A sensitive and selective Hg2+ optical sensor has been developed based on the redox interaction of Hg2+ with starch-coated silver nanoparticles (AgNPs) in the presence of 0.005 mol L−1 HNO3. The relative intensity of the localized surface plasmon absorption band of AgNPs at 406 nm is linearly dependent on the concentration of Hg2+ with positive slope for the concentration range 0–12.5 μg L−1 and negative slope for the concentration range 25–500 μg L−1. Experiments performed demonstrated that metal ions (Na+, K+, Mg2+, Ca2+, Pb2+, Cu2+, Zn2+, Cd2+, Fe3+, Co2+, and Ni2+) do not interfere under the same conditions, due to the absence of oxidative activity of these ions, which guarantees the high selectivity of the proposed optical sensor towards Hg2+. The limits of detection and quantification were found to be 0.9 µg L−1 and 2.7 µg L−1, respectively, and relative standard deviations varied in the range 9–12% for Hg content from 0.9 to 12.5 μg L−1 and 5–9% for Hg levels from 25 to 500 μg L−1. The method was validated by analysis of CRM Estuarine Water BCR505. A possible mechanism of interaction between AgNPs and Hg2+ for both concentration ranges was proposed on the basis of UV-Vis, TEM, and SAED analyses.

1. Introduction

Monitoring of toxic metals in aquatic ecosystems is an important analytical task as far as these contaminants adversely affect the environment and have serious medical effects on human health. One of the most harmful pollutants among them is Hg, which is still released in the environment and widely distributed in air, water, and soil [1]. At very low concentrations, Hg affects human’s health, causing a variety of diseases to the heart, kidneys, brain, and nervous and endocrine systems [2] Naturally occurring levels of mercury in groundwater and surface water are less than 0.5 μg L−1, although local mineral deposits may produce higher levels in groundwaters. Essential quality standard for Hg maximum permissible limit of 1 μg L−1 has been adopted at EU level and requires regular monitoring of Hg content in drinking waters. It is well known that Hg exists in natural waters as different species: Hg0, methyl-Hg, and inorganic Hg(II); however, the dominant toxic species in drinking waters is Hg(II). Various instrumental methods and techniques have been developed for Hg determination at low environmentally relevant concentrations like atomic absorption/emission spectrometry (AAS/AES) [3], atomic fluorescence spectrometry (AFS) [4, 5], and high-performance liquid chromatography (HPLC) [6, 7]. In spite of being very sensitive and precise for Hg determination, these methods often require a time-consuming sample preparation step as well as expensive instrumentation. Various colorimetric assays (based on the use of sensitive chromophores or fluorophores [811], polymers [12, 13], oligonucleotides [14, 15], DNA [16, 17], and metal nanoparticles [1820]) have been developed and reported in the literature as convenient and simple alternative methods for the detection of target analytes without the requirement of sophisticated apparatus.

Metal nanoparticles have unique properties and applications in numerous fields, which are attributed to the collective dipole oscillation known as Surface Plasmon Resonance (SPR) [21]. This phenomenon makes them very desirable for colorimetric sensing of Hg2+ ions because the interaction between the nanoparticles and the analyte changes the intensity and/or position of the absorption band in the visible spectrum, which often might be observed with the naked eye [22]. The limitations observed for these systems are mainly connected with poor selectivity, high detection limit for Hg(II), complicated synthesis of the probe materials, or complicated analytical procedures.

In this study, we present a new colorimetric assay for Hg2+ ions in 0.005 mol L−1 HNO3 using starch-stabilized silver nanoparticles (AgNPs). A change in the absorbance strength is expected as a result of the redox interaction between AgNPs and either Hg2+ ions or ions. The Hg concentration determines which of these two redox reactions dominates as the two oxidants compete with each other for Ag oxidation. This way, detection of very low environmentally relevant Hg contents is possible. Several sensing systems have been already reported based on the interaction between AgNPs and Hg(II) ions [2332]; however, detailed study of Hg behavior in the presence of another competitive oxidant is rarely performed and discussed. A dual functional sensor for determination of Hg and H2O2 has been developed based on a similar approach: addition of H2O2 to a mixture of AgNPs and Hg(II) ions [33]. The method presented in this study, however, differs not only as a mechanism of the process, but also as a behavior of Hg2+ ions at very low concentrations (below 25 μg L−1) towards AgNPs in the presence of ions as a second oxidant. A simple and fast analytical procedure for determination of Hg in drinking waters is developed and verified by the analysis of a certified reference material.

2. Materials and Methods

2.1. Apparatus

UV-Vis absorption spectra were recorded on an Evolution 300 spectrometer (Thermo Scientific, USA) within the 200–800 nm range using quartz cuvettes with 1 cm optical path length. High-purity water was used as a reference sample for background absorption. The morphology and particle sizes were examined using a high-resolution transmission electron microscope (TEM, JEOL JEM-2100 operating at an accelerating voltage of 200 kV). A volume of 5 μL AgNPs suspension was placed on a carbon-covered copper grid for TEM and air-dried. The histogram of AgNPs size distribution and the mean diameter of nanoparticles were determined by counting at least 200 nanoparticles from the different TEM images using ImageJ software. Some structural details of the nanoparticles were analyzed using the high-resolution TEM image and SAED pattern. The zeta potential of nanoparticles was measured with a ZetaSizer Nano ZS (Malvern) instrument.

2.2. Chemicals

All chemicals used were of analytical-reagent grade and all aqueous solutions were prepared in high-purity water (Millipore Corp., Milford, MA, USA). Silver nitrate (AgNO3, 99.8%), soluble starch, sodium hydroxide (NaOH, 99%), nitric acid (HNO3, 65%), salts of the different cations studied (NaCl, KCl, MgCl2, CaCl2, Pb(NO3)2, ZnCl2, CuCl2, NiCl2, CdCl2, CoCl2, and FeCl3) (from Merck, Germany), and pharmaceutical grade D-(+) glucose (from Alfa Aesar, Germany) were used. Stock Hg standard solution, Trace CEPT™, 998 μg mL−1 in 2 mol L−1 HNO3 (Sigma-Aldrich, USA), was used to prepare a working standard solution of 1000 μg L−1 Hg2+ in 0.01 mol L−1 HNO3. Standard solutions for Hg within the concentration range of 0–1000 μg L−1 were prepared weekly by serial dilution of this solution in 0.01 mol L−1 HNO3. All diluted Hg solutions were stored in dark glass flasks and kept refrigerated at 4°C.

2.3. Synthesis and Characterization of Silver Nanoparticles

The synthesis of AgNPs follows a green synthetic procedure as described in our previous study [34]. The silver nanoparticles were obtained through a reduction reaction of silver nitrate with D-glucose as a reducing agent in the presence of starch as a stabilizer and suitable sodium hydroxide amount as a reaction catalyst. Briefly, 24 mL of 0.001 M AgNO3 and 48 mL of 0.2% solution of starch were mixed and left for at least 15 minutes to form a complex under an ultrasonic treatment (ultrasonic bath, power 100 W, frequency 38 MHz). After that, 720 μL of 0.1 M D-glucose was added and sonicated for 5 minutes. The reaction was started by the addition of 3.6 mL of 0.1 M NaOH and continued for one hour at a constant temperature (30°C) in an ultrasonic bath to ensure the homogeneous formation of the silver nanoparticles.

The as-prepared AgNPs were purified and concentrated three times by ultracentrifugation (90 min, 14,000 rpm). The dispersion obtained was denoted as a stock solution of AgNPs and used in the experiments for colorimetric determination of Hg2+. The AgNPs stock solution was kept in a dark glass flask at room temperature and was homogenized in an ultrasonic bath for 30 min prior to each experiment.

2.4. Colorimetric Detection of Hg2+ Ions

The colorimetric detection of Hg2+ ions via starch-stabilized silver nanoparticles was conducted as follows: an aliquot of 200 μL AgNPs stock solution and 300 μL high-purity water were consecutively added to a small quartz cuvette, followed by addition of 500 μL Hg2+ solution with varying concentrations. The resulting mixture was equilibrated by stirring on Vortex for an optimum incubation time and then the UV-Vis spectrum in the wavelength range of 200–800 nm was recorded. In order to investigate the sensitivity of the colorimetric assay towards other ions, starch-stabilized AgNPs were allowed to interact under the same conditions with 50 μmol L−1 solutions of alkali (Na+, K+), alkaline earth (Mg2+, Ca2+), Pb2+, and transition-metal ions (Cu2+, Zn2+, Cd2+, Fe3+, Co2+, and Ni2+) (separately for each ion). The resulting solutions were monitored by optical absorption spectroscopy.

2.5. Determination of Hg in Tap/Underground Water

Tap/underground water sample (20 mL) was filtered through a 0.45 μm filter and acidified with HNO3 until reaching pH in the range 2–2.3. Sample aliquot of 500 μL was transferred to a quartz cuvette, and 200 µL stock solution of AgNPs was added and the mixture was stirred by the Vortex. After the incubation time of 5 min, the UV-Vis absorbance was measured at 407 nm. Parallel sample aliquot of 250 μL is diluted twice with 0.005 mol L−1 HNO3 and passed through the procedure described above. The response of this sample (increase or decrease, related to the original one, Figure 4) is used to distinguish the low from the high linear concentration range of Hg and to choose an appropriate calibration curve.

3. Results and Discussion

3.1. Characterization of AgNPs

The UV-Vis absorption spectrum of starch-stabilized AgNPs, recorded at 25°C, is shown in Figure 1 (inset). A single and sharp SPR band appears at 407 nm, which indicates the formation of nanometer-sized particles. This is further confirmed by the TEM observation and size distribution histogram, shown in Figure 1.

The spherical-like AgNPs exhibit a relatively narrow size distribution with a mean diameter of 15.4 ± 3.9 nm. In addition to the nanospheres, some typical polyhedral nanoparticles (multiple twined nanocrystals) can be easily observed. The crystalline nature of AgNPs is clearly observed on the HRTEM image in Figure 1 (inset) and proved by the lattice characterization (e.g., the spacing between the individual lattice fringes of 0.235 nm, which corresponds to (111) plane lattice spacing of pure silver). The colloidal stability of starch-coated AgNPs is confirmed by the -potential value of −25.3 ± 1.3 mV measured in 0.001 mol L−1 KCl at pH 6.8.

3.2. The Optimization of Colorimetric Sensing of Hg2+

Several parameters were investigated systematically in order to establish optimal conditions for the direct colorimetric detection of Hg2+. As a first step, the pH value was adjusted taking into account the analysis of real samples and HNO3 which is typically used for water sample preservation. The experiments performed showed that 0.005 mol L−1 HNO3 ensured the highest sensitivity and could be accepted as an optimal sample medium. In order to evaluate the optimum contact time, the kinetic of interaction between AgNPs and Hg2+ in the presence of 0.005 mol L−1 nitric acid was followed within one hour by measurements of UV-Vis absorbance. Typical evolution of UV-Vis absorbance spectrum with time, due to the interaction of AgNPs with 400 μg L−1 Hg2+ and respective color change of the AgNPs dispersion, is shown in Figure 2.

The changes that occurred in the LSPR absorption band of AgNPs are reflected on the color of the samples, which can be seen even with the naked eye. It is seen that the sensor’s response is significant during the first five minutes of the reaction process and a negligible change in the absorption intensity is observed over time. This fact allows convenient analytical detection of Hg2+ within only five minutes.

As a next step, the sensitivity and applicability of starch-coated AgNPs for quantitative determination of Hg2+ ions under the defined optimal conditions were studied. The colorimetric response and LSPR band behavior were monitored as a function of Hg2+ concentrations, ranging from 0 to 500 µg L−1 in the presence of 0.005 mol L−1 HNO3 (Figure 3).

As seen from the UV-Vis absorbance spectra (5-minute incubation time), the addition of 0.005 mol L−1 HNO3 results in a considerable decrease of the intensity of AgNPs characteristic plasmon band at 407 nm accompanied by a slight blue shift (Figure 3(a)). In addition, a shoulder band appears at the wavelength range of 450–600 nm. The increase of Hg2+ concentration from 0 to 12.5 μg L−1 in 0.005 mol L−1 HNO3 leads to a gradual increase of the intensity of the characteristic plasmon band of AgNPs at 407 nm and its value gradually approximates to the absorption intensity of the blank nanoparticle solution (without both and Hg2+). In addition, the intensity of the shoulder band decreases along with increasing intensity of the main plasmon absorbance band. The spectra show a clear isosbestic point at 445 nm upon addition of Hg2+ in 0.005 mol L−1 HNO3, demonstrating that the aggregation of AgNPs is directly related to the concentration of Hg2+. Contrariwise, a gradual decrease of the intensity of the characteristic plasmon band of the AgNPs at 407 nm is observed for the Hg concentration range from 25 to 500 μg L−1. The spectra presented in Figure 3(b) also show that the decrease of intensity of the absorbance maximum at 407 nm is accompanied with a slight blue shift, which is strengthened for the higher concentrations of Hg2+. This phenomenon is already reported and described as a change of the refractive index of the particles and the formation of a mercury layer around AgNPs, yielding an amalgam-like structure [25, 35, 36]. It might be suggested that, for the first Hg concentration range (0–12.5 µg L−1), a redox reaction proceeds between zero-valent silver (Ag0) and either Hg2+ or ions. The values of standard electrode potentials of the components in the system confirm this suggestion: (Ag+/Ag0) = 0.799 V; (Hg2+/Hg0) = 0.854 V; (/) = 0.864 V. Because the standard electrode potential of / is commeasurable with that of Hg2+/Hg0, two competitive oxidizing agents are involved in the studied sensing system. The most probable explanation for the decrease of LSPR band intensity in the presence of 0.005 mol L−1 HNO3 and further increase upon addition of Hg2+ (Figure 3(a)) is that, at low Hg2+ concentrations, the oxidative effect of ions towards surface silver atoms is dominant. In the presence of higher concentrations of Hg2+ (Figure 3(b)), the surface of nanoparticles is protected by the layer of Ag-Hg-amalgam due to the sorption and reduction of positively charged Hg2+ on the surface of negatively charged AgNPs followed by amalgamation. In this way, the surface of AgNPs is inaccessible for oxidation by . Evidently, within the range of 25–500 μg L−1 Hg2+, the main redox interaction is between the AgNPs and Hg2+. Such behavioral dissimilarities of the analyte (Hg2+) for different concentration ranges have not been observed and reported in the previous studies on the AgNPs-based optical sensing system for Hg2+ colorimetric detection. We have to point out, however, that none of these reports mention the acidity of the reaction media, which most probably determines the oxidizing power of reagents in the system.

For quantitative determination of Hg2+, the change of the intensity of LSPR band maximum of silver nanoparticles at 407 nm upon the addition of analyte with various concentrations was estimated as a ratio At/, where corresponds to the intensity of the absorbance maximum of blank AgNPs solution (without both and Hg2+ ions) and At corresponds to the intensity of the absorbance maximum of silver nanoparticles 5 min after the addition of Hg2+ standard solutions (Figure 4).

As indicated in Figures 4(a) and 4(b), linear correlations exist between the relative value of the absorbance maximum intensity and the concentration of Hg2+ over the concentration ranges 0–12.5 μg L−1 ( with ) and 25–500 μg L−1 ( with ), respectively. As a conclusion, the optical sensor studied using starch-stabilized AgNPs ensures a linear response over the concentration range from 0.9 to 12.5 μg L−1 which covers all environmentally relevant concentrations of Hg and might be used for fast screening of Hg in the aquatic environment. The second concentration range from 25 to 500 μg L−1 Hg2+ can be successfully applied for the determination of Hg in highly contaminated and rarely found industrial wastewaters.

3.3. Selectivity of Hg2+ Optical Sensing by Starch-Coated AgNPs

From an analytical point of view, it is very important to define the selectivity of the proposed system for colorimetric Hg2+ determination. This has been evaluated through the response of the assay to various environmentally relevant metal ions including Na+, K+, Mg2+, Ca2+, Pb2+, Cu2+, Zn2+, Cd2+, Fe3+, Co2+, and Ni2+ under the same conditions as in the case of Hg2+. The optical response of AgNPs to the tested ions (concentration level of 50 μmol L−1) after 5 min of their addition (separately for each ion) is illustrated in Figure 5. For comparison, the optical response of AgNPs to the Hg2+ ions at a concentration level of 2.5 μmol L−1 is also presented.

It is easy to observe that all other metal ions produce a much weaker signal (almost at baseline level) except Fe3+ which shows modest interference. The reason is that only Hg2+ can be reduced by surface atoms of AgNPs to form stable Ag-Hg amalgam. The addition of Fe3+ resulted in a tiny intensity decrease and red shift of the absorption band. This effect could be interpreted in terms of Fe(III) complexation with oxidized species of carbohydrates (starch and glucose) which are sorbed on the surface of silver nanoparticles [37].

3.4. Mechanism of Interaction between AgNPs and Hg2+

To elucidate the mechanism of sensing activity of the starch-coated AgNPs towards Hg2+, the nanoparticles were examined before and after Hg2+ exposure using TEM with SAED observations. Figure 6 shows TEM micrograph with the corresponding SAED pattern obtained from the agglomerate formed during interaction of AgNPs with Hg2+ solution at a concentration of 500 μg L−1.

As can be seen from Figure 6(a), the nanoparticles are of varying sizes and there is a large distribution after Hg2+ exposure. The TEM image shows a larger particle, which is surrounded by smaller particles. It seems that larger particles are undergoing Ostwald ripening. A similar observation is already reported for gold nanoparticles utilized for mercury removal from drinking water [38] and for colorimetric detection of Hg2+ using the AgNPs embedded in cyclodextrin-silicate composite [39].

The data from the analysis of SAED pattern (Figure 6(b)) are summarized with interpretation accuracy of 1% in Table 1. The analysis shows the existence of Ag2Hg3 amalgam (PDF 65-3156) and Ag (PDF 89-3722) as main phases in the aggregated mass formed during the interaction of starch-coated AgNPs with Hg2+. Some impurities of metallic Hg (PDF 01-1017) are also detected.

On the basis of TEM/SAED results, a multistep interaction of Hg2+ with the silver nanoparticles could be inferred. The interaction involves (i) the electrostatic attraction between negatively charged silver nanoparticles and positively charged Hg2+ species, decreasing the distance between nanoparticles; (ii) adsorption of Hg2+ on the surface of AgNPs and their reduction to Hg0 by the surface Ag atoms (simultaneously obtained Ag+ diffuse into the solution); (iii) amalgamation of the freshly generated mercury atoms with the surface Ag atoms [25, 32, 39]; (iv) the interaction of Hg2+ with AgNPs which decreases surface charges of nanoparticles, leading to their destabilization and aggregation. The latter one is confirmed by the shape evolution of AgNPs observed in Figure 6(a). The suggested mechanism of optical sensing of Hg2+ by starch-coated silver nanoparticles is illustrated in Figure 7.

3.5. Analytical Application

In order to test the applicability of the sensor developed for Hg2+ and total Hg determination, samples of tap water (Sofia) and mineral water (Gorna Bania, Kniagevo) were spiked at levels close to the permissible limit (drinking water) of 1 µg L−1. Total Hg content in these samples was defined preliminarily by ICP-MS and results for all samples were below 0.05 µg L−1 Hg. Recoveries achieved using the described procedure are in the range 93–97%, thus confirming the possibility of fast Hg2+ screening in drinking waters using the proposed sensor based on starch-coated AgNPs. The limits of detection (LOD) and limits of quantification (LOQ) were evaluated on the basis of repeated analysis of blank (AgNPs). The calculations were based on 3σ and 10σ criteria using the linear regression equations and slopes of calibration graphs for Hg2+ (Figure 4). The defined values for LOD (0.9 µg L−1) and LOQ (2.7 µg L−1) show that the proposed sensor is not suitable for surface water monitoring but might be successfully used for fast on-site control of the quality of sources for drinking water. Within-batch precision strongly depends on the analyte concentration in the measuring solution: 9–12% for Hg2+ in the range 0.9–12.5 µg L−1 and 5–9% for Hg2+ in the range over 25–500 μg L−1. Table 2 further summarizes the linear ranges and detection limits of various Hg2+ detection methods based on silver nanoparticles as a colorimetric sensing probe. It is evident that the proposed method ensures higher or equal sensitivity with those of earlier reported colorimetric AgNPs-based sensors [23, 26, 31, 33, 4042].

For partial validation of the procedure, CRM Estuarine Water BCR505 was analyzed after solid phase extraction (10-fold Hg enrichment) [43]. Three sample aliquots of 800 µL were analyzed according to the proposed analytical procedure. The result of 0.73 ± 0.08 nmol L−1 Hg was in reasonable agreement with the (additional material information) value of 0.69 nmol kg−1 Hg (138 µg L−1).

4. Conclusions

A simple, fast, and low cost analytical procedure is developed for easy and sensitive quantification of Hg2+ in the presence of 0.005 mol L−1 HNO3 by using starch-coated AgNPs as a LSPR-based optical sensor. The Hg2+ sensing is based on the optical response (change in the absorbance strength of LSPR band) of silver nanoparticles depending on the Hg2+ concentration. Possible mechanism of interaction between AgNPs and Hg2+ was proposed. An accurate and reliable determination of Hg is achieved in two concentration ranges: 0.9–12.5 µg L−1 and 25–500 µg L−1. The limits of detection and quantification achieved were 0.9 µg L−1 and 2.7 µg L−1, respectively, and relative standard deviations varied in the range 9–12% for Hg content from 0.9 to 12.5 μg L−1 and 5–9% for Hg levels from 25 to 500 μg L−1. The LSPR-based optical sensor for Hg(II) might be used for simple and fast on-site screening of sources for abstraction of drinking water and for Hg determination in wastewaters.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.


The authors acknowledge the support by the Horizon 2020 program of the European Commission (project Materials Networking).