A green photometric method for determination of mercuric ions in saline samples by a single-drop microextraction technique

Abstract

This study describes a rapid, simple and sensitive spectrophotometric method for determination of Hg(II) ions in saline samples by a single-drop microextraction (SDME) technique. The method is based on the extraction of dithizone (DTZ) complex of mercury(II) into an undecanol droplet which serves as the organic phase, and then the absorbance of the colored complex is measured at 490 nm by using a microvolume quartz cuvette. This procedure provides a simple, rapid, cost-efficient and, most of all, a green method for detecting mercuric ions by minimizing the organic solvent consumption. A variety of parameters affecting the signal, such as pH, DTZ concentration, sample and extraction solvent volume, extraction time and temperature and salt effects were optimized. Under optimized conditions the linear range was found between 3.2 × 10⁻⁸ and 5.0 × 10⁻⁷ mol L⁻¹ (6.4–100.8 μg L⁻¹) and the detection limit was calculated as 9.6 × 10⁻⁹ mol L⁻¹ (1.9 μg mL⁻¹) attained by a high enrichment factor of 203. The performance and accuracy of the method were compared with those of atomic fluorescence spectrometry. Validation of the proposed method was performed for determination of mercury in saline samples including sea water, mineral water, thermal spring water, and table and rock salt samples, which is difficult to analyze by conventional methods.

---

Introduction

Mercury is one of the most toxic elements for all living organisms and its monitoring in biological, environmental and industrial samples is extremely important.¹ Both organic and inorganic forms of mercury are highly toxic to humans, but most people are exposed to organic mercury through the consumption of fish and shellfish which can accumulate in the brain damaging the central nervous system.² Inorganic mercury in food is not as easily absorbed by the human body, but elemental mercury vapor can cause acute pneumonia in the case of inhalation in large concentrations. Divalent mercury can cause kidney damage and leukemia.^3,4

Inorganic mercury occurs naturally in the environment and is also released as a consequence of human activities. Surface waters in the vicinity of industrial areas are important indicators for mercury pollution.^5,6 The amount of total mercury is usually in the range of 0.5–5 ng mL⁻¹ in natural and fresh waters and increases to several μg mL⁻¹ levels in contaminated waters.⁷ For drinking water, the EPA recommends a limit of 2 ng mL⁻¹.⁸ Therefore, analytical monitoring of trace amounts of mercury in surface waters is of great significance for public health safety and environmental pollution control.

Numerous analytical techniques have been used for the determination of trace levels of mercury in water samples, including atomic absorption spectrometry,⁹ cold vapor atomic absorption spectrometry,¹⁰ atomic fluorescence spectrometry (AFS),^11,12 inductively coupled plasma with optical emission spectrometry¹³ and mass spectrometry¹⁴ along with X-ray fluorescence spectrometry.¹⁵

Although these methods provide sensitive tools for mercury determination in environmental samples, their application to saline samples is cumbersome due to the impairment of direct analysis resulting from the high ionic strength of such samples.¹⁶ For this purpose, in situ solid-phase preconcentration methods were utilized for avoiding interferences and improving the sensitivity of the method.^7,17

Electrochemical methods provide a more economical and yet sensitive tool for mercury determination in saline samples. A study carried out in the authors' lab has revealed that the mercury content of table salt samples can be determined at a gold film electrode by anodic stripping voltammetry.¹⁸ The interference arising from high chloride content was eliminated by applying a medium-exchange after the deposition step.

Spectrophotometric methods, on the other hand, are some of the most common methods due to their simplicity and inexpensive instrumentation with reasonable sensitivity for determination of mercuric ions.^19–23 However, mercuric ions usually exist in the environment at trace levels in rather complicated matrices, and thus extraction and pre-concentration procedures are essential in photometric detection.⁷ After complexation of mercury(II) with different types of reagents, the complexes are extracted in chloroform, toluene, xylene or carbon tetrachloride prior to the measurement.²⁴ However, the use of these solvents is to be avoided due to their toxicity, and they display some drawbacks such as large consumption of reagent, high volatility and low enrichment factor. As a result greener and faster methods for monitoring the trace levels of mercury are being sought.

Among the currently available extraction and preconcentration methods, miniaturized preconcentration methods have been attempted for liquid–liquid extraction systems based on single-drop microextraction (SDME),^25,26 solidified floating organic drop microextraction (SFODME),²⁷ and dispersive liquid–liquid microextraction (DLLME)²⁸ for the determination of mercuric ions in different samples. These techniques have recently aroused a great interest, due to their favorable characteristics of simplicity, rapidity, cost-effectiveness and minimized toxic and flammable organic solvent consumption.²⁹ High enrichment factor can be easily obtained by SDME since microliter volumes of liquid drops are used.³⁰ Previously, organic solvents like carbon tetrachloride, cyclohexane, toluene, m-xylene, and n-octane have been used in SDME as the extracting phases.^30–32 However, the use of such solvents in SDME is limited due to their high rate of dissolution and evaporation in addition to emulsion formation during extraction.³³ Recently, low-volatility compounds such as ionic liquids^34–36 have been used instead of toxic, flammable and water-miscible organic solvents. As an alternative to ionic liquids, a greener solvent, undecanol, was used as the extracting phase in a recent study.²⁷ Mercuric ions were complexed with diethyldithiocarbamate, and then extracted into fine droplets of 1-undecanol. After cooling in an ice bath, the solidified microdrop was transferred for determination by cold vapor AFS.

The present study includes the early findings of a novel method which utilizes a microdroplet of undecanol as the extracting solvent for the first time for Hg(II) ions complexed with dithizone (DTZ). The absorbance of the colored complex trapped in the droplet is then measured by using a microvolume quartz cuvette. The optimization of experimental conditions, possible interferences, analytical characteristics and method validation were investigated in detail. The applicability of the method for saline samples was also investigated.

Results and discussion

Optimization studies

Preliminary experiments comprise the optimization studies of the method in which a microvolume quartz cuvette is utilized for absorbance measurement of the DTZ complex of mercuric ions extracted into fine droplets of undecanol phase. Fig. 1 shows the absorption peak of the free DTZ at 420 nm has shifted to 490 nm in the presence of mercuric ions indicating Hg(II)–DTZ complex formation extracted into undecanol phase. This can also be observed from the color change of colorless undecanol phase to light green in the presence of DTZ and then to orange upon addition of mercuric ions to the solution (Fig. 1 inset). Since the extraction efficiency depends on different experimental parameters, such as medium pH, volume of undecanol and sample, DTZ concentration, extraction temperature and time, stirring rate and salt amount, these were optimized. The experiments were performed in a triplicate analysis.

Fig. 1 The absorption spectrum and the color of undecanol phase after DTZ and Hg(II)–DTZ extraction.

The effect of pH on the extraction efficiency was investigated by using 0.01 M Britton–Robinson (BR) buffer solutions in a wide range of pH. As shown in Fig. 2, the absorbance of the extracted solution has a maximum in the pH range of 6.5–8.0, and therefore pH 7.2 was chosen. Buffer type and concentration were also studied. Phosphate buffers at the same pH were tested and no significant difference was observed in a concentration range of 0.03–0.1 mol L⁻¹. Therefore, 0.1 mol L⁻¹ phosphate buffer at pH 7.2 was finally selected.

Fig. 2 Effect of pH on the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; undecanol volume: 10 μL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 5.0 × 10⁻⁶ mol L⁻¹; agitation rate: 200 rpm; extraction temperature: 25 °C; extraction time: 30 min; light path: 0.2 mm.

Temperature is another important parameter to be optimized for an efficient extraction. As shown in Fig. 3, the absorbance of Hg(II)–DTZ in undecanol phase is highly dependent on the temperature, having a maximum at 45 °C. The decline in the signal for higher temperatures can be explained by the instability of the complex. Consequently, further extraction processes were carried out at 45 °C.

Fig. 3 Effect of extraction temperature on the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; undecanol volume: 10 μL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 5.0 × 10⁻⁶ mol L⁻¹; agitation rate: 200 rpm; extraction time: 30 min; light path: 0.2 mm.

For accurate and precise analysis, extraction time should be optimized since mass transfer of the complex between water and undecanol phases is time dependent. The effect of the extraction time on the signal was investigated between 5 and 45 min and plotted against the measured absorbance (Fig. 4). Extraction efficiency is substantially increased with increasing extraction time up to 15 min and then shows a plateau after 20 min. Hence, the experiments were carried out using 20 min optimal extraction time.

Fig. 4 Effect of extraction time the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; undecanol volume: 15 μL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 5.0 × 10⁻⁶ mol L⁻¹; agitation rate: 200 rpm; extraction temperature: 45 °C; light path: 0.2 mm.

The dependence of the extraction efficiency on the solvent volume was investigated in the range of 5–100 μL. An undecanol nanodrop was picked from a solution containing 7.5 × 10⁻⁶ M Hg(II) and 1.5 × 10⁻⁵ M DTZ. A stable and relatively strong signal was observed in the range of 5–20 μL (Fig. 5). Considering that mass transfer of the complex into the organic phase occurs only by diffusion, a concentration gradient is produced in the drop for non-equilibrium conditions, thereby explaining the lower extraction efficiency observed for larger volumes.

Fig. 5 Effect of undecanol volume on the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 7.5 × 10⁻⁶ mol L⁻¹; agitation rate: 250 rpm; extraction temperature: 45 °C; extraction time: 20 min; light path: 0.2 mm.

In the next step, the sample volume was changed in the range of 1.0–20 mL with the undecanol volume kept constant at 15 μL. The extraction efficiency increased with sample to organic phase volume ratio as shown in Fig. 6. Here, 10 mL of sample volume was chosen since the reassembling of the microdrop distributed into higher sample volumes is more challenging.

Fig. 6 Effect of sample volume on the absorbance of Hg(II)–dithizonate complex. Conditions: undecanol volume: 15 μL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 7.5 × 10⁻⁶ mol L⁻¹; agitation rate: 250 rpm; extraction temperature: 45 °C; extraction time: 20 min; light path: 0.2 mm.

It is well known in liquid-phase microextraction that the addition of salt to the sample solution can increase the mass transfer of hydrophobic compounds into the extract phase (salting-out effect). On the other hand, the salt dissolved in the sample solution can change the physical properties of the Nernst diffusion layer altering the extraction kinetics.³⁷ Therefore, salt concentration in the sample solution should be optimized. In this study, the salting-out effect was investigated by addition of NaCl to the sample in the range 0.02–0.50 mol L⁻¹. As shown in Fig. 7, the extraction efficiency of Hg(II)–DTZ complex increased up to 0.1 mol L⁻¹ NaCl concentration and at higher NaCl concentration no significant change was observed. On the basis of these results, further experiments were performed at 0.1 mol L⁻¹ NaCl concentration.

Fig. 7 Effect of salt amount on the absorbance of Hg(II)–dithizonate. Conditions: sample volume: 10 mL; undecanol volume: 15 μL; DTZ concentration: 1.5 × 10⁻⁵ mol L⁻¹; Hg(II) concentration: 5.0 × 10⁻⁶ mol L⁻¹; agitation rate: 250 rpm; extraction temperature: 45 °C; extraction time: 20 min; light path: 0.2 mm.

In terms of sensitivity, instrumental parameters were also taken into consideration and the light path was changed from 0.2 mm to 1 mm to improve the absorbance signal for low Hg(II) concentration. Also, agitation of the sample solution allows an enhancement of the extraction kinetics as a result of the reduction of the Nernst diffusion film. Agitation of sample reduces the time required to reach equilibrium between the sample solution and undecanol phase. In this work, the effect of the agitation rate was studied in the range 50–250 rpm. The extraction efficiency increased with increasing stirring rate up to 250 rpm (Fig. 8). Larger agitation rates were avoided since the fine undecanol droplets dispersed into the solution cannot be easily collected. Thus, 250 rpm was selected as a compromise between sensitivity and the practicality of the method.

Fig. 8 Effect of agitation rate on the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; undecanol volume: 10 μL; DTZ concentration: 1.5 × 10⁻⁶ mol L⁻¹; Hg(II) concentration: 1.0 × 10⁻⁷ mol L⁻¹; extraction temperature: 45 °C; extraction time: 20 min; light path: 1.0 mm.

The last parameter to be optimized is the DTZ concentration which is expected to have a direct influence on the extraction efficiency of the Hg(II) complex. Generally, low ligand concentrations result in inefficient complex formation, but high concentrations would also reduce the absorbance signal as the free ligand tends to dissolve in undecanol phase. Therefore, the effect of DTZ concentration was examined in the range of 1.0 × 10⁻⁶ to 2.0 × 10⁻⁶ mol L⁻¹ and absorbance at 490 nm was plotted against DTZ concentration after baseline correction (Fig. 9). The signal has a maximum at 1.5 × 10⁻⁶ mol L⁻¹ and this concentration was selected for further studies.

Fig. 9 Effect of DTZ concentration on the absorbance of Hg(II)–dithizonate complex. Conditions: sample volume: 10 mL; NaCl: 0.1 mol L⁻¹; undecanol volume: 10 μL; agitation rate: 250 rpm; Hg(II) concentration: 1.0 × 10⁻⁷ mol L⁻¹; extraction temperature: 45 °C; extraction time: 20 min; light path: 1.0 mm.

Analytical figures of merit

Under the optimal extraction conditions, linearity, limit of detection (LOD), limit of quantification (LOQ), intra-day repeatability, inter-day reproducibility and enrichment factor of the proposed method were determined and are summarized in Table 1. Three replicate measurements were performed at each level. A good correlation coefficient (R² = 0.9989) was obtained in the working range. The detection (LOD) and quantification (LOQ) limits were calculated according the IUPAC approach as 3s/m and 10s/m (s being the standard deviation of 10 blank measurements and m the slope of the calibration line), respectively.

Table 1 Analytical figures of merit of the method

Linear working range	(3.2–50) × 10^−8 (mol L^−1)	6.4–100.8 (μg L^−1)
Linear equation	A = 1.61 × 10^6[C] − 0.0116
(R^2)	0.9989
LOD	9.6 × 10^−9 (mol L^−1)	1.9 (μg L^−1)
LOQ	3.2 × 10^−8 (mol L^−1)	6.4 (μg L^−1)
RSD % (intra-day) (n = 7)	8.5
RSD % (inter-day) (n = 7)	13.4
Enrichment factor	203

---

The repeatability of the method, expressed as relative standard deviation (RSD), was evaluated by extracting seven consecutive aqueous samples spiked at 1.0 × 10⁻⁷ mol L⁻¹ with Hg(II). Furthermore, the inter-day reproducibility was estimated by performing the calibration procedure over five consecutive days.

The enrichment factor is defined as the ratio of the final analyte concentration in the extracting phase to the initial aqueous sample concentration. Hg(II) content of a 5 × 10⁻⁸ M Hg(II) solution was determined by the AFS method prior to and after the procedure was applied. The enrichment factor was calculated as 203 which can be ascribed to the high sensitivity of the photometric method.

Interference studies

In order to investigate the selectivity of the proposed method, the effect of interfering ions that are usually present in saline samples was studied. DTZ is a versatile chelating agent; interferences may occur due to the competition of some heavy metal ions for DTZ and their subsequent co-extraction with Hg(II). The tolerance limit (M/M) was studied up to an ion/Hg(II) ratio of 5000 and defined as the concentration of the interfering ions added causing a relative error within ±10% in the true absorbance of the Hg(II)–dithizonate complex. The results are listed in Table 2. According to Table 2, the major ions in the saline sample matrices have no significant interferences in the analysis. Meanwhile, a few metallic ions including Cu(II) and Pb(II), which can compete with Hg(II) for the chelating agent, displayed much less tolerable limits at pH 7.2. One possible solution for that is to use higher concentration of DTZ, but background signal increases accordingly.

Table 2 Tolerance limits of interfering ions for the determination of Hg(II) (1.0 × 10⁻⁷ mol L⁻¹)

Foreign ion added	Interference/metal ratio (M/M)
Zn^2+	1
Cu^2+	5
Pb^2+	50
Ni^2+, Co^2+, Ca, Cd, Mg	500
Cd^2+, Cr^3+	5000
Al^3+, Bi^3+, Fe^2+, Fe^3+, K^+, Mn^2+ Br^−, NO3^−, SO4^2−	No interference

---

Another solution is to use EDTA as masking reagent. In fact, Zn(II) ions severely interfere in the measurement yielding a strong absorption at 490 nm not only due to its competition for DTZ, but also the ability of the complex to transfer to the undecanol phase Therefore, this co-extraction results in elevated signal formation which can be simply eliminated by adding EDTA so as to be 0.01 mol L⁻¹ in the mixture. By this means the interference of Zn(II) and other interfering ions can be prevented up to 5.0 × 10⁻⁴ mol L⁻¹ without any change in the signal of Hg(II) at ppb level.

Application of the method to saline samples

The developed method was applied for the determination of Hg(II) content of edible salt and natural saline water samples. Standard addition method was used in the sample analysis. SDME sample preparation technique was employed for the blank samples and for the standard spiked samples.

For this purpose the Hg(II) standard solutions were added into sample so as to be 2.0 × 10⁻⁸, 1.0 × 10⁻⁷, and 5.0 × 10⁻⁷ mol L⁻¹ and their recovery values were calculated. The recovery assays were replicated three times and acceptable recovery values were obtained (Table 3). The same samples were also analyzed with a reference AFS method for the verification of the accuracy of the method. The results are presented in Table 3. As can be seen from the results, mercury content of the samples found by the proposed method was in good agreement with AFS results.

Table 3 Mercury content and recovery values of saline samples analyzed by SDME and AFS methods

Sample type	Hg(II) found by		Spiked Hg(II) (mol L^−1)	Recovery (%)
	Proposed method	AFS
a LOD for AFS is 35 ng L^−1.
Sea water	<LOD	0.19 μg L^−1	2.0 × 10^−8	88.9 ± 7.2
			1.0 × 10^−7	100.4 ± 5.1
			5.0 × 10^−7	86.6 ± 3.7
Thermal spring water	<LOD	<LOD^a	2.0 × 10^−8	114.6 ± 8.8
			1.0 × 10^−7	98.6 ± 3.8
			5.0 × 10^−7	102.3 ± 3.4
Mineral water	<LOD	0.23 μg L^−1	2.0 × 10^−8	112.4 ± 5.5
			1.0 × 10^−7	98.9 ± 3.7
			5.0 × 10^−7	100.2 ± 2.9
Table salt	0.78 μg kg^−1	0.64 μg kg^−1	2.0 × 10^−8	113.4 ± 10.1
			1.0 × 10^−7	88.0 ± 5.6
			5.0 × 10^−7	100.3 ± 8.2
Iodized table salt	7.73 μg kg^−1	8.14 μg kg^−1	2.0 × 10^−8	110.9 ± 6.1
			1.0 × 10^−7	103.4 ± 3.5
			5.0 × 10^−7	99.2 ± 2.2
Rock salt	0.15 μg kg^−1	0.12 μg kg^−1	2.0 × 10^−8	102.3 ± 6.8
			1.0 × 10^−7	102.6 ± 4.2
			5.0 × 10^−7	100.9 ± 2.3

---

Comparison with other methods

Table 4 lists the methods developed for trace determination of Hg(II) in aqueous samples involving microextraction preconcentration techniques.

Table 4 Comparison of the proposed method with other microextraction techniques for determination of mercury in aqueous samples^a

Method	Sample type	Sample prep. technique	Extraction solvent type	LOD (ng mL^−1)	RSD (%)	Enrichment factor	Lit.
a ETV: electrothermal vaporization; AAS: atomic absorption spectroscopy; SDME: single-drop microextraction; HS-SDME: headspace single-drop microextraction; APDC: ammonium pyrrolydine dithiocarbamate; SFODME: solidified floating organic drop microextraction; ETV-ICP-MS: electrothermal vaporization inductively coupled plasma mass spectrometry; CV-AAS: cold vapor atomic absorption spectroscopy; CV-AFS: cold vapor atomic fluorescence spectrometry; HPLC: high-performance liquid chromatography; UV-VIS: UV-visible spectroscopy; DLLME: dispersive liquid–liquid microextraction; ISFME: in situ solvent formation microextraction.
ETV-AAS	River water	SDME	m-Xylene	0.01	6.1	970	1
ETV-AAS	Water fish	HS-SDME	Thiourea, APDC	5	3.3	—	2
ETV-AAS	Mineral, tap water	SFODME	Undecanoic acid	0.07	2.1	430	38
ETV-ICP-MS	Water	SDME	Ionic liquid	0.0098	5.2	50	34
CV-AAS	Sea water	HS-SDME	Ionic liquid	0.01	4.6	75	35
CV-AFS	Human saliva	SFODME	Undecanol	0.025	4.1	182	27
HPLC	Tap, river, waste water	SDME	Ionic liquid	22.8	11.6	3	36
UV-VIS	Tap, river water	SDME	Carbon tetrachloride	0.2	4.9	69	32
UV-VIS	Water	DLLME	Ionic liquid	3.9	1.7	18.8	39
UV-VIS	Mineral, river, sea water	ISFME	Ionic liquid	0.7	1.94	37	16
UV-VIS	Drinking, river, sea water	DLLME	Carbon tetrachloride	3.3	1.9–5.8	64	28
UV-VIS	Saline samples	SDME	Undecanol	1.9	8.5	203	This work

---

In comparison to these techniques, the method proposed in this study offers an inexpensive and rapid way for determining trace amounts of Hg(II) in various samples. Attention was paid to all saline samples including sea water, thermal spring and mineral water, rock and table salt, which can be complicated with other methods even those employing expensive and sophisticated instruments.

The method also has a potential to be exploited for field analysis by using a miniaturized photometer coupled with a compact extraction system with a greener solvent. The LOD level is well below the required limit⁸ and allows us to use the method for screening of pollutants in a number of environmental samples without the need of a sophisticated system. In comparison to other microextraction techniques,^38,39 this method has the advantage of extracting in a microdrop which does not require time-consuming freezing and melting steps as is the case for SFODME. By using a multi-vessel extraction system, a set of 6 samples can be simultaneously analyzed in less than 30 min.

Experimental

Reagents

All chemicals were of analytical reagent grade and working solutions were prepared in ultrapure water (Milli-Q, 18.2 MΩ cm, Millipore System Inc.). NaH₂PO₄·2H₂O, EDTA disodium (Triplex III) dihydrate and tin chloride were supplied from Merck (Darmstadt, Germany).

A standard 0.1 mol L⁻¹ solution of Hg(II) ions was prepared in 0.1 mol L⁻¹ HCl solution by dissolving a weighed portion of HgCl₂ (Merck Darmstadt, Germany) immediately before use. Working standard solution of Hg(II) was prepared by appropriate dilution of the stock standard solution with 0.1 mol L⁻¹ (pH 7.2) phosphate buffer solutions. Standard DTZ solution (7.5 × 10⁻⁴ mol L⁻¹) was prepared daily by dissolving an appropriate amount of the reagent in ethanol. Universal BR buffer solutions were prepared by mixing equimolar (0.04 mol L⁻¹) phosphoric, boric and acetic acid solutions and by dropwise addition of 0.2 mol L⁻¹ NaOH to provide a wide range (2–10) of pH. Tin chloride reducing agent (3% w/v) was used in AFS studies. All standards and extracted samples were stored at 4 °C in the dark. All glassware was soaked in 10% nitric acid for at least 24 h before use and then rinsed with ultrapure water.

Apparatus

A Transsonic 460/H ultrasonic bath was used for preparation of DTZ solution. A Jenway ion analyzer (model 3040 ion) with a combined glass electrode was used for pH measurements. Samples were placed into 20 mL clear glass screw vials with PTFE-coated caps (Agilent G1888A). Laboratory-made double-walled glass cell was connected to a circulating water bath (Nüve BS402) for temperature control. Undecanol was pipetted into the aqueous samples with a commercially available 25 μL glass syringe (Hamilton, model 1702). Heidolph Rotomax 120 model orbital shaker was used for extraction purposes. The droplet of undecanol on the surface of water was collected by a glass Pasteur pipette. Undecanol droplet was centrifuged in 250 μL pulled-point glass inserts. Mini centrifuge purchased from Combi-Spin FVL 2400 N (Boeco, Germany) was used for removing water at 2400 rpm.

A Varian, Cary 100 Bio UV-Vis spectrophotometer with a matched Hellma ultra-micro tray cell was used for recording the UV-visible absorption spectra. A PSA 10.004 Merlin Plus atomic fluorescence spectrometer (Kent, UK) was used for the determination of mercury. Cold vapor AFS measurements were made with a PSA 10.004 (PS Analytical, Sevenoaks, Kent, UK), which consisted of a PSA 20.099 random access model auto-sampler, continuous-flow vapor generation system and a fluorescence detector. Automated continuous-flow vapor generation system (PSA 10.003) was used to generate gaseous mercury. The generated mercury was then detected by utilizing a 254 nm interference filter to achieve wavelength isolation and reduction of background scatter (Merlin, PSA 10.023). Wet gas from the gas–liquid separator was continuously dried by using a semi-permeable Nafion membrane dryer tube (Perma Pure Products, USA). The salinity of the water samples was calculated by measuring the conductivity using a Metler Toledo FG3 system.

SDME procedure

A 10 mL aliquot of Hg(II) solution containing 0.1 mol L⁻¹ phosphate buffer (pH 7.2), 0.1 mol L⁻¹ NaCl and 1.5 × 10⁻⁵ mol L⁻¹ DTZ was placed into a 20 mL screw vial. 15 μL of undecanol was added as the extracting solvent and the Teflon-coated cap of the vial was closed tightly. Sample vial was placed into a laboratory-made thermostatic glass cell connected to a circulating Nüve water bath at 45 °C (Fig. 10). Six parallel samples can be extracted simultaneously in this assembly. The assembly was placed in an orbital shaker set at 250 rpm. The Hg(II)–dithizonate complex was extracted into the undecanol phase from the sample solution for 20 min, and then the droplet was carefully vacuumed in a glass posture pipette for transferring into a glass insert. Any residual water can be removed by centrifuging at 2400 rpm. Extracted sample droplet was placed onto the drop-supporting surface (pedestal) of the tray cell by an Eppendorf micropipette (0.5–5 μL). The gap was controlled by using 0.2 mm and 1 mm paths for the absorbance measurements. Background correction was performed by subtracting the absorbance of pure undecanol at 490 nm.

Fig. 10 A laboratory-made extraction vessel for SDME consisting of a thermostatic glass cell connected to a circulating water bath.

Sample preparation

Sea water, mineral water and thermal spring water samples were stored at 4 °C and filtered through 0.45 μm pore-sized cellulose acetate filters prior to analysis. The salinity of the samples was determined by conductivity measurements and then salinity was adjusted to 0.1 mol L⁻¹ NaCl by adding the necessary amount of solid NaCl. Then, the medium pH was adjusted to 7.2 by adding 0.1 mol L⁻¹ phosphate solution containing 0.01 mol L⁻¹ EDTA. Upon addition of DTZ to be 1.5 × 10⁻⁶ mol L⁻¹ the sample was made up to 100 mL. 10 mL aliquots of this mixture were then subjected to SDME.

Iodized table salt, non-iodized table salt and rock salt samples were obtained commercially from a local market and 0.6000 g of the salt samples were weighed precisely. The same procedure as described above was applied to the salt samples except addition of NaCl. The resultant samples were then subjected to SDME and subsequently analyzed by UV-visible spectrophotometry.

Conclusions

The present study provides a simple, rapid and yet sensitive spectrophotometric method for trace determination of Hg(II) in saline samples. Without any need of expensive instrumentation, spectrophotometry was coupled with single-drop microextraction for enrichment of Hg(II), thus minimizing organic solvent consumption. Analytical characteristics of the method were found to be comparable with sophisticated methods, and trace amounts of Hg(II) in saline water and salt samples were shown to be detected with good repeatability and high recoveries. The accuracy of the method was compared with atomic fluorescence spectrometry.

Consequently, the method is appropriate for automation and can be adapted to portable systems for field analysis.

Acknowledgements

The authors thank Ege University for financial support and also thank Professor Dr Emür Henden and Dr Onur Yayayürük for guidance during AFS measurements.

Notes and references

1. S. I. Honda, L. Hylander and M. Sakamoto, Environ. Health Prev. Med., 2006, 11, 171–176.
2. D. W. Boening, Chemosphere, 2000, 40, 1335–1351.
3. T. W. Clarkson and L. Magos, CRC Crit. Rev. Toxicol., 2006, 36, 609–662.
4. M. Yoshida, M. Satoh, A. Yasutake, A. Shimada, Y. Sumi and C. Tohyama, Toxicology, 1999, 139, 129–136.
5. M. Saber-Tehrani, M. Givianrad and H. Hashemi-Moghaddam, Talanta, 2007, 71, 1319–1325.
6. M. Saber-Tehrani, H. Hashemi-Moghaddam, M. H. Givianrad and P. Abroomand-Azar, Anal. Bioanal. Chem., 2006, 386, 1407–1412.
7. K. Leopold, M. Foulkes and P. J. Worsfold, TrAC, Trends Anal. Chem., 2009, 28, 426–435.
8. U. S. E. P. A., National Primary Drinking Water Regulations, EPA 816-F-09–004, May 2009.
9. D. P. Torres, V. L. Frescura and A. J. Curtius, Microchem. J., 2009, 93, 206–210.
10. M. Tuzen, I. Karaman, D. Citak and M. Soylak, Food Chem. Toxicol., 2009, 47, 1648–1652.
11. Y. Yin, J. Qiu, L. Yang and Q. Wang, Anal. Bioanal. Chem., 2007, 388, 831–836.
12. V. N. Tirtom, Ş. Goulding and E. Henden, Talanta, 2008, 76, 1212–1217.
13. Z. Zhu, G. C.-Y. Chan, S. J. Ray, X. Zhang and G. M. Hieftje, Anal. Chem., 2008, 80, 7043–7050.
14. H. Chen, J. Chen, X. Jin and D. Wei, J. Hazard. Mater., 2009, 172, 1282–1287.
15. P. R. Aranda, L. Colombo, E. Perino, I. E. De Vito and J. Raba, X-Ray Spectrom., 2013, 42, 100–104.
16. M. Baghdadi and F. Shemirani, Anal. Chim. Acta, 2009, 634, 186–191.
17. J. Fan, Y. Qin, C. Ye, P. Peng and C. Wu, J. Hazard. Mater., 2008, 150, 343–350.
18. F. Okçu, H. Ertaş and F. Sönmez, Talanta, 2008, 75, 442–446.
19. E. Y. Hashem, Spectrochim. Acta, Part A, 2002, 58, 1401–1410.
20. A. S. Amin and G. O. El-Sayed, Monatshefte für Chemie, 2001, 132, 587–596.
21. A. Niazi, A. Azizi and M. Ramezani, Spectrochim. Acta, Part A, 2008, 71, 1172–1177.
22. S. Chatterjee, A. Pillai and V. Gupta, Talanta, 2002, 57, 461–465.
23. H. Khan, M. J. Ahmed and M. I. Bhanger, Anal. Sci., 2005, 21, 507–512.
24. D. Nambiar, N. Patil and V. Shinde, Fresenius. J. Anal. Chem., 1998, 360, 205–207.
25. S. Gil, S. Fragueiro, I. Lavilla and C. Bendicho, Spectrochim. Acta, Part B, 2005, 60, 145–150.
26. D. Y. Sarıca and A. R. Türker, Clean: Soil, Air, Water, 2012, 40, 523–530.
27. C.-G. Yuan, J. Wang and Y. Jin, Microchim. Acta, 2012, 177, 153–158.
28. V. A. Lemos, L. O. d. Santos, E. d. S. Silva and E. V. d. S. Vieira, J. AOAC Int., 2012, 95, 227–231.
29. L. Xu, C. Basheer and H. K. Lee, J. Chromatogr. A, 2007, 1152, 184–192.
30. H. Bagheri and M. Naderi, J. Hazard. Mater., 2009, 165, 353–358.
31. M. A. Jeannot and F. F. Cantwell, Anal. Chem., 1996, 68, 2236–2240.
32. F. Yang, R. Liu, Z. Tan, X. Wen, C. Zheng and Y. Lv, J. Hazard. Mater., 2010, 183, 549–553.
33. F. Pena-Pereira, I. Lavilla and C. Bendicho, Spectrochim. Acta, Part B, 2009, 64, 1–15.
34. L. Xia, X. Li, Y. Wu, B. Hu and R. Chen, Spectrochim. Acta, Part B, 2008, 63, 1290–1296.
35. E. M. Martinis and R. G. Wuilloud, J. Anal. At. Spectrom., 2010, 25, 1432–1439.
36. F. Pena-Pereira, I. Lavilla, C. Bendicho, L. Vidal and A. Canals, Talanta, 2009, 78, 537–541.
37. F. Pena-Pereira, N. Cabaleiro, I. De la Calle, M. Costas, S. Gil, I. Lavilla and C. Bendicho, Talanta, 2011, 85, 1100–1104.
38. I. López-García, R. Rivas and M. Hernández-Córdoba, Anal. Bioanal. Chem., 2010, 396, 3097–3102.
39. M. Gharehbaghi, F. Shemirani and M. Baghdadi, International Journal of Environmental and Analytical Chemistry, 2009, 89, 21–33.

---