Introduction

Silver ions (Ag+) is widely used in the electrical industry, photography imaging industry and pharmacy [1–3]. It has been considered to be a high-effect fungicide for a long time, which is always thought to be a precious metal safe to humans [4]. The standard safe concentration of Ag+ for human is <0.05 ppm according to the World Health Organization (WHO) [5]. But thousands of tons of Ag+ are released to the environment from industrial wastes and emissions annually. Excessive amounts of Ag+ are distributed and contaminated in ambient air, water, and soil, which may lead to serious environmental problems [6, 7]. And it can accumulate in human body through the food chain, and causing various problems to the human health such as cytotoxicity, organ failure, and reduction in mitochondrial function [4, 8]. Therefore, it has received considerable attention in recent years and recent studies emphasized bioaccumulation and the potential negative impact of Ag+ [9]. Hence, the quantitative determination of Ag+ is of great significance in diagnosing related diseases.

Nowadays, several technologies have been developed for Ag+ detection, including atomic absorption/emission spectroscopy [10–13], inductively coupled plasma-mass spectroscopy [14, 15], ion-selective electrodes (ISEs) [16, 17], voltammetry [18–20], ultraviolet–visible (UV–vis) [6, 21], florescence method [3, 22, 23] and electrochemical chemosensing [4, 5]. Despite tremendous efforts being made in detection of Ag+ with high selectivity and sensitivity, all these above-mentioned methods suffer from more or less drawbacks such as sophisticated equipment, complicated sample preparation processes, high cost, relatively high temperature and time-consuming immobilizing processes, which limit their wide application. Compared to other choices, the florescence methods in detection of Ag+ are fast, efficient and sensitive. Recently, the method that visual and trap emission spectrometric detections of Ag+ with mesoporous ZnO/CdS@SiO2core/shell nanocomposites has been developed [1]. However, it is difficult to eliminate the background interference resulted from the detection systems, which affects the sensitivity and detection limit. Therefore, it is essential to monitor Ag+ without background interference through fluorescence methods.

In recent years, increasing attention has been paid to the upconversion nanoparticles (UCNPs) and their potential applications [24–31]. UCNPs doped with certain rare earth ions which the low energy near-infrared light (typical 980 nm) can be converted into a higher energy visible emission via a two-photon or multiphoton mechanism [32–34]. Comparing with conventional fluorescent labels, as fluorescent biolabels, UCNPs possess several outstanding features including narrow emission peaks, large Stokes shifts, low toxicity, high photostability and thermal stability, minimum photodamage to living organisms, no autofluorescence and high signal-to-noise ratio [24, 35]. Thus, these excellent fluorescence properties make the UCNPs exploited to develop biosensors for a wide variety of target analytes.

In this study, enlightened by the above facts, we have developed a new UCNPs-based optical biosensor for Ag+ determination. The proposed detection strategy was shown in Scheme 1. Ag+ can oxidize o-phenylenediamine (OPD) to induce a yellow color and an absorption peak centered at 450 nm. The fluorescence of UCNPs can be significantly quenched by oxOPD through inner filter effects (IFE), which was proportional to Ag+. The high selective and sensitive detection of Ag+ was possible using the fluorescent method. Furthermore, this method has been successfully applied to the determination of Ag+ in real samples.

Scheme 1
scheme 1

Schematic illustration of the UCNPs fluorescence assay for the detection of Ag+

Full size image

Experimental

Materials

Rare earth oxides, including Y2O3, Yb2O3, Tm2O3, were purchased from china national pharmaceutical group corporation (Shanghai, China). AgNO3, CaCl2, CdCl2, CoCl2, CuCl2, FeCl3, Hg(NO3)2, Mg(OAc)2, MnCl2, NiCl2, PbCl2, ZnCl2 and FeCl2 were purchased from Beijing Chemical Corp. All the other chemicals (99 %, Merck) used in this work were of analytical grade and Millipore Milli-Q ultrapure water was used throughout the experiments.

Apparatus

The UV-2450 spectrophotometer (Shimadzu Co., Japan) was used to collect the absorption spectra of oxOPD. The fluorescent emission spectra of UCNPs were recorded by the F-4500 fluorescence spectrophotometer (Hitachi Co., Japan). Transmission electron microscopy (TEM) images were collected by the JEOL-1230 transmission electron microscope (JEM-3010 Joel, Japan). Fourier transform infrared (FTIR) spectra were collected on a Nicolet Nexus 670 FTIR spectrometer (Nicolet Instrument Co., U.S.A.).

Preparation of UCNPs

PEI-modied water-soluble NaYF4:Yb3+, Tm3+ were synthesized via a solvothermal method according to the previously reported method [36]. NaCl (0.15 g), PEI (0.4 g), Y(NO3)3 (0.4 mmol, 0.295 mL), Yb(NO3)3 (0.20 mmol, 0.6 mL), and Tm(NO3)3 (0.05 mmol, 0.04 mL) were dissolved in ethylene glycol (15 mL) under vigorous stirring. When the solution became transparent, NH4F (4 mmol) in EG (10 mL) was added to it. After being stirred for another 10 min, the mixture was transferred into a teflon-lined autoclave and heated to 200 °C for 4 h. Then, the solution was cooled down to room temperature naturally. The nanocrystals were precipitated from the solution by centrifugation and washed sequentially three times by using deionized water and ethanol. The product was dried in vacuum before it was used.

Assay for Ag+ in Aqueous Buffer

The detection procedure of Ag+: various amounts of Ag+ were added to same OPD (1 mM), UCNPs (50 μL, 0.1 mg mL−1), citrate buffer (100 mM, pH = 5.0), which were then kept for 10 min at 30 °C. The fluorescence emission spectra were measured with excitation wavelength at 980 nm. The relationships between the fluorescence intensity and the concentrations of Ag+ were plotted into calibration curves.

Detection of Ag+ in Real Samples

The proposed method applied to detect Ag+ was applied for the direct analysis of Ag+ in tap water to verify feasibility of this new Ag+ probe. 100 mL tape water was mixed with Ag+ aqueous solution, OPD (1 mM), and UCNPs (0.1 mg mL−1). Then citrate buffer (100 mM, pH = 5.0) was added into mixed solution and made up the final volume of 500 mL. And then the fluorescence emission spectra were measured with excitation wavelength at 980 nm.

Results and Discussion

Characterization of the UCNPs

The morphology, structure, and optical properties of UCNPs were characterized firstly (Fig 1). As shown in Fig. 1a, the TEM images of UCNPs are uniform in morphology with the average sizes of 40 nm. The typical XRD patterns of the UCNPs are presented in Fig. 1b. NaYF4:Yb, Tm nanocrystals were obtained consisting of a dominant cubic phase (JCPDS no.77–2042) by the XRD pattern, indicating that the highly crystalline nature possessed the synthesized material. FT-IR spectra were used to analyze the functional groups on the surface of UCNPs. As shown in Fig. 1c, the methylene asymmetric and symmetric C − H stretching (2960, 2850 cm−1), N − H stretching vibration (3441 cm−1), amine N − H bending (1640 cm−1), and amide bonds internal vibration (1380 cm−1) were correspond to the characteristic peak of amine groups (−NH2), which exhibited characteristic absorption peaks of PEI molecules (Fig. S1). Figure 1d shows the upconversion fluorescence spectrum of the aqueous dispersion under 980 nm laser excitation, a characteristic peak at 490 nm observed. Upon further investigation, we found that the fluorescence intensity of UCNPs possessed excellent photostability over 60 min in air (Fig. S2A) and was not affected by temperature (Fig. S2B) and the pH (Fig. S2C).

Fig. 1
figure 1

(a) TEM images of the as-prepared water-soluble PEI-NaYF4:Yb,Tm UCNPs. (b) XRD patterns of the NaYF4:Yb,Tm UCNPs. (c) FTIR spectra of the NaYF4:Yb,Tm UCNPs. (d) Emission spectrum of NaYF4:Yb,Tm UCNPs

Full size image

The Principle for Detection of Ag+

The detection mechanism is based on the facts that the fluorescence of UCNPs can be significantly quenched by oxOPD. OPD could be oxidized by Ag+ and the fluorescence of UCNPs can be quenched by oxOPD. As shown in Fig. 2, UCNPs solution exhibits a maximum fluorescent signal at 490 nm (1). Compared to other signal source, UCNPs has been broadly applied in biosensor because it generates visible emission with low-energy NIR excitation (normally 980 nm). UCNPs could eliminate the autofluorescence from biological macromolecules and thus provides an enlarged signal-to-background ratio and increased sensitivities. The fluorescence of UCNPs was quenched when UCNPs was mixed with Ag+/OPD system (4, 5, 6). In order to evaluate the feasibility of this assay for Ag+, we investigated the effects of the related factors. As shown in Fig. 2, the fluorescent intensity of UCNPs has no significant change when UCNPs individually incubated with OPD and AgNO3 under the same conditions (2, 3). The results indicated that the factors mentioned above should not affect the experimental results and OPD plays an important role in the quenching effect of the system.

Fig. 2
figure 2

Fluorescence spectra of UCNPs solution. (1) only UCNPs, (2) UCNPs + OPD, (3) UCNPs + AgNO3, (4) UCNPs + OPD + AgNO3 (0.15 mM), (5) UCNPs + OPD + AgNO3 (0.35 mM), (6) UCNPs + OPD + AgNO3 (0.5 mM), respectively. [UCNPs]: 0.1 mg/mL. [OPD]: 1 mM

Full size image

In this work, the possible mechanism for UCNPs capped PEI quenching was related to inner filter effects (IFE) or fluorescence resonance energy transfer (FRET). It has been reported that the absorption of the excitation and/or emission light by absorbers may reduce the FL intensity of the fluorophore; i.e., FL can be quenched by the so-called inner filter effect [37–39]. The UV–vis absorption spectroscopy and fluorescence emission spectra were measured to study the mechanism. UV-vis absorptions pectroscopy of oxOPD solution in the absence and presence UCNPs were recorded (Fig. S3(4,5)). No change in the absorption bands position and absorption intensity of oxOPD was observed after adding UCNPs, which indicated that no complex was formed and fluorescence resonance energy transfer (FRET) between oxOPD and UCNPs does not occur. From the absorption spectrum of oxOPD (Fig. S3(4)), we found that the aqueous pink solution of oxOPD displayed intense absorption at 450 nm. The water-soluble UCNPs were prepared through aqueous-phase synthesis, and they showed a fluorescence emission maximum at 490 nm (Fig. S3(6)), which overlapped very well with the absorption spectrum of oxOPD to induce IFE between them [39–41]. Simultaneously, we investigated the effects of the related factors. As shown in Fig. S3, the absorption intensity of them has no significant change when UCNPs individually incubated with OPD and AgNO3 under the same conditions (1–3). The results indicated that the factors mentioned above should not affect the experimental results. For further confirmation, the zeta potential was then measured to examine the change of surface charges in buffer solution and the experimental results showed the zeta potentials of dispersed UCNPs and oxOPD were all negative. Because both UCNPs and oxOPD are negatively charged, oxOPD could not be adsorbed on the surface of UCNPs via electrostatic interaction and their distanc should be greater than 10 nm. Therefore, the FRET between UCNPs and oxOPD can not occur in buffer solution. Judging from the above analysis, it is obvious that IFE contributed to the observed fluorescence decrease of UCNPs in the presence of oxOPD.

Optimization of the Detection Conditions

In order to enhance sensitivity of the sensor system for Ag+ determination, the parameters including pH, incubation time, temperature and the concentration of OPD were optimized. Firstly, the effect of pH on the fluorescence quenched efficiency was investigated in the range from pH 3.0 to 6.5. The pH value of the reaction solution has a significant influence on the response of the sensing system. OPD was easily oxidized only in the acidic environment, and so we chose citrate buffer which buffer range is pH 3.0–6.6. As shown in Fig. 3a, the maximal fluorescence quenched efficiency reached a maximum at pH 5.0. Therefore, pH 5.0 was chosen as the optimal pH for the detection of buffer solution. Fluorescence quenched efficiency was defined as (I 0-I)/I 0, and I 0 and I refers to the fluorescent intensity of UCNPs-OPD system in the absence and presence of Ag+, respectively. As shown in Fig. 3b, the fluorescent quenching of UCNPs-OPD system reached a maximum and sustained a stable value when the incubation time reached to 10 min. Thus, the ideal incubation time for Ag+ detection is 10 min. We also investigated the temperature of the reaction between the UCNPs-OPD system and Ag+. Figure 3c shows that the quenching efficiency reached a maximum when the temperature reached 30 °C. Thus, 30 °C was chosen for the following experiment. The concentrations of OPD have an influence on the sensitivity of the proposed sensing strategy for Ag+ detection. Figure 3d display the effect of OPD concentrations of the sensing system. With the increasing concentration of OPD, the fluorescence intensity decreased rapidly and reached a stable value when the concentration of OPD was 1 mM. Hence, the concentration of OPD was selected to be 1 mM for this experiment.

Fig. 3
figure 3

Effects of pH (a), incubation time (b), temperature (c), and OPD concentration (d) on the fluorescence responses of the sensor to Ag+ detection. I 0 and I stand for the fluorescent intensities of UCNPs in the absence and presence of Ag+, respectively. [UCNPs]: 0.1 mg/mL. [OPD]: 1 mM. [Ag+]: 0.1 mM

Full size image

Quantitative Analysis of Ag+

Under the optimal detection conditions, the UCNPs-OPD system in the presence of various concentration of Ag+ was investigated. As shown in Fig. 4a, the fluorescent intensity of the UCNPs-OPD mixture decreased (from top to bottom) with the increase of the Ag+ concentration. The fluorescent spectra of the system exhibits a good linear relationship with Ag+ in the concentration ranging from 0.0 to 0.5 mM and a linear regression equation was F = −0.0176 + 1.8098C with the correlation coefficient (R 2) of 0.99 (Fig. 4b), where F refers to the measured fluorescence intensity and C refers to the concentration of Ag+, respectively. A comparison between the method and other reported biosensors for Ag+ determination was summarized in Table S1. The detection limit can reach as low as 33 nM, which is lower than that of the some reported methods for Ag+ [5, 6, 22]. Our UCNPs-based fluorescent biosensor assay only needs cheap instrument and simple design, and can be performed without the presence of trained technician. In addition, our strategy has the best anti-interference performance for complex sample matrices analysis simultaneously. Thus it makes the assay with the advantages of convenience, low cost, fast, time saving and ease of operation.

Fig. 4
figure 4

(a) Quenching effects of UCNPs towards different concentrations of AgNO3. (from top to bottom, the final concentration of AgNO3 is 0, 0.001, 0.01, 0.1, 0.15, 0.22, 0.3, 0.35, 0.4, 0.45, 0.5, and 0.8 mM, respectively. (b) The linearity of the system towards different concentrations of AgNO3

Full size image

Selectivity of the Method towards Ag+

For determination of Ag+ in tap water, the main interference includes other metal ions such as CaCl2, CdCl2, CoCl2, CuCl2, FeCl3, Hg(NO3)2, Mg(OAc)2, MnCl2, NiCl2, PbCl2, ZnCl2 and FeCl2 were investigated. Typically, under optimum experimental conditions, the detection system was incubated respectively with 0.4 mM Ag+ and 4 mM other metal ions in buffer. As is shown in Fig. 5, compared with the Ag+ system, the existence of the other commonly seen metal ions had almost no influence on detection of Ag+. These results demonstrated that these species did not interfere the specificity of the method for Ag+.

Fig. 5
figure 5

Fluorescent quench of UCNPs solution mixture with Ag+ (0.4 mM) or other substances (4 mM): (a) Ca2+, (b) Cd2+, (c) Co2+, (d) Cu2+, (e) Fe2+, (f) Hg2+, (g) Mg2+, (h) Mn2+, (i) Ni2+, (j) Pb2+, (k) Zn2+, (l) Fe3+, (m) Ag+

Full size image

Analysis of Ag+ in Real Samples

In order to further investigate the potential application of the sensor in the practical samples, the assay was used to detect Ag+ in tap water. The real samples were spiked with certain amounts of Ag+. The assay medium and the fluorescence response of UCNPs-OPD system was detected by using the proposed strategy. The results obtained by standard addition method were listed in Table 1. As shown in Table 1, excellent quenching in the range from 98.40 % to 103.6 % was obtained. The desirable experimental results demonstrate the reliability of the proposed method for detection of Ag+ in practical applications. In particular, low-energy NIR excitation (normally 980 nm) can overcome the interference from background fluorescence.

Table 1 Recovery ratios of Ag+ in tap water
Full size table

Conclusion

In summary, we have successfully developed a novel, convenient, sensitive, accurate and reliable fluorescent biosensor for the determination of Ag+ based on the quenching of the fluorescence of UCNPs without complicated modification and expensive instruments. The Ag+ concentration can be analyzed by the fluorescence of UCNPs, and the normalized fluorescence of UCNPs is quantitatively correlated to the concentration of Ag+. Most importantly, the interference from background fluorescence and scattered light were eliminated by using the upconversion nanoparticles (UCNPs). In particular, the proposed method is successfully applied to the detection of Ag+ in real samples. Therefore, the convenient assay provides a promising tool for developing simple, fast, low cost and sensitive sensors for Ag+ detection.