Introduction

Mercury ions (Hg2+) are one of the most toxic heavy metal ions [1, 2]. They can cause serious injury to brain, lungs, kidney, nervous system, immune system, etc [3,4,5,6]. In order to protect human beings from the toxic effects of long-term exposure to Hg2+, The World Health Organization (WHO) has defined the maximum level of inorganic mercury in drinking water as 30 nM [7]. Therefore, sensitive and selective detection of Hg2+ is very important in monitoring of aqueous environment.

Many methods have been utilized to detect Hg2+, including atomic emission spectroscopy (AES) [8], atomic absorption spectroscopy (AAS) [9, 10], fluorescence [11], inductively coupled plasma mass spectrometry (ICP-MS) [12, 13], and electrochemical method [14, 15]. However, these methods require time-consuming and tedious sample preparation and treatment, and expensive instrumentation, which seriously limit their applications in on-site detection of Hg2+.

Colorimetry is simple and inexpensive, and color changes can be detected visually [16,17,18,19]. Nobel metals, especially gold nanoparticles (AuNPs) have used as colorimetric probes for Hg2+ detection [20,21,22,23,24] due to unique optical properties known as stronger localizedsurface plasmon resonance (LSPR) and readout distinguishable to the bare eye [25,26,27]. However, most of AuNP surface needs to be modified with specific surface functionalized molecules for selective interaction with Hg2+. In addtion to strong interaction with Hg2+, these AuNP surface functionalized molecules also interact with other transition metal ions, affecting the selectivity.

To solve the above issue, herein, we present a highly selective colorimetric method for Hg2+ detection based on gold amalgam-induced deaggregation of AuNPs in the presence of aminopropyltriethoxysilane (APTES). The APTES molecules absorbed on the surface of AuNPs destabilize the dispersion state of AuNPs, leading to a distinct color change from red to blue-purple. Interestingly, in the presence of different concentrations of Hg2+, the formation of gold amalgam makes that reduced Hg occupy the former location of APTES, and prevents AuNPs from aggregation, resulting in a blue-purple-to-red color change. The concentration dependent color and absorbance changes suggest that the assay can detect Hg2+ up to nanomolar level.

Experimental section

Materials

Hydrogen tetrachloroaurate trihydrate (HAuCl4·3H2O, 99.9%), trisodium citrate, APTES, and Tris-base were purchased from Sigma-Aldrich. All other chemicals are of analytical reagent grade. Ultrapure water obtained from a Millipore water purification system (>18.2 MΩ cm, Milli-Q, Millipore) was used in all assays and solutions. All glasswares were cleaned with freshly prepared aqua regia and rinsed thoroughly with DI water prior to use.

Apparatus

UV-vis spectra were recorded using a U-3900 UV-vis spectrophotometer and processed using OriginLab software (Hitachi, Japan). The size and morphology of AuNPs were observed with a Hitachi H-7500 transmission electron microscope (TEM, Tokyo, Japan).

Synthesis of AuNPs [28]

AuNPs with an average diameter of 13 nm were prepared through the previous reported method (Supporting Information).

Detection of Hg2+

10 μL of various concentrations of Hg2+ solutions (final concentration: 15.4, 46.2, 61.5, 76.9, and 92.3 nM) was added into 540 μL AuNP solution (2.5 nM), respectively. After reaction for 1.5 h, 70 μL of 20 mM Tris-HCl buffer (pH 7.2) was separately added into the mixtures. Then, 30 μL 4.2 mM APTES was injected into the above solutions for 20 min of incubation at room temperature. Finally, the solutions were used for absorbance measurement at a wavelength of 520 and 620 nm.

Results and discussion

Detection principle

Scheme 1 shows the schematic illustration of the Hg2+ assay. In the absence of target Hg2+, APTES molecules were absorbed onto the surfaces of AuNPs via Au-NH2 bonds. The positive charges of APTES weakened the negative charges of AuNPs, leading to the AuNP aggregation. The solution color quickly changed from red to blue-purple. While in the presence of target Hg2+, citrate ions attached to the AuNP surfaces reduced Hg2+ to Hg, and then formed the gold amalgam deposited onto the surfaces of AuNPs. The formation of the gold amalgam may effectively inhibit the accessibility of APTES molecules to the surfaces of AuNPs, and prevented the aggregation of AuNPs. The solution underwent the color change from blue-purple-to-red, thus deaggregation of AuNPs resulted in a bare eye-based assay for Hg2+.

Scheme 1
scheme 1

Schematic illustration of colorimetric detection of Hg2+ based on gold amalgam-induced deaggregation of AuNPs

Full size image

Optimization of detection conditions

To achieve an optimal sensing performance, the experimental conditions including APTES concentration, solution pH, and reaction time between Hg2+ and AuNPs were examined. Fig. S1A shows the absorbance change (K0/K) of the solution toward APTES with varying concentrations in the presence of 1 μM Hg2+, where K0 and K were the absorbance (K = A620 nm/A520 nm) of the AuNP solution in the absence and presence of the target Hg2+, respectively. As seen from Fig. S1A, K0/K increased with the increasing Hg2+ concentration (66.7-193.8 μM) and then decreased after 193.8 μM. Therefore, a Hg2+ concentration of 193.8μM was used for the following experiment. Next, we investigated the optimum reaction between Hg2+ and AuNPs in the assay. Fig. S1B shows that the leaping range in K0/K occurred in 1.5 h, it was all downhill after 1.5 h. Thus, the optimal reaction time was 1.5 h. The pH value of the solution was also a key factor that affected the sensitivity of the assay. In this case, solutions with various pH values (4-8) were used to investigate the performance of the assay. As shown in Fig. S1C, the maximum K0/K was obtained when pH was 7.2. However, higher pH caused the decrease of K0/K. Thus, all subsequent experiments were conducted at pH 7.2.

Colorimetric detection of Hg2+

Pursuant to characterizing the sensitivity and the limit of detection (LOD) of the assay for Hg2+. The UV-vis absorption spectra of AuNP solutions were recorded after adding various concentrations of Hg2+ under the optimized conditions. As shown in Fig. 1a, the absorbance decreased with the increment of Hg2+ (0-92.3 nM). An obvious color change from blue-purple to red was observed with the increase of target Hg2+ concentration in the solution (inset of Fig. 1b). Also, a linear relationship between the absorption ratio (A620 nm/A520 nm) and the Hg2+ level (0-92.3 nM) was obtained (Fig. 1b). The LOD was 10 nM estimated by the 3σ rule, which was lower than the guideline Hg2+ concentration of (30 nM) in drinking water set by the WHO. Compared with other methods for Hg(II) detection, as shown in Table 1, the sensitivity of our method was comparable or more sensitive. Furthermore, the other direct evidence for the deaggregation of AuNPs induced by the addition of Hg2+ was confirmed by the TEM measurement. As shown in Fig. 2, the AuNPs underwent an aggregation-dispersion process in the presence of target Hg2+.

Fig. 1
figure 1

a UV-vis spectra of the solutions response for the Hg2+ concentration (0-92.3 nM). b Plot of A620 nm/A520 nm vs Hg2+ concentration (0-92.3 nM). Error bars indicate standard deviation of the mean of three experiments

Full size image
Table 1 Comparison of our approach with other reported methods for Hg(II) detection
Full size table
Fig. 2
figure 2

TEM images of AuNPs in the (a) absence and (b) presence of 92.3 nM Hg2+ in the presence of 0.2 mM APTES

Full size image

Selectivity of the assay

An essential feature of a metal ion assay is its selectivity not only to isolated metal ions but also to mixtures that are environmentally relevant. We tested 8 individual metal ions including Sn2+, Mn2+, Na+, Cd2+, Co2+, Zn2+, Pb2+, and Fe2+ at a concentration of 1 μM, and the mixture of the 8 metal ions. As shonwn in Fig. 3, the interference from the 8 metal ions and the mixture only caused a slight decrease in the absorbance, whereas 0.1μM Hg2+ gave rise to a tremendous decrease, compared with that of the background. Furthermore, the corresponding color photographs also show that only Hg2+ caused a blue-purple-to-red color change while other environmentally relevant metal ions remained blue-purple. Obviously, the interference of these metal ions to the assay was negligible. The excellent selectivity can be attributed to the specific interaction of gold with Hg2+.

Fig. 3
figure 3

The absorption ratio (A620 nm/A520 nm) change for Hg2+ and other environmentally relevant metal ions. Inset: color changes of the solution in the presence of 0.1μM Hg2+ and 8 individual metal ions and the mixture of the 8 metal ions, each at 1 μM. Error bars indicate standard deviation of the mean of three experiments

Full size image

The practicality of this assay

To explore whether the method was feasible to real samples, the practicality of Hg2+ detection in river water samples using the method was tested. The river water was collected from city moat (Beijing, China) and filtered through 0.22 μM nitrocellulose membranes to remove physical impurities. There was no detectable Hg2+ existing in the river water samples analyzed by ICP-MS. Therefore, employing the standard addition method, Hg2+ was respectively spiked into river water samples at different concentrations (15.4, 30.8, 46.2, 61.5, and 76.9 nM) and then measured with our method. As shown in Fig. 4a, it is seen that the absorbance decreased with the Hg2+ concentration increasing from 0 to 92.3 nM. The calibration curve for detecting Hg2+ in river water samples was obtained by plotting the values of (A620 nm/A520 nm) versus Hg2+ concentrations (0-92.3 nM) (Fig. 4b). In spite of the interference from organics and many minerals existing in river water, the results from the river samples were in good agreement with those from Tris-HCl buffer, indicating that the assay can satisfy the practical determination of Hg2+ in real samples.

Fig. 4
figure 4

a UV-visible spectra of AuNP solution in the presence of various Hg2+ concentrations (from top to bottom: 0, 46.2, 61.5, 76.9, and 92.3 nM) in river water. b The dependence of A620 nm/A520 nm on the Hg2+ concentrations in the range of 0-92.3 nM. Error bars indicate standard deviation of the mean of three experiments

Full size image

Conclusion

In summary, we have demonstrated a sensing strategy to be utilized for the determination of Hg2+ in aqueous solution. In the sensing system, target Hg2+ absorbed on the surfaces of citrate-capped AuNPs was reduced to Hg, the specific interaction between Au and Hg to form gold amalgam was employed to prevent the AuNP aggregation in the presence of APTES. The aggregation-to-deaggregation change can be easily accomplished by the bare eyes or UV-vis spectrometer. The method can provide a LOD of 10 nM with outstanding selectivity in mixed solutions containing eight other metal ions. Moreover, the experiments for detection of Hg2+ in river water had been demonstrated with satisfactory results. The limitation of the study is that the sensitivity needs to be further improved.