ were also tested by the same conditions. Each experiment was repeated three times. The error bars represent standard deviation from 3 replicates.

Real water samples analysis

The tap water sample was collected from the lab’s tap in the authors’ institute. The mineral water was collected from the Yuelu mountain (Changsha, China). The river water was collected from the Xiangjiang river (Changsha, China). The real water samples (tap water, mineral water and river water) were simply filtered with cylinder membrane filter (0.22 μm, Pall 66,191, Pall Corporation, USA) and diluted 10 times before the analysis. The following steps were similar to that of fluorescence detection of Cu2+, and the test solutions were spiked with standard solutions including varied concentration of Cu2+. Besides, the analytical results were found by atomic absorption spectrometry (data were performed by an atomic absorption spectrophotometer, Beijing Purkinje General Instrument CO. LTD, China) and from the Chinese National Standard GB/T 7475-1987 (water quality--determination of copper, zinc, lead and cadmium--atomic absorption spectrometry).

Results and discussion

Choice of materials

The morphology and property of nanomaterials have important influence on the performance of fluorescence detection methods. Compared with semiconductor quantum dots (e.g. CdSe, CdS and ZnS etc.), which are usually composed of heavy metals, SiNPs have the advantages of good water dispersibility, low toxicity, resistance to photobleaching and good biocompatibility [27]. On the other hand, the larger specific surface area and abundant surface functional groups of SiNPs make it easy interaction with guest species. Moreover, the SiNPs also have strong fluorescence, which is helpful to improve the detection sensitivity. Therefore, based on the above advantages, the authors developed a fluorescence method to detect Cu2+ by using SiNPs as fluorophore.

Characterization of SiNPs

Transition electron microscopy (TEM) experiments were performed to investigate the morphology of the SiNPs. Fig. 1a shows that a great deal of spherical nanoparticles can be observed in the TEM image and that the SiNPs possess excellent monodispersity and uniform size.. The diameters were calculated to be about 6-8 nm (Fig. 1b). In order to demonstrate the surface structure of the SiNPs, the FT-IR spectra information was collected. As shown in Fig. 1c, the absorption peaks at 1100 cm−1, 1384 cm−1, 1577 cm−1, and 3423 cm−1 are assigned to the Si-O stretching vibration C-O bending vibrations, N-H bending vibrations and the N-H stretching vibration, respectively. The FT-IR results indicated that amino groups are rich on the surface of SiNPs, which is similar with the literature [26]. In addition, as shown in Fig. 1d, the absorption, excitation and emission spectra demonstrated that as-synthesized SiNPs were an outstanding candidate for fluorescent materials. The UV-Vis spectrum shows absorption peak at 350 nm while the maximum fluorescence emission intensity centered at 450 nm when excited at 350 nm.

Fig. 1
figure 1

a TEM image of the SiNPs. b Size distribution histograms of SiNPs. c FT-IR spectrum of the SiNPs. (D) UV-vis absorption spectrum, excitation spectrum and emission spectrum of SiNPs

Full size image

Mechanism of the fluorescent detection method

The feasibility of using SiNPs for Cu2+ detection was explored. Fig. 2 shows the fluorescence spectra of SiNPs solution under various conditions. It can be seen that under the 350 nm light excitation, the SiNPs exhibits a strong fluorescence emission peak centre at 450 nm. The presence of Cu2+ or Cys alone does not affect the fluorescence of SiNPs. However, presence of both Cu2+ and Cys leads to an obvious decrease of fluorescence intensity of SiNPs, indicating that Cu2+ and Cys can quench the fluorescence of SiNPs effectively. These results indicate that it is feasible to quantify Cu2+ by measuring the fluorescence intensity of SiNPs.

Fig. 2
figure 2

The fluorescence spectra of SiNPs in the presence of Cu2+, Cys and Cu2++Cys. Experiments were performed in HEPES buffer (pH 8.0, 10 mM), the concentration of SiNPs, Cu2+ and Cys are 12 μg mL−1, 10 μM and 0.5 mM, respectively, the excitation/emission wavelengths are 350/450 nm, slits width are both 5 nm)

Full size image

The quenching mechanism was investigated. Previously study indicated that the Cu2+ can catalyze the oxidation of Cys to cystine and produce H2O2 under the O2 atmosphere [22, 23]. The overall process of Cys oxidation as:

$$ 2\ \mathrm{RSH}+{\mathrm{O}}_2\ \overset{{\mathrm{Cu}}^{2+}}{\to }\ \mathrm{RSSR}+{\mathrm{H}}_2{\mathrm{O}}_2 $$

The authors inferred that the Cu2+ can further catalyze the decomposition of H2O2 to produce hydroxyl radicals through a Fenton-like reaction [24, 25]. Similar to the reported carbon quantum dots [28, 29] and semiconductor quantum dots [30], the authors further inferred that the surface states of SiNPs may be destroyed and changed by the generated hydroxyl radicals, which results the fluorescence quenching of SiNPs. Control experiments were carried out to demonstrate the fluorescence quenching mechanism. As shown in Fig. S1, the H2O2, cystine, cystine+H2O2 cystine+Cu2+ do not affect the fluorescence of SiNPs. Conversely, as shown in Fig. S1B, the fluorescence of SiNPs was quenched effectively in the presence of both Cu2+ and H2O2. Furthermore, the fluorescence quenching efficiency was decreased to a certain degree when dimethyl sulfoxide (DMSO, a trapping agent of hydroxyl radicals) was added to the solution (Fig. S2A). These results indirectly indicated that the hydroxyl radicals are associated with the fluorescence quenching of SiNPs. Similar to the published literatures that the high activity hydroxyl radicals can destroy the original surface state of SiNPs, resulting the fluorescence quenching of SiNPs [31]. Fig. S2B shows the fluorescence recovers to a certain degree when NaBH4 is added to the solution. Therefore, the authors speculated that the fluorescence quenching of SiNPs was caused by the changes of the surface state of SiNPs through hydroxyl radicals. The detection mechanism is shown in Fig. 3.

Fig. 3
figure 3

Schematic illustration for the detection of Cu2+

Full size image

Optimization of method

The following parameters were optimized: (a) the concentration of Cys; (b) the concentration of SiNPs; (c) pH and (d) incubation time. Respective data and figures are given in the Electronic Supporting Material (Fig. S3). The following experimental conditions were found to give the best results: (a) the concentration of Cys is 0.5 mM; (b) pH = 8; (c) the concentration of SiNPs is 12 μg mL−1 and (d) an incubation time of 30 min.

The analytical performance of the method

Under optimal conditions, the analytical performance of the method for Cu2+ detection was evaluated. Fig. 4a shows the fluorescence emission spectra of the SiNPs in the presence of various concentrations of Cu2+. It can be seen that the fluorescence of SiNPs at 450 nm decreased gradually with the increasing of Cu2+. The relationship between the fluorescence quenching rate of SiNPs [(F0-F)/F0] and the concentration of Cu2+ is shown in Fig. 4b. It is clearly seen that a good linear correlation is observed over the concentration range of 0.05 μM to 15 μM. The corresponding calibration equation is y = 0.0424× + 0.0663 (R2 = 0.9984), where y and x are the fluorescence quenching rate [(F0-F)/F0] and the concentration of Cu2+, respectively. According to the ratio of 3 times the standard deviation of the blanks to the slope of the linear equation of the calibration curve, the limit of detection (LOD) is calculated as 29 nM. The LOD is lower than the maximum concentration of Cu2+ in drinking water by the US Environmental Protection Agency (20 μM). Table 1 summarizes some nanomaterials-based fluorescence method for Cu2+ analysis [2, 18,19,20,21, 32,33,34,35,36,37,38,39]. Some metal ions with similar chemical properties or ionic radius may affect the interaction between Cu2+ and the fluorescence nanomaterials. Thus, these detection methods had lower selectivity to Cu2+. J. Hwang et al. reported a fluorescence method for the detection of Cu2+ by using electrochemically etch method synthesized porous silicon nanoparticles as a fluorescent probe [39]. However, the detection mechanism of J. Hwang and his coworker is different from the present method. The method of J. Hwang and his coworker was based on the interaction between Cu2+ and porous silicon nanoparticles, which may affected by Fe2+. Conversely, the present method has good selectivity toward Cu2+ due to the dual catalytic effects of Cu2+. Therefore, the authors believe that they provide an effective new approach for the selective detection of Cu2+.

Fig. 4
figure 4

a Fluorescence spectra of SiNPs+Cys in the presence of different concentrations of Cu2+. b The relationship between the fluorescence quenching rate and the concentration of Cu2+; Inset shows the calibration curve at low concentrations. (Performed in HEPES buffer (pH 8.0, 10 mM), the concentration of SiNPs, and Cys are 12 μg mL−1 and 0.5 mM, respectively. F0 and F correspond to the fluorescence intensity of the SiNPs at 450 nm in the absence and presence of Cu2+, respectively. The excitation wavelength is 350 nm, slit width are both 5 nm. The error bars represent standard deviation from 3 replicates, and magnified by a factor of three)

Full size image
Table 1 Comparisons of analytical performances of various nanomaterials-based methods for Cu2+ detection
Full size table

Next, the fluorescence quenching data were further analyzed by the Stern-Volmer equation:

$$ {I}_0/I={K}_{\mathrm{SV}}\times {c}_{\mathrm{Q}}+1 $$

Where I0 and I are the fluorescence intensities of SiNPs in the absence and presence of Cu2+, respectively; cQ is the concentration of the Cu2+; KSV is the Stern-Volmer quenching constant. As shown in Fig. S4, the data do not fit a conventional linear Stern-Volmer equation, indicating both dynamic and static quenching processes occur in the present detection method [19, 36].

In order to evaluate the selectivity of this method for the detection of Cu2+, various relevant metal ions such as Na+, K+, Ag+, Mg2+, Ni2+, Zn2+, Ca2+, Co2+, Cd2+, Hg2+, Pb2+, Mn2+, Fe2+, Fe3+, and Al3+ were investigated. As shown in Fig. 5a, the fluorescence quenching rate caused by Cu2+ is greater than that of other metal ions, indicating that this method has a good selectivity for Cu2+ detection. Although the Fe2+ can react with H2O2 to generate hydroxyl radicals, the Fe2+ can not catalytic the Cys to produce H2O2. Thus, the authors speculate that the selectivity may be due to the unique catalytic reaction between Cu2+ and Cys [22, 23].

Fig. 5
figure 5

The fluorescence quenching rate of SiNPs in the presence of Cu2+ and other compounds. (Performed in HEPES buffer (pH 8.0, 10 mM), the concentration of SiNPs, and Cys are 12 μg mL−1 and 0.5 mM, respectively. The concentration of Cu2+ is 10 μM, other compounds is 200 μM. F0 and F correspond to the fluorescence intensity of the SiNPs at 450 nm in the absence and presence of interfering substance, respectively. The excitation wavelength is 350 nm, slit width are both 5 nm. The error bars represent standard deviation from 3 replicates, and magnified by a factor of three)

Full size image

Analysis of Cu2+ in real water samples

The practicality of this method to detection of Cu2+ in real samples was further validated by a standard addition method. Certain concentration of Cu2+ standard solutions were directly spiked into the real water samples (tap water, mineral water, river water) before sample pre-treatment. The analysis results are summarized and listed in Table S1. Satisfactory recoveries (94.0%-105.8%) were obtained for all samples, implying that this method has potential application in the detection of Cu2+ in real applications.

Conclusions

In summary, a novel and effective fluorometric method for the detection of Cu2+ has been developed. It is based on the excellent fluorescence properties of SiNPs as well as the the dual catalytical effects of Cu2+. Under the optimal conditions, a relative low detection limit and a good linear for Cu2+ were obtained. Compared with other reported Cu2+ methods which commonly based on the interaction between Cu2+ and the surface groups of the nanomaterials, this method has better selectivity due to the dual catalytical effects of Cu2+. In addition, this method can be applied to the detection of Cu2+ in real samples and obtained satisfactory results, indicating its potential application in practice. This research will provide an effective new approach for the design of Cu2+ selective detection method. Nonetheless, it should be noted that this novel method also has a limitation. As the rapid and on-site detection of Cu2+ is of great significance to human health and pollution treatment. The detection time of this method is 30 min and the response is irreversible, which limits its online detection application. However, this method is still expected for sensitive and reliable detection of Cu2+ due to its high selectivity.