Copper-cobalt hexacyanoferrate modified glassy carbon electrode for an indirect electrochemical determination of mercury

Accepted Manuscript Title: Copper-cobalt hexacyanoferrate modified glassy carbon electrode for an indirect electrochemical determination of mercury Author: Vivek Vishal Sharma Domenica Tonelli Lorella Guadagnini Massimo Gazzano PII: DOI: Reference:

S0925-4005(16)31039-5 http://dx.doi.org/doi:10.1016/j.snb.2016.07.005 SNB 20503

To appear in:

Sensors and Actuators B

Received date: Revised date: Accepted date:

5-2-2016 15-6-2016 1-7-2016

Please cite this article as: Vivek Vishal Sharma, Domenica Tonelli, Lorella Guadagnini, Massimo Gazzano, Copper-cobalt hexacyanoferrate modified glassy carbon electrode for an indirect electrochemical determination of mercury, Sensors and Actuators B: Chemical http://dx.doi.org/10.1016/j.snb.2016.07.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Copper-cobalt hexacyanoferrate modified glassy carbon electrode for an indirect electrochemical determination of mercury Vivek Vishal Sharma, Domenica Tonelli*, Lorella Guadagnini, Massimo Gazzano1 Department of Industrial Chemistry “Toso Montanari”, University of Bologna, INSTM, UdR Bologna, Viale Risorgimento, 4 40136 Bologna, Italy; *corresponding author: tel. +39 0512093667; fax no. +39 0512093690; e-mail address: [email protected] 1

ISOF-CNR, at Department of Chemistry “G. Ciamician”, University of Bologna, Via Selmi 2, I-

40126 Bologna, Italy

________________________________________________________________________________

Graphical Abstract

Highlights 

A Cu/Co hybrid hexacyanoferrate was electrosynthesized on GC electrodes (CME)



Determination of Hg2+exploited the decrease in the oxidation current of L-cysteine



The major interfering cations were identified



Interference from Cu2+ was removed exploiting the slower kinetics towards L-cysteine



The CME was applied for the determination of Hg2+ in spiked mineral water samples

Abstract An electrochemical sensor based on a glassy carbon electrode modified with a hybrid film of copper cobalt hexacyanoferrate was prepared for the indirect electrochemical determination of Hg2+. It exploits the formation of a redox inactive complex with thiols. L-Cysteine and 1,4-Butanedithiol were used, with the former giving more sensitive results. Interference studies were carried out which led to Cu2+ being the major interferent. Exploiting the fact that the response of L-cysteine to Hg2+ is much more rapid compared to Cu2+, the interference from Cu2+ was avoided or minimized by quick measurement of the amperometric current after the addition of the analyte. The modified electrode showed promising results in detecting Hg2+ in spiked mineral water samples.

________________________________________________________________________________ Keywords: Mercury, Hexacyanoferrates, Glassy carbon, L-Cysteine, Interfering species

1. Introduction Mercury is a highly dangerous element because of its accumulative and persistent properties in the environment and biota. The toxicological interest of mercury is linked to its natural occurrence in the environment. The major source of mercury is the natural degassing of the earth‟s crust [1] while current emissions due to human activities are essentially from coal-burning power stations, municipal incinerators, mining activities and automobiles when they are recycled. Furthermore, because of its unique properties, mercury finds many applications in industry, agriculture, military

field, medicine and dentistry. This has resulted in worldwide contamination of large areas of industrial wastewater and led to elevated atmospheric mercury levels [1,2]. It is well known that the toxicity, biogeochemical behavior and transportation of mercury in the environment are strongly dependent on its chemical form. The inorganic forms of mercury include liquid metallic mercury and its vapor, compounds containing the mercurous (Hg22+) and mercuric Hg2+ ions, whereas the most common form of organic mercury is methylmercury, present in most if not all aquatic species and causing health risks which have been the subject of several large epidemiological investigations and continue to be the topic of intense debate [3]. Inhaled mercury vapor can damage the central nervous system due to its ability to cross the blood– brain barrier and this may cause some mental disabilities. Mercuric mercury (Hg2+) has a limited capacity to cross the blood–brain and placental barriers but it can be absorbed readily by humans and other organisms and is avidly accumulated by the kidneys [4]. Therefore, rapid, sensitive and inexpensive methods for Hg2+ determination are highly desirable. The determination of mercury has been generally carried out by cold-vapor atomic absorption or fluorescence spectrometry [5,6] and inductively coupled plasma spectroscopy techniques [7]. In addition, fluorimetric methods [8-10] and optical sensors [11,12] have been proposed. More recently, Quartz Crystal Microbalance (QCM) has been used to detect mercuric ions in wastewater [13] and a highly sensitive and specific monoclonal antibody has been obtained and used to develop an enzyme-linked immunosorbent assay for the determination of Hg2+ traces in environmental and food samples [14]. Among the analytical techniques currently employed for the trace determination of mercury in environmental samples, electrochemical ones have been extensively used due to their lower cost, major rapidity, and less laborious analysis [15,16]. Chemically modified electrodes (CMEs) have received increasing attention in the past decades and some papers have discussed the determination of Hg(II) ion by CMEs [17-21].

Since Hg(II) can inhibit the activity of numerous enzymes the use of electrodes modified with these biomolecules as transducers in amperometric and potentiometric techniques has paved the way for the development of biosensors [22-24]. A very sensitive and selective microbial sensor has been developed based on the inhibition of the alkaline phosphatase (ALP) enzyme present (or bound) in the cell wall of Chlorella sp. algae, immobilized on glassy carbon surface [25]. Recently, a screenprinted electrode modified with a nanostructured carbon black film has been proposed for the detection of Hg2+ ions in drinking water based on the fact that, in the presence of mercuric ions, thiol compounds form a complex which is not electroactive, so leading to a decrease of the electrochemical signal [26]. Following this analytical approach and thanks to our expertise in the field of electrodes modified with hexacyanoferrates (HCFs) as redox mediators, we propose in this work a sensor for the determination of mercury(II) ions in polluted environmental water samples, which is robust, cheap, simple to prepare and with a rapid response time. The sensor exploits a mixed hexacyanoferrate of copper and cobalt (Cu-CoHCF) deposited on a glassy carbon electrode (GCE) which acts both as a modifier (protecting the bare GC from fouling) and as an electrocatalyst for thiols oxidation mediated by the Fe2+/Fe3+ couple.

2. Experimental 2.1 Reagents and apparatus L-cysteine, 1,4-butanedithiol, Co(NO3)2 ∙ 6H2O and KNO3 were provided by Sigma-Aldrich (St. Louis, Missouri). The solutions of L-cysteine and 1,4-butanedithiol were prepared daily fresh and deoxygenated before use and were stored at 4º C when not in use. CuCl2 ∙ 2H2O was purchased from Merck (Germany), HgCl2 from Farmitalia Carlo Erba SpA (Milan, Italy) and K3[Fe(CN)6] from Riedel-de Haën (Germany). 0.1 M phosphate buffer solution (PBS) was prepared from H3PO4 (Farmitalia Carlo Erba) with the pH being set to the chosen value by subsequent addition of KOH (Farmitalia Carlo Erba). KNO3 (Sigma-Aldrich Steinheim, Germany) was added to the phosphate buffer solution as the supporting electrolyte up to a concentration of 0.1 M. All chemicals were of

analytical reagent grade and were used as received, without further purification. Aqueous solutions were prepared with doubly distilled (DD) water. The electrochemical experiments were performed on a CHI 660C potentiostat/galvanostat electrochemical workstation (CH Instruments Inc, Austin, TX, USA) interfaced with a personal computer. A three-electrode conventional electrochemical cell was employed, with a 3 mmdiameter glassy carbon electrode (GCE) the working, a saturated calomel electrode (SCE) as the reference and a Pt wire as the counter electrode.

2.2 Modification of the electrode surface Prior to surface modification, the GCE surface was gently polished on a fine (4000 grit) wet SiC sandpaper until a mirror finish was obtained. Then it was thoroughly rinsed with DD water and cycled many times in 0.25 M KNO3 at 0.050 V s-1 between 0 and +1.2 V, until superimposable voltammograms were recorded (blank signal). In our case this required 20 cycles. The working electrode, soaked in a de-aerated solution of 0.25 M KNO3, was modified with a thin film of Cu-CoHCF by a classic potentiodynamic method [27]. To this aim the solutions of the required salts were added, under magnetic stirring and in the specified order so as to obtain the following concentrations: 0.125 mM CuCl2, 0.125 mM Co(NO3)2 and 0.125 mM K3[Fe(CN)6]. To induce the precipitation of the Cu-CoHCF film, 40 cycles between 0 and +1.0 V, at 0.050 V s-1, were applied at the GCE [27]. After rinsing with water, some electrodes were kept at room temperature (25°C) until dryness and some were dried at 100°C for 1 h (thermal treatment). This treatment was aimed to increase the stability of the Cu-CoHCF film as previously demonstrated for Prussian Blue (PB). In fact, the presence of water molecules inside the PB polycrystal affects the conducting properties and, as a consequence, the stability of the corresponding PB modified electrodes. When GC electrodes modified with a PB film were kept at 100°C for 1 hour, an increase

in the amount of electroactive PB was demonstrated [28]. A similar procedure as that for glassy carbon was deployed for graphite paper in the deposition of Cu-CoHCF which was used to record directly the XRD pattern and to find out the molar ratio of Cu to Co by flame atomic absorption spectroscopy (FAAS).

2.3 Characterization of the Cu-CoHCF chemically modified electrode X-ray diffraction patterns were obtained with Cu-K radiation in reflection mode directly on the graphite electrodes by means of an X‟Pert PANalytical diffractometer equipped with a fast X‟ Celerator detector, 0.065° step, 20s/step. The slow scans were recorded with 600s/step integration time. Infrared spectra (IR) of the Cu-CoHCF deposited on graphite paper were taken on a Perkin Elmer Spectrum Two™ Attenuated Total Reflectance (ATR) FTIR spectrometer. An EVO 50 Series (LEO ZEISS) Scanning Electron Microscope (SEM) using an OXFORD INCA 350 EDS was employed to investigate the morphology of HCF films. The instrument acquired magnified topographic images of the sample at an accelerating voltage of 20 kV with the beam current as 100 pA. Flame atomic absorption spectrometry (ICE 3000 Series, ThermoScientific, Rockford, IL, USA) was used to analyze copper to cobalt molar ratio in the thermally treated Cu-CoHCF. The spectrometer was equipped with a copper/cobalt hollow cathode lamp, and a deuterium lamp for background correction. An air-acetylene flame was used for atomization using a solution acidified with HNO3 after dissolving the HCF in 0.5 M aqueous LiOH. The instrument was operated under the conditions recommended by the manufacturer and calibrated using standard solution of Co2+ and Cu2+ in HNO3 (Carlo Erba Reagents, Val-de-Reuil, France). 2.4 Chronoamperometric determination of mercury The Cu-CoHCF modified GCE was immersed in a conventional three electrodes electrochemical cell in the presence of 0.1 M KNO3 and 0.1 M PBS, pH 4. It was poised at +0.65 V and the current

was recorded until a stable background was reached. Then the thiol (CySH or BdSH) was added to the solution so that an appropriate concentration was obtained. When the oxidation current was stable, subsequent additions of HgCl2 were carried out, and, consequently, rapid decreases of current were recorded.

3. Results and discussion 3.1 Electrosynthesis and characterization of the Cu-CoHCF modified GCE Fig. 1 shows the cyclic voltammograms recorded during the electrodeposition process of the CuCoHCF film. The steadily increasing currents both for the anodic and cathodic peaks with increasing number of scans proves that the HCF films are deposited continuously on the GC surface. Insert Fig. 1 here To better characterize the nature in terms of crystallinity of the phase present on the electrode surface, an X-ray diffraction pattern was obtained. As shown in Fig. S1 in Supplementary information (SI), the scans recorded at 20s/step revealed only the crystal phase (graphite) of the electrodes, but some noise in the background suggested carrying out a deeper investigation. The need for strong conditions is due to the low amount of deposited material and to the presence of nanosized Cu-CoHCF aggregates, as shown by SEM images (Fig. S2 in SI). In case of the thermally treated CME, the film looks more uniform and it is evident an increase in particle size due to aggregation. By using very long data collection times, it was possible to distinguish the contribution of the graphitic substrate from the one due to the Cu-CoHCF, but only for the deposited material which had been thermally treated. Among the peaks, the principal ones are attributable to KNO3 crystals (see below for explanation), but the small reflections at 17.7° and 35.3° (d=0.4995 and 0.2553 nm) are consistent with the most intense peaks (200) and (400) of CoHCF, as reported by

Kaplun et. al. [35]. The presence of copper which can substitute for cobalt can be a further reason that makes it difficult to obtain good quality crystals. Cu-CoHCF was deposited on a graphite paper under the same electrochemical conditions as those of glassy carbon. For ATR-IR studies the graphite paper was chosen as the substrate because the sample holder of the equipment requires a very good contact between the sample and the diamond crystal for accurate measurement, which did not occur in the case of GCE. In fact, in case of glassy carbon, there was not a proper contact with the diamond crystal. We recorded the ATR-IR spectra of both the as prepared and thermally treated Cu-CoHCF deposited on graphite support. As shown, both spectra display the typical C-N stretching band which is clearly visible at around 2100 cm-1 (Fig. S3 in SI). The main difference between the two spectra is related to the presence of the broad absorption centred at  3400 cm-1 and the medium intensity band centred at 1762 cm-1 in the as prepared HCF which can be attributed to the water O-H stretching and H-O-H bending vibrations, respectively. After submitting the material to the thermal treatment the former band completely disappears and the latter becomes less intense, so confirming the loss of water molecules from the HCF, as already suggested by our group [29] for Cu-CoHCF and by De Mattos et al. for Prussian Blue [28]. In both spectra a very strong band at  1370 cm-1 and two other bands of medium intensity centred at  2394 and 824 cm-1, respectively, are also present, which could be attributed to the presence of potassium nitrate after performing the following experiment. A piece of graphite paper was dipped in an aqueous solution containing only KNO3 for 15 min, then it was removed from the solution and dried at room temperature. The ATR-IR spectrum of this graphite support was also recorded (Fig. S4 in SI). The copper to cobalt molar ratio was evaluated for the thermally treated HCF by FAAS and resulted about 1:4, a value which is in agreement with the result reported by Cui et al. which performed a study to correlate the ratio displayed by the electrosynthesized HCF with respect to the concentration ratio of Cu2+ and Co2+ in the electrolytic solution [36].

3.2 Mercury detection The analytical determination of CySH and BdSH by chronoamperometry (CA) using a Cu-CoHCF modified GCE was previously established by our group [29]. In the current work, the two thiols were also quantitated by differential pulse voltammetry (DPV) in order to evaluate which analytical technique would lead to a better sensitivity. As an example, in Figs. 2 A and B the calibration plots obtained for CySH by CA and DPV, respectively, are shown. In the case of DPV, the sensitivity was 0.118 A M-1 cm-2, which is about seven times lower than the value obtained by CA (0.814 A M-1 cm-2). Hence, CA was selected to find out the optimized conditions for the indirect electrochemical determination of Hg2+.

Insert Fig. 2 here

In the absence of mercury, thiols are oxidized at the electrode following the reaction: 2RSH → RSSR + 2e- + 2H+

(1)

In Fig S5 the chronoamperometric response for the addition of 10 M CySH to the buffer solution is shown. The response was recorded for 1000 s and shows that the oxidation current is not changing for such a long time and this result demonstrates the stability of the HCF modifier in the presence of the thiol. Calibration curves for Hg2+ determination could be obtained from the % decrease of the current (R%), after subsequent additions of a standard solution of Hg2+, applying the following equation [27]:

R% 

i0  iR  100 i0

(1)

where i0 and iR are the oxidation currents in the absence and in the presence of Hg2+, respectively, and R% is proportional to Hg2+ concentration,.

In order to find the optimum match in terms of limit of detection and limit of linearity for Hg2+ quantitative determination different concentrations of thiols were investigated. As an example in Fig. 3 is shown the calibration line obtained for CySH at a 10 µM concentration. The same procedure was adopted in the case of BdSH obtaining similar results. The limit of detection (LOD) for Hg2+ was found to be 80 nM for CySH and 130 nM for BdSH, respectively, and was calculated as suggested by Miller and Miller [31]. Moreover, CySH showed a higher sensitivity (315.5 µM-1 cm-2) as compared to BdSH (144.3 µM-1 cm-2). This result can be attributed to the stoichiometry of the complex between CySH and Hg2+ which is 2:1, while in the case of BdSH is 1:1. Insert Fig. 3 here

Due to the greater sensitivity displayed by CySH, it was further used for detection of Hg2+ in spiked mineral water samples and for interference studies. Obviously, the CySH concentration determines the detectable levels of mercury. We investigated three levels of concentration, i.e. 1, 5 and 10 M, and the analytical parameters of the relevant calibration lines are shown in Table 1. The limit of quantification (LOQ) was also calculated as suggested by Miller and Miller [31]. Actually, it was not possible to accurately evaluate the current decrease after mercury addition for 1 M CySH solution due to the very small value of current and the high noise associated with the signal.

Table 1 : Analytical parameters for the calibration plot of Hg2+ determination using two different concentrations of CySH. L-cysteine Conc.

LOQ -1

-2

Sensitivity (µM cm )

Linearity Range (µM)

LOD (µM)

(µM) 5

(µM) 380

0.25 – 2.5

0.080

0.25

10

0.25 – 5.0

365

0.080

0.25

A reproducibility study was carried out with five CMEs. Working with a thiol concentration of 5.0 µM, the average slope value of the five calibration plots was found to be (375 ± 20) µM-1 cm-2. Since the percentage coefficient of variation was about 5 %, we can conclude that the reproducibility of the proposed sensor is very good.

3.2 Interference study

Insert Fig. 4 here

Various cations were tested for their possible interference with a given concentration (10 µM) of CySH. All the cations were added to the solution at the same concentration (5 µM) as Hg2+ which caused a full disappearance of the CySH current. Fig. 4 shows that Cu2+ and Ag+ are the major interferents, with only negligible interference from other metal ions. CA responses of the various cations towards CySH are given in the Supplementary Information (Figs. S6, S7 and S8). In the case of Fe(II), the increase in current is due to the fact that such an ion is oxidized at the operative potential; the probability that iron is present in a real sample in its reduced form is, however, very low. If this is the case, the interference from Fe(II) can be eliminated, e.g., by adding a selective masking agent, like 1,10-phenanthroline. Since Ag+ is present in environmental water samples at extremely low concentrations (of the order of 10-10 M), the interference from Ag+ can be neglected for practical purposes. However, Cu2+ is a major interferent and various methods (both chemical and electrochemical) were deployed to remove its adverse effect on mercury determination. First we tried to avoid the interference of Cu2+ by chemical methods, using masking agents, such as 8-hydroxyquinoline (oxine) [32], 2,2‟-bipyridyl (bPy) [33], and allyl alcohol [34]. In the case of

oxine, we observed that the ligand does not form a precipitate with Cu2+ if the concentration is at µM level. For bPy there was a complex formation with Cu2+, however, the ligand had an adverse effect on the stability of the HCF film, which was evident from the reduced current value in the cyclic voltammogram of the blank buffer solution (Fig. S9 in SI). In the case of allyl alcohol, the ligand forms a stable complex with Cu(I), therefore a reducing agent like L-ascorbic acid was first added to Cu2+ containing solution, followed by allyl alcohol. However, allyl alcohol was unable to mask the interference from Cu2+ since the produced Cu(I) could oxidize cysteine before forming the complex with the alcohol, as evident from the CA plot in Fig. S10 (in SI). Therefore, the chemical approach was not successful. An electrochemical approach was also attempted to remove the interference from Cu2+ by reduction to Cu(0), without affecting Hg2+, working both in acetate buffer and aqua regia (7 times diluted) solutions. The latter was used to simulate the oxidative treatment often applied to environmental water samples. The reduction of the solutions containing Cu2+ or Hg2+ was carried out by linear sweep voltammetry at both the bare GC and the HCF modified electrodes. In all cases we found that Hg2+ was reduced at less cathodic potentials than Cu2+, so also the electrochemical approach was not feasible. Fig. 5 shows the two CA responses obtained by adding Hg2+ and Cu2+ (separately) at a concentration of 5 M, to a 10 M CySH solution. It is well evident that the kinetics of the complex formation is much slower in the case of Cu2+ as compared to Hg2+. Therefore, the Hg2+ determination could be carried out accurately also in the presence of Cu2+, if the current measurement was performed soon after (within 10 s) the addition of the mixture. Additional measurements were conducted by adding solutions with a constant Hg2+ (2 M) and varying Cu2+ (from 1 to 4 M) concentrations to a PBS containing 10 M CySH. The results showed that the response of the CME was as expected for the presence of only Hg2+. In conclusion, it can be said

that Hg2+ determination is accurate also in the presence of Cu2+, exploiting the different kinetics of complexes formation. Some common anions were also studied to verify if their presence could interfere in mercury determination. No interference was observed for CH3COO-, Cl-, NO3-, SO4-, HCO3- ions, for a concentration up to 1 mM. Insert Fig. 5 3.3 Experimental studies on packaged mineral water samples spiked with Hg2+ Packaged mineral water samples were spiked with Hg2+ (250 and 500 nM) and examined by the CME. The average concentration values (n = 5) were found to be 235 and 524 nM so that the percentage recovery ranged between 96 and 105, respectively, so confirming that the proposed sensor can be applied for the accurate determination of Hg2+ in polluted environmental water samples (Figure 6). Furthermore, the reproducibility of the analytical determination was good since the percentage coefficient of variation was about 6 %. Insert Fig. 6 here

Additional tests were also carried out on spiked mineral water samples containing both Hg2+ and Cu2+ ([Hg2+] : [Cu2+] = 1 : 2) which gave accurate results for the concentration of Hg2+ (Fig. 7). Insert Fig. 7 here

4. Conclusions A fast, sensitive and efficient electroanalytical method for the indirect electrochemical determination of Hg2+ has been successfully developed. The approach is able to determine

concentrations of Hg2+ as low as 250 nM, which is considered the LOQ, even though LOD is much lower (80 nM). Interference studies showed that Cu2+ was the major interferent with negligible interference from other cations and no interference from the most common anions. Amperometric studies demonstrated that the response of CySH towards Hg2+ was much faster compared to Cu2+ and this observation was exploited to overcome the interference from Cu2+ by a rapid measurement of the current after the addition of the analyte. The CME was used to successfully determine Hg2+ in spiked mineral water samples. It could also be employed to have a preliminary screening of the level of pollution in an environmental water sample.

Acknowledgements This work was funded by the University of Bologna. This project has been also funded with the support of the European Commission (Erasmus Mundus Action 2 India4EU II). This publication reflects the views only of the author, and the Commission cannot be held responsible for any use which may be made of the information contained therein.

Author Biographies Vivek Vishal Sharma obtained his Integrated Masters of Science degree in Chemistry from the Indian Institute of Technology (IIT) Roorkee, India in 2013. Currently, he is a PhD student in the Department of Industrial Chemistry, University of Bologna, on an Erasmus Mundus scholarship. He has been conducting research in the field of electrochemical sensors with mixed hexacyanoferrates and carbonaceous materials as modifiers for detection of various oxidizable analytes and heavy metals like mercury.

Domenica Tonelli obtained her degree in Chemistry in 1976. In 1992 she became an Associate Professor and since 2002 she is Full Professor of Analytical Chemistry at the University of Bologna. Since 1995 she carries out her research activity at the Department of Physical and Inorganic Chemistry of the University of Bologna (now named Department of Industrial Chemistry

“Toso Montanari”). At present her research interests deal with development of electrochemical sensors, characterization of materials, environmental and food chemistry by the use of chromatographic, spectrophotometric and electrochemical techniques. Lorella Guadagnini graduated first in “Industrial Chemistry” in October 2004, then in “Products, Materials and Processes in Industrial Chemistry” in July 2006, at the University of Bologna. In September 2006 she earned a scholarship and in June 2010 obtained the PhD title in Chemical Sciences. Later she carried out experimental activity as a post-doc researcher at the University of Bologna. Her interests are mainly focused on the fabrication and electrochemical characterization of amperometric sensors and biosensors by means of conventional electroanalytical techniques and Scanning Electrochemical Microscopy.

References [1] D.P. Ruys, J.F. Andrade, O.M. Guimarães, Mercury detection in air using a coated piezoelectric sensor, Anal. Chim. Acta 404 (2000) 95-100. [2] A. Safavi, N. Maleki, M.M. Doroodmand, Fabrication of a selective mercury sensor based on the adsorption of cold vapor of mercury on carbon nanotubes: Determination of mercury in industrial wastewater, J. Hazard. Mat. 173 (2010) 622-629. [3] C.T. Driscoll, R.P. Mason, H.M. Chan, D. J. Jacob, N. Pirrone, Mercury as a global pollutant: Sources, pathways, and effects, Environ. Sci. Technol. 47 (2013) 4967-4983. [4] T.W. Clarkson, L. Magos, The toxicology of mercury and its chemical compounds, Crit. Rev. Toxicol. 36 (2006) 609-662. [5] P. Rivaro, C. Ianni, F. Soggia, R. Frache, Mercury speciation in environmental samples by cold vapour atomic absorption spectrometry with in situ preconcentration on a gold trap, Microchim. Acta 158 (2007) 345-352. [6] W.-b. Zhang, Z.-f. Su, X.-f. Chu, X.-a. Yang, Evaluation of a new electrolytic cold vapor generation system for mercury determination by AFS, Talanta 80 (2010) 2106-2112. [7] X. Zhu, D.S. Alexandratos, Determination of trace levels of mercury in aqueous solutions by inductively coupled plasma atomic emission spectrometry: Elimination of the „memory effect‟, Microchem. J. 86 (2007) 37-41. [8] N. Aksuner, B. Basaran, E. Henden, I. Yilmaz, A. Cukurovali, A sensitive and selective fluorescent sensor for the determination of mercury(II) based on a novel triazine-thione derivative, Dyes and Pigments 88 (2011) 143-148. [9] Y.-Q. Wanga, Y. Liu, X.-W. He, W. Y. Li, Y.-K. Zhang, Highly sensitive synchronous fluorescence determination of mercury (II) based on the denatured ovalbumin coated CdTe QDs, Talanta 99 (2012) 69-74.

[10] J. Ma, Y. Chen, Z. Hou, W. Jiang, L. Wang, Selective and sensitive mercuric (II) ion detection based on quantum dots and nicking endonuclease assisted signal amplification, Biosens. Bioelectron. 43 (2013) 84-87. [11] Z. Sun, L. Jin, S. Zhang, W. Shi, M. Pu, M. Wei, D,G. Evans, X. Duan, An optical sensor based on H-acid/layered double hydroxide composite film for the selective detection of mercury ion, Anal. Chim. Acta 702 (2011) 95-101. [12] A. Yari, F. Papi, Ultra trace mercury(II) detection by a highly selective new optical sensor, Sens. Actuators B 160 (2011) 698-704. [13] H. Lou, Y. Zhang, Q. Xiang, J. Xu, H. Li, P. Xu, X. Li, The real-time detection of trace-level Hg2+ in water by QCM loaded with thiol-functionalized SBA-15, Sens. Actuators B 166-167 (2012) 246-252. [14] Y. Wang, H. Yang, M. Pschenitza, R. Niessner, Y. Li, D. Knopp, A. Deng, Highly sensitive and specific determination of mercury(II) ion in water, food and cosmetic samples with an ELISA based on a novel monoclonal antibody, Anal. Bioanal. Chem. 403 (2012) 2519-2528. [15] J. Wang, Analytical Electrochemistry, Wiley, New York, 2000. [16] A. Economou, Recent developments in on-line electrochemical stripping analysis-An overview of the last 12 years, Anal. Chim. Acta 683 (2010) 38-51. [17] F. Wang, X. Wei, C. Wang, S. Zhang, B. Ye, Langmuir–Blodgett film of p-tertbutylthiacalix[4]arene modified glassy carbon electrode as voltammetric sensor for the determination of Hg(II), Talanta 80 (2010) 1198-1204. [18] T. Alizadeh, M. Reza Ganjali, M. Zare, Application of an Hg2+ selective imprinted polymer as a new modifying agent for the preparation of a novel highly selective and sensitive electrochemical sensor for the determination of ultratrace mercury ions, Anal. Chim. Acta 689 (2011) 52-59. [19] A. Sánchez, S. Morante-Zarcero, D. Pérez-Quintanilla, I. Sierra, I. del Hierro, Determination of Hg(II) in natural waters using a carbon paste electrode modified with hybrid mesostructured silica nanoparticles, Sens. Actuators B 163 (2012) 38-43.

[20] X.-C Fu, X. Chen, Z. Guo, C.-G. Xie, L.-T Kong, J.-H. Liu, X.-J- Huang, Stripping voltammetric detection of mercury(II) based on a surface ion imprinting strategy in electropolymerized microporous poly(2-mercaptobenzothiazole) films modified glassy carbon electrode, Anal. Chim. Acta 685 (2011) 21-28. [21] A.J. Tchinda, E. Ngameni, A. Walcarius, Thiol-functionalized porous clay heterostructures (PCHs) deposited as thin films on carbon electrode: Towards mercury(II) sensing, Sens. Actuators B 121 (2007) 113-123. [22] C. Malitesta, M.R. Guascito, Heavy metal determination by biosensors based on enzyme immobilised by electropolymerisation, Biosens. Bioelectron. 20 (2005) 1643-1647. [23] O. Domínguez-Renedo, M.A. Alonso-Lomillo, L. Ferreira-Gonҫalves, M.J. Arcos-Martínez, Development of urease based amperometric biosensors for the inhibitive determination of Hg(II), Talanta 79 (2009) 1306-1310. [24] S. Cosnier, C. Mousty, A. Guelorget, M.Sanchez-Paniagua Lopez, D. Shan, A fast and direct amperometric determination of Hg2+ by a bienzyme electrode based on the competitive activities of glucose oxidase and laccase, Electroanalysis 23 (2011) 1776-1779. [25] J. Singh, S. K. Mittal, Chlorella sp. based biosensor for selective determination of mercury in presence of silver ions, Sens. Actuators B 165 (2012) 48-52. [26] F. Arduini, C. Majorani, A. Amine, D. Moscone, G. Palleschi, Hg2+ detection by measuring thiol groups with a highly sensitive screen-printed electrode modified with a nanostructured carbon black film, Electrochim. Acta 56 (2011) 4209-4215. [27] A. Abbaspour, A. Ghaffarinejad, Electrocatalytic oxidation of L-cysteine with a stable coppercobalt hexacyanoferrate electrochemically modified carbon paste electrode, Electrochim. Acta 53 (2008) 6643-6650 and references therein. [28] I.L. De Mattos, L. Gorton, T. Ruzgas, A.A. Karyakin, Sensor for hydrogen peroxide based on prussian blue modified electrode: Improvement of the operational stability, Anal. Sci. 16 (2000) 795-798.

[29] V. V. Sharma, L. Guadagnini, M. Giorgetti, D. Tonelli, Electrocatalytic determination of thiols using hybrid copper cobalt hexacyanoferrate modified glassy carbon electrode, Sens. Actuators B 228 (2016) 16-24. [31] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry, Wiley, New York, 1988. [32] I.M. Kolthoff, E.B. Sandell, E.J. Meehan, S. Bruckenstein, Analisi chimica quantitiva, Vol. 1, Piccin Editore, Padova, 295-297. [33] B. Narayana, N.G. Bhat, K.S. Bhat, C.H.R. Nambiar, B. Ramachandra, A. Joseph, Selective complexometric determination of copper in ores and alloys using 2,2‟-bipyridyl as masking agent, Microchem. J. 64 (2000) 221-225. [34] G. D. Ulibarri, T. Ogura, J. H. Tzeng, N. Scott, Q. Fernando, Allyl alcohol as masking agent in determination of copper(I) and copper(II) in aqueous mixtures, Anal. Chem. 54, 13 (1982) 23072310. [35] M. M. Kaplun, Yu. E. Smirnov, V. Mikli, V. V. Malev, Structure of Cobalt Hexacyanoferrate Films Synthesized from a Complex Electrolyte, Russ. J. Electrochem. 37 (2001) 914-923. [36] X. Cui, L. Hong, X. Lin, Electrochemical preparation, characterization and application of electrodes modified with hybrid hexacyanoferrates of copper and cobalt, J. Electroanal. Chem. 526 (2002) 115-124

Captions to figures : Fig. 1 : CVs for electrodeposition process of a hybrid Cu-CoHCF film on bare GC electrode, 40 cycles between 0 and +1.0 V (vs. SCE). Scan rate: 0.050 Vs-1

Fig. 2 Calibration plot for CySH additions (10 µM each) to 0.1 M PBS using (A) Chronoamperometry and (B) Differential pulse voltammetry. Fig. 3 Calibration plot of percentage decrease in current versus Hg2+ concentration when CA was performed in the presence of 10 µM CySH. Inset: CA plot for addition of 10 µM CySH to pH 4 PBS and subsequently six Hg2+ additions (each one 500 nM). In the calibration plot, the error bar represents the standard deviation of three replicated measurements. Fig. 4 Percentage decrease in current of 10 µM CySH on 5 µM additions of the interferents. Tests carried out in 0.1 M PBS, pH 4, with the thermally treated CuCoHCF modified GCE biased at +0.65 V. Fig. 5 Chronoamperometric response for the addition of 5 µM Hg2+ (A) and 5 µM Cu2+ (B) to a 10 µM CySH solution (same conditions as in Fig. 4). Fig. 6 Chronoamperometric response for a spiked (500 nM Hg2+) mineral water sample after the addition to a 10 µM CySH solution (same conditions as in Fig. 4). Fig. 7 Chronoamperometric plot for the addition of a mixture containing 1 µM Hg2+ and 2 µM Cu2+ spiked mineral water sample to a 10 µM CySH solution (same conditions as in Fig. 4).

Figures : Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7