Hg²⁺ ion-imprinted polymers sorbents based on dithizone–Hg²⁺ chelation for mercury speciation analysis in environmental and biological samples

Abstract

For mercury speciation analysis in environmental and biological samples, novel Hg²⁺ ion-imprinted polymers (IIPs) were synthesized by a sol–gel process using the chelating agent dithizone, and then dithizone–Hg²⁺ chelate as a template and 3-aminopropyltriethoxysilane as a functional monomer, followed by solid-phase extraction (SPE) and atomic fluorescence spectroscopy (AFS) detection. The resultant Hg-IIPs offered high binding capacity, fast kinetics, and their adsorption processes followed a Langmuir isotherm and pseudo-second-order kinetic models. The IIPs displayed excellent selectivity toward Hg²⁺ over its organic forms and other metal ions with selectivity factors of 19–34, as well as high anti-interference ability for Hg²⁺ confronting with common coexistent ions. Through 10 adsorption–desorption cycles, the IIPs showed a good reusability with a relative standard deviation within 5%. Moreover, because of the chelation of dithizone, the IIPs could readily discriminate Hg²⁺ from organic mercury. Thus, mercury speciation analysis could be attained by using IIPs-SPE-AFS, presenting high detectability of up to 0.015 μg L⁻¹ for Hg²⁺ and 0.02 μg L⁻¹ for organic mercury. This method was validated by using two certified reference materials with very consistent results. Satisfactory recoveries ranging from 93.0–105.2% were attained for spiked seawater and lake water samples with three concentration levels of Hg²⁺. Furthermore, the analytical results for the spiked mercury species in real biological samples, such as human hair and fish meat, confirmed that the methods are practically applicable to speciation analysis. The IIPs-SPE-AFS demonstrated significant application perspectives for rapid and high-effective cleanup, enrichment and determination of trace mercury species in complicated matrices.

---

1. Introduction

Heavy metals have received wide international concern because of their difficult degradation and high bioaccumulation properties, which results in environmental and ecotoxicological threats.^1–4 Among them, mercury (Hg) has become an attractive research area^1–6 because it has negative impacts on humans usually due to exposure through food, especially fish.^7,8 Its toxicity depends on its chemical forms; organic mercury compounds are generally much more toxic than inorganic mercury due to their strong lipophilic characteristics.^7–10 Hence, speciation analysis of mercury ion (Hg²⁺) and its organic species is being increasingly investigated because the monitoring and remediation of heavy metal pollution has become a crucial global issue. The speciation and determination of mercury species is generally performed using chromatographic techniques, such as gas chromatography (GC), liquid chromatography (LC), ion chromatography (IC) and capillary electrophoresis (CE) for separation,^5,6,8,11 which are often coupled with element selective detectors, e.g. inductively coupled plasma mass spectrometry (ICP-MS),¹² atomic absorption spectroscopy (AAS),¹³ and atomic fluorescence spectroscopy (AFS).^14–16 However, these instrumental techniques are usually complex, costly, and time-consuming. On the other hand, the trace/ultra-trace presence of targeted analytes in complicated matrices has become a bottleneck for actual application analysis.¹⁷ Hence, it is important to develop high-efficiency pretreatment and enrichment techniques/processes. Among them, solid-phase extraction (SPE) with selective sorbents is gaining popularity.^18,19

Ion-imprinted polymers (IIPs), a branch of molecularly imprinted polymers (MIPs), have attracted significant interest as selective SPE sorbents for particular chemical forms of given elements.^4,20–24 In contrast to MIPs, IIPs recognize inorganic ions after imprinting, especially metal ions.^{20,21,24–27} This recognition ability can be explained by the memory effects of polymers toward metal ion interaction with a specific ligand, coordination geometry, metal ion coordination number, charge and size.²⁸ Metal IIPs have attracted increasing attention due to their unique features: first, the coordination reaction is stronger than the hydrogen bond strength, especially in aqueous solution; second, it is quick to achieve the balance of the thermodynamics and kinetics.^29,30 Lately, a number of studies on IIPs and their applications for selective preconcentration and separation of metal ions have been reported.^{20,21,28,30,31} However, IIPs related studies are often limited by the challenges of incomplete ion removal and low selectivity because many metal ions have the same charges, similar ionic radius and properties.^28,32,33 To overcome these limitations effectively, a grafting method on the surface of polymer/silica gel beads and utilizing selective ligands and/or functional monomers for ion imprinting have been developed.^33–35 For example, Dakova et al. prepared Hg(II) layer-coated silica gel particles (Hg(II)-IIP) for the speciation and determination of mercury in wine.³⁴ Yan et al. proposed IIPs mesoporous materials for Zn²⁺ and Cd²⁺ detection, offering an alternative to immobilize small molecules.³⁵ Chen et al. synthesized specific ligand 3-isocyanatopropyltriethoxysilane and a specific functional monomer for specially imprinting Hg²⁺, and attained highly selective preconcentration of Hg²⁺ in water samples.³³

Dithizone, a well-known traditional ligand that acts as a chelating agent for metal ions, has excellent selectivity advantages and it has been increasingly used for mercury.^36–38 For instance, Liu and his coworkers developed a dithizone-functionalized SPE procedure followed by HPLC-ICP-MS for mercury in water samples.³⁷ Talpur et al. used surfactant coated alumina modified by dithizone for the preconcentration and determination of trace amounts of mercury coupled with cold vapor AAS.³⁸ On the other hand, sol–gel technology, an easy-to-handle and ecofriendly strategy, has flourished for the synthesis of IIPs and MIPs.^33,39–41 However, to the best of our knowledge, no work combining IIPs, dithizone–Hg²⁺ chelate and sol–gel for mercury speciation analysis has been reported.

Inspired by these studies, herein, we propose the synthesis of novel Hg²⁺ imprinted IIPs by sol–gel process, using dithizone–Hg²⁺ chelate as template and 3-aminopropyltriethoxysilane as functional monomer, for mercury speciation analysis in environmental and biological samples via selective SPE coupled with AFS detection. The selectivity of Hg-IIPs for Hg²⁺ versus other interfering metal ions (Zn²⁺, Cd²⁺, Pb²⁺) and organic mercury species (MeHg⁺, EtHg⁺) was investigated. Two certified reference materials were used for validation, and the developed method was successfully applied for Hg²⁺ determination in environmental water samples, as well as organic mercury species determination in human hair and fish meat samples with satisfactory results.

2. Experimental

2.1. Reagents and materials

Standard solution of (60 mg L⁻¹) mercury chloride (HgCl₂) and methylmercury (MeHg) chloride, as well as ethylmercury chloride (EtHg) were purchased from the CRM/RM Information Center of China (Beijing, China). Dithizone, 3-aminopropyltriethoxysilane (APTES) and tetraethoxysilicane (TEOS) were obtained from Sigma-Aldrich (Shanghai, China). Other reagents and materials were supplied by Sinopharm Chemical Reagent (Shanghai, China). All the reagents were at least analytical grade and were used directly without further purification, unless otherwise specified. Throughout the experiment aqueous solutions work were prepared using doubly purified deionized (DDI) water, which was produced by a Milli-Q Ultrapure water system with the water outlet operating at 18.2 MΩ (Millipore, MA, USA); all pH measurements were performed using a pHs-3TC digital pH meter equipped with a combined glass–calomel electrode (Shanghai, China).

2.2. Instrumentation

Atomic fluorescence measurements of mercury species were carried out with an AFS-3000 double-channel nondispersive atomic fluorescence spectrometer (Beijing Kechuang Haiguang Instrument Co., China). ICP-MS analyses were performed on a PerkinElmer Elan DRC II (USA). A FT-IR spectrometer (Nicolet iS10, Thermo scientific) was employed to record the infrared spectra of samples using a pressed KBr tablet method. A scanning electron microscope (SEM, Hitachi S-4800, Japan) was employed to investigate the size and morphology of the polymers. Specific surface area of the polymers was determined by Brunauer–Emmett–Teller (BET) analysis by performing nitrogen adsorption experiments, and the measurement was operated on an AUTOSORB 1 (Quantachrome Instruments, Germany). The polymers were degassed in vacuum at 300 °C prior to adsorption measurements. Zeta potential measurements were performed on a Malvern Zetasizer Nano-ZS90 (ZEN3590, UK).

2.3. Preparation of Hg-IIPs

Hg-IIPs based on dithizone–Hg²⁺ interactions were prepared by a sol–gel process similar to a reported procedure³³ with some modifications. Dithizone (256 mg) and HgCl₂ (108 mg) were dissolved in a 100 mL ethanol solution to form a template molecules complex, followed by adding APTES (467 μL) as functional monomers, which were stored at 4 °C in dark for 6 h. Then, TEOS (2 mL), as crosslinking agent, and aqueous ammonia (5 mL, 14%) were added in the mixing solution. The polymerization reaction was undertaken under magnetic stirring at room temperature for 12 h, and the product was further aged by stirring at 60 °C for 6 h to obtain high cross-linking density and interconnected macroporous structures. The resultant IIPs were obtained by centrifugation and then rinsed with anhydrous ethanol for three times to remove the residues. Hg²⁺ was removed from the prepared IIPs by several sequential elution steps with 0.5 mol L⁻¹ HCl under vigorous stirring. The final products were washed with DDI water up to pH 6–7 and were dried under vacuum at 40 °C. For simplicity, the obtained polymers were named Hg-IIPs. For comparison, the non-imprinted polymers (NIPs) as control materials were prepared using the same procedure and conditions, only in the absence of the template molecules.

2.4. Batch procedure

To evaluate the ion recognition properties of the Hg-IIPs, the effect of pH in the solution, static and dynamic adsorption and selectivity and reusability were tested by a batch procedure method. The procedures were carried out as follows. The Hg-IIPs (20 mg) were equilibrated with 10 mL of aqueous solutions containing 20 mg L⁻¹ of Hg²⁺ with pH varying from 2 to 9 (adjusted using 0.1 mol L⁻¹ of phosphate buffer) at room temperature for 12 h. The pH was maintained in a range of ±0.1 units. The tested solutions were centrifuged and the supernatant solutions were collected and filtered through a 0.45 μm membrane. Finally, the filtrate solutions were characterized by AFS. The instrumental response was periodically tested with known metal standard solutions. The experiments were performed in triplicates.

Static adsorption was examined using 10 mL aqueous solutions containing Hg²⁺ at various concentrations, i.e. from 2 to 20 mg L⁻¹, incubating for 12 h under optimal sorption conditions. The binding amount (Q) of Hg²⁺ was calculated by subtracting the free concentrations from the initial concentrations. Meanwhile, dynamic adsorption test was carried out by monitoring the temporal amount of Hg²⁺ in the solutions with the initial concentration of 10 mg L⁻¹. The mixture was continuously shaken in a thermostatically controlled water bath at room temperature for 0 to 160 min. The polymers were removed by centrifugation and the supernatant solutions were collected and determined using AFS, which was similar to that of the static adsorption test.

The selectivity adsorption experiments were performed by using Zn²⁺, Cd²⁺, Pb²⁺, MeHg⁺ and EtHg⁺ at the same concentration as Hg²⁺ (4 mg L⁻¹), which were then treated with the Hg-IIPs and NIPs. The supernatant solutions were collected and determined using ICP-MS. Interference tests were conducted by using MeHg⁺, EtHg⁺, Na⁺, K⁺, Zn²⁺,Cu²⁺, Pb²⁺, Mg²⁺, Ca²⁺, Mg²⁺, Cd²⁺ and Fe³⁺ as co-existent ions at concentrations 10 times higher than of Hg²⁺ (40 mg L⁻¹). The experimental processes were similar to the abovementioned static adsorption test with the same polymer mass of 20 mg. All the tests were performed in triplicates. The average data from triplicate independent results were used for the following discussion.

In addition, the recognition ability of the Hg-IIPs was evaluated by using the imprinting factor (α), which is defined as:

α = Q_IIP/Q_NIP (1)

where, Q_IIP and Q_NIP are the adsorption amounts of template or analogues on IIPs and NIPs at equilibrium, respectively.

The reusability of Hg-IIPs was assessed by the adsorption–desorption experiments. Briefly, the Hg-IIPs were first put in Hg²⁺ solution (4 mg L⁻¹) for saturated adsorption, and then the adsorbed Hg²⁺ ions could be desorbed from the Hg-IIPs by using a 0.5 mol L⁻¹ HCl solution as desorption medium, after vigorously stirring for 2 h at room temperature. The final Hg²⁺ in the aqueous solution was detected by AFS. The adsorption–desorption studies were repeated ten times using the same imprinted polymers.

2.5. Standard solution and practical sample preparation

The stock standard solutions of MeHg⁺ at 1000 mg L⁻¹ were obtained by dissolving appropriate amounts of methylmercury chloride in HPLC grade methanol. Lake water samples were collected in a teflon bottle from an artificial lake located in Laishan District of Yantai City. Surface seawater samples were collected in a teflon bottle from the Fisherman's Wharf of the Yellow Sea located in the coastal zone area of Yantai City. All the water samples were filtered through 0.45 μm PTFE syringe filters (Phenomenex, Los Angeles, CA, USA) to remove suspended particles. The resultant filtrates, which were directly analyzed or alkalified to pH 7.0 using 0.1 mol L⁻¹ NaOH, were kept in a refrigerator at 4 °C prior to use.

Two certified reference materials (CRMs), namely freeze-dried fish powder (BCR-463) and human hair (GBW07601(GSH-1)), purchased from the CRM/RM Information Center of China (Beijing, China), were used to validate the accuracy and precision of the developed method. Moreover, for practical sample tests, human hair samples were randomly collected from healthy volunteers in our lab and fish meat was purchased from a local supermarket. The sampled were treated according to previous reports,⁴² as follows: 1.0 g samples were added to a centrifuge tube containing 1.4 mL of 10 mol L⁻¹ NaOH. The tube was heated at 90–95 °C for 30 min. Then, the solution was cooled to room temperature and its pH was adjusted to 7.0 using concentrated hydrochloric acid, followed by centrifugation to obtain the supernatant. The treatment procedure was repeated three times, and all the three portions of supernatants were merged for AFS detection.

2.6. SPE procedure

A PTFE column (Phenomenex, Los Angeles, CA, USA) was packed with 200 mg of Hg-IIPs using a syringe with 4.0 mm i.d. The column was preconditioned successively with 5 mL of 0.1 mol L⁻¹ HCl, 5 mL of DDI water and 5 mL of blank solutions (standard or sample solutions without spiking). Then, 50 mL of the sample solution containing Hg²⁺ at three concentrations (2, 5 and 10 μg L⁻¹) was passed through the column at a flow rate of 1.0 mL min⁻¹. The column was washed with 10 mL of DDI water and the adsorbed Hg²⁺ was eluted with 2.5 mL of 0.5 mol L⁻¹ HCl. Then, the extractants were detected by AFS. Each concentration was analysed using three replicates.

3. Results and discussion

3.1. Preparation of Hg²⁺ imprinted IIPs

The synthesis and imprinting process of Hg²⁺ imprinted IIPs via the sol–gel process is schematically shown in Fig. 1. First, the Hg²⁺ and dithizone complex was produced based on the strong chelation of dithizone to Hg²⁺ ions, which was used as template molecules. The template molecules could easily pre-polymerize with the functional monomers APTES by non-covalent interactions. APTES, a commonly used silane coupling agent for sol–gel process, was employed due to the hydrophobic interaction. Then, the polymerizable group of APTES copolymerized with TEOS. Therefore, APTES played an important role in the preparation of Hg-IIPs. Finally, Hg²⁺ was removed by elution, leaving behind specific recognition sites on the surface of the polymers, with functional groups in a predetermined orientation and proper size cavities for Hg²⁺ ions, which contributes to the high adsorption efficiency and selectivity for Hg²⁺ ions due to metal–ligand chemistry.⁴³ Meanwhile, the adopted sol–gel process, with distinct characteristics such as easy fabrication, using ecofriendly reaction solvents, and mild polymerization conditions,³³ allowed dithizone–Hg²⁺ chelate, as template, to be readily embedded in the highly cross-linked host structure without the problem of thermal or chemical decomposition.

Fig. 1 Schematic illustration of the preparation and imprinting process of Hg-IIPs.

Moreover, it was found that dithizone chelated Hg²⁺ but not organic mercury compounds (Fig. S1†). Thus, it is possible to realize the speciation analysis of mercury species using the Hg-IIPs based on the specific, strong chelation of dithizone. Meanwhile, AFS can only measure total mercury with the mercury hollow cathode lamp without discriminating between mercury species. Therefore, the prepared Hg-IIPs coupled with AFS would not only provide a superior selectivity for Hg²⁺ but also a possible speciation analysis.

3.2. Characterization of the Hg-IIPs

To confirm the presence of dithizone in the silica gel sorbents and its chelation for imprinting, the IR spectra of IIPs, Hg-contained IIPs and NIPs were measured. As shown in Fig. 2 (upper), the wide and strong absorption band at around 1089 and 945 cm⁻¹ could be attributed to Si–O–Si and Si–O–H stretching vibrations, respectively, and the absorption peaks at around 799 and 467 cm⁻¹ belonged to Si–O vibrations, indicating the presence of silica matrices in the three materials. The strong and broad absorption band observed at 3425 cm⁻¹ could be assigned to the N–H stretching vibration of APTES. Compared with NIPs, the characteristic absorption peak at 1625 cm⁻¹ (Fig. 2a), which most probably represents the stretching vibrations of N=N double bonds of the pyridine ring in dithizone, became significantly weak in the IIPs (Fig. 2b) and Hg-contained IIPs (Fig. 2c). Interestingly, a new peak appeared at 1607 cm⁻¹ in Fig. 2c, which is just the characteristic peak of Hg–N bond. These observations demonstrated the coordination of dithizone through the N atom from its fragment, indicating that the metal–ligand bond was formed between Hg²⁺ and the chelating groups of IIPs. The FT-IR results confirmed that the Hg-IIPs were successfully prepared based on the metal–ligand chemistry by sol–gel polymerization.

Fig. 2 (Upper) FT-IR spectra of (a) Hg-NIPs, (b) Hg-IIPs and (c) Hg contained Hg-IIPs. (Below) SEM images of Hg-IIPs.

SEM images were taken to characterize the surface morphologies of the Hg-IIPs. As observed, the IIPs displayed a dendritic structure and appeared to be covered with particles in several areas (Fig. 2, below). The particles were likely to be aggregated silicon particles because of the use of APTES in the sol–gel process. Because the imprinting polymerization process occurred, the polymers demonstrated a dendritic structure. However, in the absence of imprinting process, silicon particles aggregation materials could be easily formed. Thus, herein, the dendritic-structured IIPs were reasonable with partial coverage of aggregated silicon particles. In addition, by BET analysis, a specific surface area of 43.27 m² g⁻¹ was attained for Hg-IIPs, which is higher than that of the corresponding NIPs (22.92 m² g⁻¹). Fig. 3A shows N₂ adsorption–desorption isotherms of the Hg-IIPs and the corresponding NIPs. As seen, when the relative pressure P/P₀ was less than 0.8, the slope of the curves was small, indicating that little amount of small pores were present on the surface of IIPs. However, when the relative pressure P/P₀ was higher than 0.8, the slope of the curves significantly increased. As seen in the figure, the desorption curve was closer but leveled slightly above the adsorption curve, which could be the evidence of a small quantity of micropores on the IIPs.⁴⁴ As shown in Fig. 3B, the average pore size for IIPs and NIPs was 9.42 and 5.62 nm, respectively. The narrow pore diameter distribution and the low average pore diameter suggested that the size of the cavities formed in the IIPs played a key role in binding capacity. Related morphological structure parameters of IIPs and NIPs are listed in Table S1.† The results indicate that the IIPs have uniform and regular structure. The large cumulative pore volume was very likely because of the microspores on the surface of IIPs, which revealed that the Hg²⁺ ions were almost completely eluted and removed, and thereby leading to considerable amounts of imprinting cavity sites.

Fig. 3 (A) N₂ adsorption–desorption isotherms and (B) pore size distribution of IIPs and NIPs.

3.3. Binding properties of the IIPs for Hg²⁺

Solution acidity plays an important role in the adsorption capacity because it generally affects the complexing reaction between metal ions and ligands. To evaluate the effect of varying pH on Hg²⁺ adsorption, a batch procedure was used. As seen in Fig. 4, the adsorption capacity was very low at pH values below 4.0. The N atoms of dithizone are highly protonated at low pH, and the electronic-supply ability of N atoms significantly weakens, thereby reducing the chelating effect of dithizone for Hg²⁺ ions,^31,32 which in turn reduces the adsorption capacity of Hg-IIPs. The adsorption capacity increased rapidly along with an increase in pH to 4–6 because the protonation of N atoms of dithizone became weak and enhanced the adsorption capacity of dithizone for Hg²⁺. Above pH 6.0, the increase was relatively slow, and adsorption capacity approached a maximum at pH 8.0. At pH > 9, the formation and precipitation of metal hydroxide, as well as quick decomposition, easily occurs. At the same time, the IIPs would become electronegative, enabling them to adsorb various other metal ions to a certain extent, thus causing a loss of selectivity. In addition, as can be seen from the inset of Fig. 4, Zeta potential of Hg-IIPs varied in different pH values with an excellent linearity in the range of pH 2–9. Thus, the adsorption capacity of IIPs for Hg²⁺ was dependent on pH. The optimum pH for the adsorption of Hg²⁺ from aqueous solutions ranged from 7.0 to 8.0. Within this range, neither the precipitation of the metal hydroxide nor the protonation of the N atoms were obvious. Considering their possible applications in physiological conditions, pH 7.0 was selected to be optimum for further experiments.

Fig. 4 Effect of pH on the adsorption of Hg²⁺ for Hg-IIPs. (Inset) Effect of pH on the Zeta potential of Hg-IIPs. Experimental conditions: IIPs, 20 mg; C₀, 10 mg L⁻¹; V, 10 mL; adsorption time, 12 h.

Then, static adsorption experiments were investigated to evaluate the adsorption capacities of Hg-IIPs for Hg²⁺ within 2–20 mg L⁻¹ at pH 7.0. As shown in Fig. 5A, the amounts of Hg²⁺ adsorbed per unit mass of IIPs increased along with an increase in the initial concentrations of Hg²⁺, and leveled off after 16 mg L⁻¹. Excitingly, the experimentally obtained maximum adsorption capacity of Hg-IIPs was about 10 times than that of NIPs, presenting the imprinting factor α of 9.58. This result suggests that the IIPs possess high imprinting effect due to the high binding affinity for Hg²⁺. Furthermore, the obtained static capacity data for IIPs were fitted using Langmuir, Freundlich and Langmuir–Freundlich isotherm models. As shown in Fig. 5B, the Langmuir isotherm model yielded the best fit, showing the highest correlation coefficient of R² = 0.998 (Table S2†). Therefore, the data obtained from the equilibrium studies for the adsorption of Hg²⁺ ions onto Hg-IIPs may follow the Langmuir model, which assumes that adsorption occurs at specific homogeneous sites on the adsorbent and can be successfully applied to monolayer adsorption processes.⁴⁵ This indicates that Hg²⁺ ions are adsorbed as a monolayer onto the surface of Hg-IIPs. In addition, the Hg-IIPs present the highest concentration of binding sites per gram of polymers and the largest median binding affinity (Table S2†), further confirming their excellent imprinting effect due to a number of specific binding sites. Dynamic binding experiments were carried out to evaluate the binding rate and ion transfer properties of Hg²⁺ onto the IIPs. As shown in Fig. S2A,† 80 min were required to reach an adsorption equilibrium for Hg-IIPs. In the same adsorption time, the binding capacity of Hg-IIPs was remarkably higher than that of NIPs, indicating that rebinding via metal–ligand bond favors ion transfer and binding capacity. Furthermore, the dynamic binding was investigated by different models, including pseudo-first-order, pseudo-second-order, Elovich and intra-particle diffusion.⁴⁶ Showing the highest correlation coefficient of R² = 0.999 (Table S3†), the pseudo-second-order model provided the most suitable correlation for the Hg-IIPs adsorption, which can be expressed as follows:

where Q_t is the instantaneous binding capacity at time t, Q_e is the equilibrium binding capacity, and k₂ is the rate constant. The obtained Q_e of 3.74 μmol g⁻¹ calculated from the model was in good agreement with the Q_e of 3.68 μmol g⁻¹ obtained from experimental results. In addition, as shown in Fig. S2B,† the curve showing the entire time period was found to be the best predicting the kinetic process. Therefore, it can be deduced that adsorption follows the pseudo-second-order kinetics model. Furthermore, it could be expected that the rate-limiting step might be chemisorption involving valency forces through the sharing or exchange of electrons between sorbent and sorbate.⁴⁷ Therefore, chemisorption could be the rate-limiting step in the adsorption process of IIPs for Hg²⁺, further confirming that the Hg-IIPs were prepared based on metal–ligand chemistry.

Fig. 5 (A) Static adsorption isotherm curves of IIPs and NIPs for Hg²⁺ in aqueous solutions, and (B) comparison of Langmuir, Freundlich and Langmuir–Freundlich isotherm models for Hg²⁺ adsorption process onto the IIPs. Experimental conditions: V, 10 mL; polymer, 20 mg; adsorption time, 12 h; room temperature.

3.4. Ionic selectivity and recognition reliability of the IIPs

High ionic selectivity and recognition reliability are recognized as the chief requirements and features of IIPs. Competitive adsorption from binary mixtures, including Hg²⁺/MeHg⁺, Hg²⁺/EtHg⁺, Hg²⁺/Zn²⁺, Hg²⁺/Cd²⁺ and Hg²⁺/Pb²⁺, was investigated in selective binding experiments because these ions often coexist in practical samples. As can be seen in Table 1, the competitive adsorption capacity of Hg-IIPs for Hg²⁺ is considerably higher than that of NIPs. The selective coefficients of Hg-IIPs for binary mixtures were obtained in the range of 19–34, which is likely due to the highly selective Hg²⁺–dithizone interactions. Interestingly, this implied that the determination of organic mercury can be realized using the IIPs by subtracting the amounts of Hg²⁺ from total mercury.

Table 1 Selectivity of the prepared Hg-IIPs and NIPs

Metal ions	Hg-IIPs		NIPs
	D^a (mL g^−1)	S^b	D (mL g^−1)	S
a Distribution ratio, D = Q/Ce.b Selective coefficient, S = Dtemplate/Dcompetetive ion. Dtemplate and Dcompetetive ion are distribution ratio of the template and the competitive ions on IIPs or NIPs, respectively.
Hg^2+	452.8	—	47.3	—
MeHg^+	23.8	19.06	28.1	1.67
EtHg^+	21.1	21.56	26.2	1.82
Zn^2+	13.5	33.54	15.3	3.10
Cd^2+	16.9	26.63	21.8	2.17
Pb^2+	20.5	22.09	23.3	2.03

---

To further evaluate the anti-interference ability and reliability of the IIPs for Hg²⁺, the mixture solutions containing Hg²⁺ and various possible interfering ions were examined with concentrations of interfering ion 10 times higher than Hg²⁺ in every case. As shown in Fig. 6, those ions did not cause significant interference without remarkable reduction in the capacity for Hg²⁺ determination after the IIPs preconcentration procedure. It is possible to estimate that only <10% of the binding sites were taken over by the interfering ions, even when their concentration was 10 times higher. All the results indicate that the Hg-IIPs are highly selective and reliable for Hg²⁺ recognition.

Fig. 6 Adsorption capacities for Hg²⁺ in the presence of 4 mg L⁻¹ Hg²⁺ and 40 mg L⁻¹ of other metal ions, including MeHg⁺, EtHg⁺, Na⁺, K⁺, Zn²⁺, Cu²⁺, Pb²⁺, Mg²⁺, Ca²⁺, Mg²⁺, Cd²⁺ and Fe³⁺. Experimental conditions: IIPs, 20 mg; V, 10 mL; pH, 7.0.

3.5. Reusability of the IIPs

As an important index, the reusability of the IIPs is also investigated because of their pivotal role in cost-saving. Desorption of Hg²⁺ from the IIPs were studied by using 0.5 mol L⁻¹ of HCl, and 0.5 mol L⁻¹ of HCl plus 0.1 mol L⁻¹ thiourea. It was found that the rebinding capacity was significantly reduced when using the latter as the desorption medium, which was very likely due to the disruption of the non-covalent interaction between functional monomers and dithizone, and thereby the complex of dithizone and Hg²⁺ was desorbed. In contrast, after treatment with 0.5 mol L⁻¹ of HCl, the IIPs displayed high rebinding capacity without obvious decrease in activity. It could be inferred that the chelation of dithizone for Hg²⁺ was destroyed and Hg²⁺ ions were subsequently released into the desorption medium, which satisfied the requirements of IIPs. Therefore, 0.5 mol L⁻¹ of HCl was utilized as an ideal desorption medium for the repeated 10 adsorption–desorption cycles. Finally, the Hg-IIPs still presented high rebinding capacities for Hg²⁺ with only a slight decrease within 5%. Consequently, the Hg-IIPs present excellent reusability and regeneration ability, which favors actual applications.

3.6. Analytical performances for Hg²⁺ and organic mercury determination by using IIPs-SPE

Based on the abovementioned results, the Hg-IIPs could be employed as SPE sorbents for the preconcentration of Hg²⁺. The analytical performances of Hg-IIPs-SPE coupled with AFS for Hg²⁺ and total Hg determination were investigated. Under the optimized SPE conditions, an excellent linearity from 0.05 to 15 μg L⁻¹ for Hg²⁺ was observed, and the relative standard deviation (RSD) at 2 μg L⁻¹ was 5.2%. The limit of detection (LOD) based on three times the standard deviation and for 10 replicate measurements of a blank solution was 0.015 μg L⁻¹, which is significantly lower than the mandated upper limit of 2 μg L⁻¹ for Hg²⁺ in drinking water by the United States Environmental Protection Agency (USEPA). This indicates that the Hg-IIPs-SPE-AFS are potentially applicable for the sensitive and accurate determination of Hg²⁺ in drinking water.

Using the developed method, the organic mercury value could be obtained by taking the total Hg value and subtracting the Hg²⁺ value, thereby the organic mercury species could be determined. The LOD achieved for MeHg⁺ was 0.02 μg L⁻¹, higher than the permitted maximum concentration level (MCL) of 0.001 μg L⁻¹ by China. Nevertheless, the method sensitivity might contribute to mercury speciation analysis in water samples, especially industrial and sanitary wastewater. The LOD for EtHg⁺ was 0.02 μg L⁻¹, lower than the MCL of 0.1 μg L⁻¹ formulated by China. In addition, the RSDs at 2 μg L⁻¹ were in a range of 6–11%. Hence, the Hg-IIPs-SPE-AFS could recognize and detect organic mercury. Thus, this method clearly proves to be potentially feasible to quantitatively analyze organic mercury in real water samples.

To validate this established method, the Hg-IIP-SPE procedure was applied to the CRMs, BCR-463 (fish meet) and GBW07601 (human hair). As shown in Table 2, the determined values for MeHg and total Hg were 2.95 ± 0.15 μg g⁻¹ and 3.21 ± 0.15 μg g⁻¹, respectively; in BCR-463, the total Hg was 0.38 ± 0.07 μg g⁻¹ in GBW07601, which agrees well with the certified values. The t-test confirmed that there was no significant difference between the developed method and the ICP-MS determination at 95% confidence level. Therefore, the Hg-IIPs-SPE-AFS was proven to be accurate and reliable for mercury speciation in real solid and semi-solid samples.

Table 2 Determination of MeHg⁺ and total Hg in certified reference materials (CRMs)^a

CRMs	Certified (μg g^−1)		Determined (μg g^−1)^b		Determined (μg g^−1)^c
	MeHg^+	Total Hg	MeHg^+	Total Hg	MeHg^+	Total Hg
a The comparison of the developed Hg-IIPs-SPE-AFS and ICP-MS results was made by using the t-test at 95% confidence level.b Results obtained from this method using AFS.c Results obtained by ICP-MS.d Average value of three determinations ± standard deviation.e Not detected.
BCR-463	2.85 ± 0.16	3.04 ± 0.16	2.95 ± 015^d	3.21 ± 0.15	2.89 ± 0.14	3.15 ± 0.14
GBW07601	—	0.36 ± 0.08	ND^e	0.38 ± 0.07	ND	0.36 ± 0.06

---

3.7. Applications of the IIPs-SPE to environmental and biological samples

To evaluate the practical applicability of the developed method for the mercury speciation analysis, the IIPs-SPE was applied in environmental water samples and biological samples. As shown in Table 3, recoveries were in the range of 93.0–105.2% with RSD of 2.9–5.3%, suggesting that the obtained IIPs, via the chelating interaction of dithizone and Hg²⁺ accompanied with dual functional monomers, were ideal candidates for SPE sorbents and were applicable for the effective enrichment and quantitative determination of trace Hg²⁺ in real water samples. At the same time, it suggested that the matrix effects were sharply reduced after the IIPs-SPE procedure. In addition, the t-test confirmed that the determined data was in good consistency with that obtained by ICP-MS (Table 3). Hence, the method was feasible for real sample analysis with a high accuracy and good reliability. Moreover, the endogenous contents of Hg²⁺ were detected at 0.21 and 0.31 μg L⁻¹ in the tested seawater and lake water samples, respectively. This value is significantly lower than the permitted concentration of Hg²⁺ in drinking water by USEPA (2.0 μg L⁻¹). Therefore, the validated IIPs-SPE coupled to AFS method might be an ideal alternative to the simultaneous enrichment, separation and determination of Hg²⁺ in complicated water samples for environmental monitoring and remediation.

Table 3 Analytical results of the determination of seawater and lake water samples spiked with Hg²⁺

Sample	Added (μg L^−1)	Found (μg L^−1)	Recovery^a ± RSD^b (%)	ICP-MS results (μg L^−1)
a Average value from five individual experiments.b Relative standard deviation.
Seawater	0	0.21	—	0.23
	2	2.15	97.0 ± 5.3	2.05
	5	5.47	105.2 ± 4.7	5.07
	10	9.76	95.5 ± 3.2	9.84
Lake water	0	0.31	—	0.29
	2	2.24	95.5 ± 4.5	2.06
	5	4.96	93.0 ± 2.9	5.09
	10	10.74	104.4 ± 3.6	10.13

---

In addition, the practicality of the IIPs-SPE for organic mercury in biological samples was investigated. As observed in Table 4, recoveries for organic mercury ranged from 97.8–104.5% with RSDs of 7.9–8.1% for human hair and 96.2–97.8% with RSDs of 8.6–9.3% for fish meat samples. Therefore, the IIPs-SPE-AFS was demonstrated to be capable of simply and sensitively identifying and detecting organic mercury species.

Table 4 Analytical results of the determination of Hg²⁺, MeHg⁺ and EtHg⁺ in biological samples

Sample	Added (μg L^−1)			Found (μg L^−1)		Recovery ± RSD (%)
	MeHg^+	EtHg^+	Hg^2+	Organic mercury^a	Total Hg	Organic mercury	Total Hg
a Organic mercury value was obtained subtracting the Hg^2+ value to the total Hg value.b Average value from three individual experiments.
Human hair	0	0	10	ND	9.85	—	98.5 ± 2.4
	10	0	10	10.45^b	21.23	104.5 ± 8.1	106.2 ± 4.2
	0	10	10	9.78	18.52	97.8 ± 7.9	92.6 ± 3.4
Fish meat	0	0	10	ND	10.24	—	102.4 ± 4.1
	10	0	10	9.62	18.76	96.2 ± 9.3	93.8 ± 5.3
	0	10	10	9.78	19.75	97.8 ± 8.6	98.8 ± 3.7

---

3.8. Comparison with other methods

Performance of this developed chelate imprinting strategy toward mercury species was compared with that of some reported IIPs-SPE based methods. As shown in Table S4,† as a new material, the prepared IIPs are ideal sorbent candidates in SPE for the preconcentration of Hg²⁺. The attained LOD is lower or comparable to the reported values.^33,34,48 Species including Hg²⁺, MeHg⁺ and EtHg⁺ are analyzed. The presented IIPs are simpler and easier to prepare using a cheaper classic ligand (dithizone) than 1-pyrrolidinedithiocarboxylic acid³⁴ and diazoaminobenzene⁴⁸ by a milder sol–gel process, as well as avoiding synthesizing procedures.³³ Moreover, they are applicable in more samples, including environmental and biological samples.

4. Conclusions

In summary, the novel Hg²⁺ IIPs based on dithizone–Hg²⁺ chelation were successfully prepared and applied to environmental and biological samples for the preconcentration of Hg²⁺ and mercury speciation analysis by a facile sol–gel process. Using dithizone–Hg²⁺ complex as a template, the easily obtained IIPs displayed fast adsorption kinetics and high binding capacity for Hg²⁺, excellent selectivity towards Hg²⁺ over organic mercury and other metal ions, as well as high anti-interference ability for Hg²⁺ recognition and adsorption. Coupled to AFS detection, a high detection ability of up to 0.015 μg L⁻¹ for Hg²⁺ and 0.02 μg L⁻¹ for organic mercury could be achieved, comparable to or even surpassing that afforded by traditional chromatography-mass spectroscopy methods. The method accuracy was validated with consistent results for CRMs. In addition, the applicability was demonstrated for the analysis of Hg²⁺ in seawater and lake water samples, together with organic mercury species in human hair and fish meat samples. Therefore, the IIPs-SPE-AFS proved to be a very promising, sensitive, simple and ecofriendly alternative to conventional methods for mercury speciation analysis in aqueous, solid and semi-solid environmental and biological samples.

On the other hand, we plan to develop a simple general applicable IIPs based platform for heavy metal ions and species by means of the model of Hg²⁺ imprinting. For example, MeHg⁺ can also be imprinted by using specific ligands and/or functional monomers. With a suitable choice and reasonable utilization of chelation, traditional ligands and functional monomers (especially those commercially available), and by smartly devising and synthesizing new ligands and functional monomers, the IIPs-based heavy metal detection/removal platform can be developed for the selective, sensitive routine monitoring/remediation of environmental and biological samples. Such an excellent platform has significant potential for imprinting various heavy metals, which certainly will push the development of heavy metal speciation analysis and enrich the research of molecular imprinting technologies. We believe that it will provide new opportunities to explore novel IIPs/MIPs for potential utilizations. Our efforts in this regards are currently undergoing.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (21275158, 21105117, 21107057, 21275068), the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry, and the Innovation Projects of the Chinese Academy of Sciences (KZCX2-EW-206).

Notes and references

1. Q. H. Wu, N. F. Y. Tam, J. Y. S. Leung, X. Z. Zhou, J. Fu, B. Yao, X. X. Huang and L. H. Xia, Ecotoxicol. Environ. Saf., 2014, 104, 143.
2. G. Aragay, J. Pons and A. Merkoc, Chem. Rev., 2011, 111, 3433.
3. A. R. Timerbaev, Chem. Rev., 2013, 113, 778.
4. Y. G. Yin, J. F. Liu and G. B. Jiang, Chin. Sci. Bull., 2013, 58, 150.
5. Z. F. Liu, Z. L. Zhu, H. T. Zheng and S. H. Hu, Anal. Chem., 2012, 84, 10170.
6. J. H. Li, W. H. Lu, J. P. Ma and L. X. Chen, Microchim. Acta, 2011, 175, 301.
7. L. H. Reyes, G. M. Mizanur Rahman, T. Fahrenholz and H. M. Kingston, Anal. Bioanal. Chem., 2008, 390, 2123.
8. F. A. Duarte, B. M. Soares, A. A. Vieira, E. R. Pereira, J. V. Maciel, S. S. Caldas and E. G. Primel, Anal. Chem., 2013, 85, 5015.
9. J. B. Shi, L. N. Liang, G. B. Jiang and X. L. Jin, Environ. Int., 2005, 31, 357.
10. D. Karunasagar, M. V. Balarama Krishna, S. V. Rao and J. Arunachalam, J. Hazard. Mater., 2005, 118, 133.
11. F. F. Yang, J. H. Li, W. H. Lu, Y. Y. Wen, X. Q. Cai, J. M. You, J. P. Ma, Y. J. Ding and L. X. Chen, Electrophoresis, 2014, 35, 474.
12. D. Yan, L. M. Yang and Q. Q. Wang, Anal. Chem., 2008, 80, 6104.
13. Y. Li, Y. Jiang and X. P. Yan, Anal. Chem., 2006, 78, 6115.
14. T. Yordanova, I. Dakova, K. Balashev and I. Karadjova, Microchem. J., 2014, 113, 42.
15. J. Margetínová, P. H. Pelcová and V. Kubáñ, Anal. Chim. Acta, 2008, 615, 115.
16. W. B. Zhang, X. A. Yang, Y. P. Dong and J. J. Xue, Anal. Chem., 2012, 84, 9199.
17. Y. Y. Wen, J. H. Li, J. P. Ma and L. X. Chen, Electrophoresis, 2012, 33, 2933.
18. W. A. W. Ibrahim, L. I. Abd Ali, A. Sulaiman, M. M. Sanagi and H. Y. Aboul-Enein, Anal. Chem., 2014, 44, 233.
19. Y. Y. Wen, L. Chen, J. H. Li, D. Y. Liu and L. X. Chen, TrAC, Trends Anal. Chem., 2014, 59, 26.
20. L. X. Chen, S. F. Xu and J. H. Li, Chem. Soc. Rev., 2011, 40, 2922.
21. X. Q. Cai, J. H. Li, Z. Zhang, F. F. Yang, R. C. Dong and L. X. Chen, ACS Appl. Mater. Interfaces, 2014, 6, 305.
22. N. T. Hoai, D. K. Yoo and D. Kim, J. Hazard. Mater., 2010, 173, 462.
23. C. R. T. Tarley, F. N. Andradea, F. M. Oliveirac, M. Z. Corazzac, L. F. M. Azevedoa and M. G. Segatelli, Anal. Chim. Acta, 2011, 703, 145.
24. S. L. Wei, Y. Liu, M. D. Shao, L. Liu, H. W. Wang and Y. Q. Liu, RSC Adv., 2014, 4, 29715.
25. N. Zhang and B. Hu, Anal. Chim. Acta, 2012, 723, 54.
26. A. Bhaskarapillai and S. V. Narasimhan, RSC Adv., 2013, 3, 13178.
27. B. B. Prasad, D. Jauhari and A. Verma, Talanta, 2014, 20, 398.
28. C. Branger, W. Meouche and A. Margaillan, React. Funct. Polym., 2013, 73, 859.
29. C. R. Preetha, J. M. G. Ladis and T. P. Rao, Environ. Sci. Technol., 2006, 40, 3070.
30. T. Alizadeh, M. R. Ganjali and M. Zare, Anal. Chim. Acta, 2011, 689, 52.
31. B. BaliPrasad, D. Jauhari and A. A. Verma, Talanta, 2014, 120, 398.
32. S. Clemens, M. Monperrus, O. F. X. Donard, D. Amouroux and T. Guerin, Talanta, 2012, 89, 12.
33. S. F. Xu, L. X. Chen, J. H. Li, Y. F. Guan and H. Z. Lu, J. Hazard. Mater., 2012, 237–238, 347.
34. I. Dakova, T. Yordanova and I. Karadjova, J. Hazard. Mater., 2012, 231, 49.
35. J. Tan, H. F. Wang and X. P. Yan, Biosens. Bioelectron., 2009, 24, 3316.
36. K. Leopold, M. Foulkes and P. Worsfold, Anal. Chim. Acta, 2010, 663, 127.
37. Y. G. Yin, M. Chen, J. F. Peng, J. F. Liu and G. B. Jiang, Talanta, 2010, 81, 1788.
38. Z. A. Chandio, F. N. Talpur, H. Khan, H. I. Afridi, G. Q. Khaskheli and M. A. Mughal, RSC Adv., 2014, 4, 3326.
39. X. L. Song, J. H. Li, S. F. Xu, R. J. Ying, J. P. Ma, C. Y. Liao, D. Y. Liu, J. B. Yu and L. X. Chen, Talanta, 2012, 99, 75.
40. H. He, D. L. Xiao, J. He, H. Li, H. He, H. Dai and J. Peng, Analyst, 2014, 139, 2459.
41. J. E. Lofgreen and G. A. Ozin, Chem. Soc. Rev., 2014, 43, 911.
42. Y. W. Liu, Y. H. Zai, X. J. Chang, Y. Guo, S. M. Meng and F. Feng, Anal. Chim. Acta, 2006, 575, 159.
43. J. J. Becker and M. R. Gagnea, Acc. Chem. Res., 2004, 37, 798.
44. X. Y. Wang, Q. Kang, D. Z. Shen, Z. Zhang, J. H. Li and L. X. Chen, Talanta, 2014, 124, 7.
45. C. Herdes and L. Sarkisov, Langmuir, 2009, 25, 5352.
46. J. H. Zhu, S. Y. Wei, H. B. Gu, S. B. Rapole, Q. Wang, Z. P. Luo, N. Haldolaarachchige, D. P. Young and Z. H. Guo, Environ. Sci. Technol., 2012, 46, 977.
47. Y. S. Ho and G. McKay, Process Biochem., 1999, 34, 451.
48. Y. W. Liu, X. J. Chang, D. Yang, Y. Guo and S. M. Meng, Anal. Chim. Acta, 2005, 538, 85.

---

Footnote

† Electronic supplementary information (ESI) available: The UV-vis spectra of dithizone alone and dithizone with mercury species, the adsorption kinetics curves of Hg-IIPs/NIPs for Hg²⁺ and kinetics models data, the specific surface area and other related data of Hg-IIPs/NIPs obtained by BET, the isotherm model parameters for Hg-IIPs/NIPs, and the parameters obtained of Hg²⁺ adsorption towards Hg-IIP from four kinetic models. See DOI: 10.1039/c4ra08163c

---