Environmentally, mercury contamination of ecosystems occurs through a variety of natural and anthropogenic sources. These sources include oceanic and volcanic emissions, forest fires, gold mining, solid waste incineration, and the combustion of fossil fuels, resulting in increasing levels of Hg2+
emissions into the air, soil and water. Inorganic mercury can cause a wide range of diseases such as digestive, heart, kidney and especially neurological disorders [1
]. Mercury is one of the most harmful pollutants because it can easily pass through skin, respiratory, and gastrointestinal systems. A major absorption source is related to daily food such as fish. Therefore, it is important to monitor Hg2+
levels in aquatic ecosystems as a potential source of contamination, since the situation becomes increasingly serious to the living environment of humans and other species. The United States Environmental Protection Agency (US-EPA) has set the safety level of Hg2+
concentration for drinking water at 10 nM (2 μg·L−1
], and the World Health Organization [3
] and the European Union [4
] indicate for this ion a value of 5 nM (1 μg·L−1
), highlighting the necessity of developing sensitive methods for its determination.
Many current techniques for mercury determination, such as atomic absorption/emission spectroscopy, inductively coupled plasma mass spectrometry, and selective cold vapor atomic fluorescence spectrometry, require sophisticated instrumentation and/or complicated manipulations and time-consuming sample preparation processes [5
]. In this context, considerable attention has recently been focused on the exploration of simple and fast detection methods for aqueous Hg2+
ions and, particularly, on the design of luminescence chemodosimeters due to the highly sensitive, quick, and nondestructive advantages of luminescence methods, which make them an ideal probe displaying very low detection limits and high selectivity towards Hg2+
in the presence of other metals [6
The chemodosimeter strategy is based on the use of a selective reaction that is induced by the target species and gives rise to an observable luminescence signal. Generally, Hg2+
ions are known to produce fluorescence quenching when binding to fluorophore molecules via the spin-orbit coupling effect. However, a turn-on response is preferable, since the ubiquitous nature of fluorescence quenching reduces to some extend the practical utility of the turn-off probes. Therefore, fluorescent turn-on type molecular probes for monitoring the level of Hg2+
in environmental and biological samples is of current interest, and rhodamine and boron-dipyrromethene (BODIPY) derivatives are examples of families of small molecular probes successfully reported for selective turn-on Hg2+
Rhodamine 6G spirocyclic phenylthiosemicarbazide derivative (FC1) has been previously proposed for Hg2+
determination in diverse matrices, such as water and fish samples [10
], reporting a limit of detection of 1 μg·L−1
]. Unlike other chemosensors, the probe responds to Hg2+
in an irreversible manner, as the rhodamine 6G derivative undergoes oxadiazole formation when the thiosemicarbazide moiety is liberated by Hg2+
, facilitating ring opening of the spirocycle group. The reaction is based on the well-known spirolactam (nonfluorescent) to ring opened amide (fluorescent) equilibrium of rhodamine derivatives [13
], coupled to the irreversible Hg2+
-promoted reaction of thiosemicarbazides to form 1,3,4-oxadiazoles [16
]. The same probe and mechanism were also used in a LED excited portable fiber-optic system, lowering the limit of detection to 0.7 μg·L−1
]. The probe has been also proposed for methylmercury fluorescent detection [19
It is interesting to point out that, although the number of examples of rhodamine derivative probes reported for Hg2+
detection is high, there are only few examples of probes immobilized in solid supports [20
] and, in an effort to develop a probe for the construction of an optosensor, several approaches have been explored to immobilize FC1 in a solid support. In this sense, the procedure was improved by firstly immobilizing the reagent in a nylon membrane [23
]. With this approach, the limit of detection of the method was lowered to 0.4 μg·L−1
, and the probe was applied to the determination of Hg2+
residues in drinking and environmental waters. Later on, the probe was immobilized on hydrophilic water-insoluble poly(2-hydroxyethylmethacrylate-co-methylmethacrylate) co-polymer [25
], and encapsulated in electrospinning generated microfiber nonwoven mats [26
]. Then, the developed sensing films was applied to the detection of mercury and methylmercury ions in tap and mineral waters, although in these two approaches, the detection limit was higher than in solution.
AuNPs with well-defined structures have become an active research area, and combination of organic chromophores and AuNPs for producing typical organic/inorganic hybrid materials and unique photophysical and photochemical properties is still a challenge. Interestingly, they are excellent quenchers of organic fluorophores via resonance energy transfer (RET) and can efficiently quench the molecular-excitation energy in the chromophore–AuNP composite [27
]. In this sense, the quenching property of AuNPs has been employed in several chemosensing schemes [28
], and many approaches have been used for Hg2+
determination using AuNPs, being an area of current interest [30
]. Usually, the RET process occurring in a bimolecular dye-quencher system through single dipole–dipole interaction is the Förster RET (FRET) [37
]. Some authors postulate that, in the case of AuNPs, the dipole–dipole interaction is replaced by a dipole–multipole interaction due to the metallic surface, and the process is known as nanomaterial surface energy transfer (NSET) [38
]. In the case of rhodamine–AuNPs systems, different studies suggested the occurrence of FRET [32
], NSET [40
], and both FRET and NSET simultaneously [42
]. It is important to mention that the partial or complete overlap between the SPR band of the AuNPs and the emission band of the dye, acting in this case the rhodamine as the donor and the AuNPs as the acceptor, is a prerequisite for the occurrence of the efficient energy transfer.
Therefore, it is expected that in the presence of AuNPs FC1 exhibits lower fluorescence than in its absence (i.e., free FC1) due to the RET effect. In this paper, we report on the use of this effect for enhancing the sensitivity of the former method based on the reaction of FC1 towards Hg2+, maintaining its high selectivity.
2. Material and Methods
Fluorescence measurements were carried out on a Perkin Elmer LS 55 luminescence spectrometer (Llantrisant, UK) equipped with a xenon discharge lamp, using a 1.00 cm quartz cell and 10 nm of excitation and emission slit widths. The photomultiplier tube voltage was set to 800 V. The emission spectra were recorded between 540 and 600 nm at the excitation wavelength of 528 nm, with the exception of some cases mentioned below. The fluorescence measurements were made using a thermostated cell holder and a thermostatic bath (ORL Hornos Eléctricos S.A., Buenos Aires, Argentina). Absorbance measurements between 200 and 800 nm were obtained using a 1.00 cm quartz cell with a Perkin Elmer Lambda 20 spectrophotometer (Waltham, MA, USA).
The pH of solutions was measured with an Orion 410 A potentiometer (Beverly, MA, United States) equipped with a Boeco BA 17 (Hamburg, Germany) combined glass electrode.
Measurements of average diameter and size distribution of AuNPs were performed using a Zetasizer Nano S90 Malvern Dynamic Light Scattering (DLS) instrument (Malvern, UK).
2.2. Reagents and Procedures
Analytical reagent-grade chemicals and ultrapure water, obtained from a Milli-Q water purification system from Millipore (Kankakee, IL, USA), were used throughout the experiments. HPLC-grade MeOH and ACN were purchased from Merck Millipore (Danvers, MA, USA).
FC1 was synthesized following the procedure previously described in the literature [10
]. The stock solution of FC1 1 mmol·L−1
was prepared by dissolving 5.0 mg of FC1 in 10.00 mL of ACN. This solution was further diluted by 100 times in water:MeOH (80:20, pH = 7) to prepare an intermediate solution, which was kept in the fridge before being used.
AuNPs were synthesized by chemical reduction according to the procedure reported by Turkevich-Frens [43
] utilizing aqueous solutions of AuCl3
and sodium citrate 50.0 g·L−1
. Using Beer’s law, the concentration of the citrate-capped AuNPs solution was estimated to be 5.17 × 10−8
(absorption at 520 nm and an extinction coefficient of 3.67 × 10−8
for particles of 15 nm diameter were used for the calculations [45
]). The as-prepared AuNPs solution was diluted into solutions of different concentrations for subsequent characterization, i.e., a 1.29 × 10−9
solution in Milli-Q water for the registration of the UV-Vis spectrum between 400 and 800 nm, and an 8.61 × 10−9
solution for DLS measurements.
The stock solution of HgCl2 (2.173 g·L−1) was prepared by dissolving 217.3 mg of HgCl2 in 100.00 mL of Milli-Q water containing a few drops of concentrated HNO3 and further diluted whenever necessary.
In order to perform the interference study with metal cations, several 2.00 μg·L−1 stock solutions were prepared by weighing and dissolving appropriate quantities of each salt in Milli-Q water: Li2CO3, NaCl, KCl, AgNO3, MgCl2, CaCl2, BaCl2, FeSO4, CuCl2, ZnSO4, Cd(CH3COO)2, CoSO4·7H2O, Pb(NO3)2, Fe2(SO4)3 and Al2(SO4)3. All stock solutions were stored in amber glass bottles at 4 °C.
2.3. Sensor Development and Characterization
In order to qualitatively evaluate the interaction between FC1 and AuNPs in terms of the occurring energy transfer process, the emission spectrum of a FC1 solution 1 μmol·L−1 and the UV spectrum of a AuNPs solution 1.29 × 10−9 mol·L−1, both prepared in water:MeOH (80:20, pH = 7), were recorded and graphically overlapped.
With the purpose of establishing the quantity of AuNPs needed to suppress the signal attributable to FC1, increasing volumes between 5.0 and 140.0 μL of the AuNPs stock solution were added to 2.00 mL of a FC1 solution 1 μmol·L−1 prepared in water:MeOH (80:20, pH = 7), and the emission fluorescence spectrum was recorded after each addition. In order to calculate the energy transfer efficiency, an emission spectrum of a solution having 250.0 μL of the FC1–AuNPs sensor, i.e., 200.0 μL of FC1 10 μmol·L−1 and 50.0 μL of AuNPs stock solution in 2.00 mL of water:MeOH (80:20, pH = 7), was registered and, then, the FC1–AuNPs emission spectrum was compared with the emission spectrum of a FC1 solution 1 μmol·L−1.
Besides, in order to evaluate the influence of the Hg2+ concentration on the sensor response, a solution containing 250.0 μL of the FC1–AuNPs sensor in 2.00 mL of water:MeOH (80:20, pH = 7) was prepared and transferred to a quartz cuvette. Then, 5.0 μL aliquots of Hg2+ solutions of several concentrations (0.4, 4.0, 20.0 and 40.0 mg·L−1) were added to gradually increment the Hg2+ concentration before recording each fluorescence spectrum.
The stability of the sensor response in the presence of Hg2+ was studied by adding 250.0 μL of the FC1–AuNPs sensor into a 2.00 mL volumetric flask and completed to the mark with water:MeOH (80:20, pH = 7). After homogenizing, the solution was spiked with Hg2+ in order to obtain a final Hg2+ concentration of 5.00 μg·L−1 and the emission spectra were registered every 30 min, during 480 min, at the working temperature of 40 °C.
The influence of the organic solvent was evaluated by substituting different volumes of water for MeOH in 2.00 mL solutions having 250.0 μL of the FC1–AuNPs sensor and completed to the mark with a solution having 2.00 μg·L−1 of Hg2+ in order to reach percentages of MeOH ranging from 0% to 20% each 5% intervals. The experiment was performed in triplicate and the fluorescence emission spectra were registered.
An interference study was performed to gather information regarding the selectivity of the FC1–AuNPs sensor towards Hg2+. In order to evaluate different metallic ions, i.e., Li+, Na+, K+, Ag+, Mg2+, Ca2+, Ba2+, Fe2+, Cu2+, Zn2+, Cd2+, Co2+, Pb2+, Fe3+ and Al3+, several 2.00 mL solutions containing 250.0 μL of FC1–AuNPs sensor and completed to the mark with solutions containing 2.00 μg·L−1 of each metallic ion dissolved in water:MeOH (80:20, pH = 7) were prepared and monitored in the established conditions.
It is important to mention that the pH was set to 7 in every experiment, as the sensor is sensitive to pH. This behavior was firstly studied and reported by Yang and coworkers [10
], and further confirmed by Bohoyo and coworkers [18
]. These latter authors informed that pH-controlled emission measurements showed that FC1 responds to Hg2+
in the pH range from 5.5 to 12.0 with the fluorescent intensity varying less than 10%. When the pH value is lower than 5.0, the luminescence intensity of free FC1 increases greatly with decreasing pH values, and then a value of pH = 7.0 was selected as optimum.
All the experiments were carried out in the darkness.
2.4. Method Validation: Limit of Detection (LOD), Limit of Quantitation (LOQ), Linearity, Precision and Accuracy
The proposed method was validated in order to test its performance regarding linearity, precision and accuracy.
In order to calibrate, six Hg2+ solutions having 0.00, 0.50, 1.00, 1.50, 2.00 and 3.00 μg·L−1 in Milli-Q water were prepared in triplicate by adding appropriate amounts of Hg2+ stock solution into 10.00 mL volumetric flasks. Then, 1600.0 μL of each standard was mixed with 400.0 μL of MeOH and the pH was adjusted to 7.00 with NaOH 0.1 mol·L−1. Finally, 250.0 μL of FC1–AuNPs sensor were added to 2.00 mL volumetric flasks, completed to the mark with each corresponding standard solutions and incubated at 40 °C during 7 h before registering the emission spectra.
The method precision was estimated by preparing and analyzing 10 replicates of samples having 2.00 μg·L−1 of Hg2+ in Milli-Q water during three consecutive weeks. Both interassay and intermediate precisions were estimated by performing an ANOVA test.
Finally, with the aim of evaluating accuracy and testing the applicability of the method both in Milli-Q water and tap water from Santa Fe city, appropriate aliquots of Hg2+ stock solution were added to 10.00 mL volumetric flasks in order to have three Hg2+ concentration levels: 1.00, 1.50 and 2.00 μg·L−1. The samples were processed in triplicate according to the proposed method, and before analyzing, they were filtered through a membrane filter of 0.45 μm particle size.