The renewable bismuth bulk annular band working electrode: Fabrication and application in the adsorptive stripping voltammetric determination of nickel(II) and cobalt(II)

Analytica Chimica Acta 881 (2015) 44–53

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

The renewable bismuth bulk annular band working electrode: Fabrication and application in the adsorptive stripping voltammetric determination of nickel(II) and cobalt(II)
ska Bogusław Ba
s * , Krystian We˛giel, Katarzyna Jedlin Faculty of Materials Science and Ceramics, AGH University of Science and Technology, Mickiewicza 30, 30-059 Kraków, Poland

H I G H L I G H T S

G R A P H I C A L A B S T R A C T

 Design of a novel voltammetric sensor with the renewable bismuth bulk (annular band) electrode (RBiABE) is presented.  RBiABE is proposed as an alternative to bismuth film electrodes.  The activation principles and properties of the RBiABE sensor are characterized.  Simultaneous determination of Ni(II) and Co(II) in real samples was investigated.  The LOD of this sensor was 0.29 mg L 1 and 0.018 mg L 1 for Ni(II) and Co (II), respectively.

A R T I C L E I N F O

A B S T R A C T

Article history: Received 8 January 2015 Received in revised form 30 April 2015 Accepted 5 May 2015 Available online 7 May 2015

The paper presents the first report on fabrication and application of a user friendly and mercury free electrochemical sensor, with the renewable bismuth bulk annular band working electrode (RBiABE), in stripping voltammetry (SV). The sensor body is partly filled with the internal electrolyte solution, in which the RBiABE is cleaned and activated before each measurement. Time of the RBiABE contact with the sample solution is precisely controlled. The usefulness of this sensor was tested by Ni(II) and Co(II) traces determination by means of differential pulse adsorptive stripping voltammetry (DP AdSV), after complexation with dimethylglyoxime (DMG) in ammonia buffer (pH 8.2). The experimental variables (composition of the supporting electrolyte, pre-concentration potential and time, potential of the RBiABE activation, and DP parameters), as well as possible interferences, were investigated. The linear calibration graphs for Ni(II) and Co(II), determined individually and together, in the range from 1 10 8 to 70  10 8 mol L 1 and from 1 10 9 to 70  10 9 mol L 1 respectively, were obtained. The calculated limit of detection (LOD), for 30 s of the accumulation time, was 3  10 9 mol L 1 for Ni(II) in case of a single element’s analysis, whereas the LOD was 5  10 9 mol L 1 for Ni(II) and 3  10 10 mol L 1 for Co(II), when both metal ions were measured together. The repeatability of the Ni(II) and Co(II) adsorptive stripping voltammetric signals obtained at the RBiABE were equal to 5.4% and 2.5%, respectively (n = 5). Finally, the

Keywords: Bismuth bulk electrode Stripping voltammetry Nickel Cobalt Water analysis Mercury-free electroanalysis

* Corresponding author. Tel.: +48 12 634 1201; fax: +48 12 634 1201. E-mail address: [email protected] (B. Ba
s). http://dx.doi.org/10.1016/j.aca.2015.05.005 0003-2670/ ã 2015 Elsevier B.V. All rights reserved.

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

45

proposed method was validated by determining Ni(II) and Co(II) in the certified reference waters (SPS-SW1 and SPS-SW2) with satisfactory results. ã 2015 Elsevier B.V. All rights reserved.

1. Introduction Electrode materials have frequently been considered to be of utmost importance in analytical electrochemistry. A great variety of different materials has been suggested and exploited to produce useful working or indicator electrodes, from mercury and noble metals (e.g., Ag, Au, Pt), other metals (e.g., Cu, Ni, Pb), and different configurations of carbon based to the advanced metal-oxide electrodes [1–4]. Unfortunately, all the solid-state electrodes (SSEs) are characterized by low repeatability and reproducibility of the surface and very short periods of parameters’ stability. Hence, many different mechanical and electrochemical procedures of renovation/activation of SSEs are recommended in the literature. However, these procedures are time-consuming, require removal of the working electrode from the sample solution, and often are subjected to changes of its parameters. Therefore in electro analysis most often the metal film electrodes (MeFEs) are used. The latter are characterized by better stability and sensitivity than SSEs, ensure high precision and accuracy of measurements but are much less durable than SSEs. A bismuth film electrode (BiFE) was for the first time proposed by Wang et al. in 2000 [5]. Due to its less-toxic character, wide operational potential window and even improved performance in the presence of dissolved oxygen, bismuth has been accepted as an efficient replacement metal for mercury. However, for a wider application of Bi there are still some problems to be solved: relatively low detection limit and requirement of additional washing or polishing step of the substrate surface. Bismuth films have been shown to be very fragile. Furthermore, there are instances where preparation of the pre-plated BiFE or introduction of Bi ions into the solution is undesirable or impossible. Nowadays in SV, beside the BiFEs, bismuth-modified carbon paste electrodes, bismuth-modified carbon nanotubes, bismuth nanoparticles modified screen-printed carbon electrodes and bismuth singlecrystal-plane electrodes are frequently used [6–13]. Different microfabricated bismuth working electrodes were used to construct bismuth microelectrode arrays, miniaturized sensor chip (lab-on-a-chips), and also bismuth vibrating electrodes for stripping voltammetric detection of trace metals [14–20]. Among the bismuth-based electrode options, little work has been done using a bismuth bulk electrode (BiBE). The work with the BiBE can, however, offer many advantages; these are: minimization of wastes production (solution of bismuth salt is not needed to prepare the film), there is no need of time-consuming cleaning and deposition or regeneration of the film, feasibility of its use and fabrication, cost-efficiency and stability (under potential control). Up to now, application of the BiBEs in trace metals analysis has been mainly focused on the anodic stripping voltammetric (ASV) determination of Cd, Pb and Zn [19,21–23] and rarely on the other metals, for example speciation determination of Cr(III) and Cr(VI) in the presence of diethylenetriaminepentaacetic acid (DTPA) [7], determination of Ni and Co by square wave adsorptive cathodic stripping voltammetry (SW AdCSV) [20] and Fe by adsorptive constant-current stripping chronopotentiometric analysis (CC PSA) [24]. Also few organic compounds were analyzed using the BiBE [25,26]. Nickel and cobalt belong to the class of elements considered to be essential for man, plants and animals. Cobalt serves its paramount, established function as a component of vitamin B12 while nickel has recently been found to be essential for a number of

metabolic processes [27]. Ni(II) and Co(II) determinations by means of the AdSV measurements were carried out in ammonia or Britton Robinson (BR) buffers after complexation with dimethylglyoxime (DMG) [12,14,15,17,20,28–35]. In such buffers, bismuth salts undergo hydrolysis and the measurements cannot be performed at the in situ plated bismuth film electrode. The BiFEs applied till now for Ni(II) and Co(II) determinations are formed ex situ, usually from bismuth salts dissolved in an acetate buffer [6] or in LiBr + HCl solution [36]. Preplating influences total analysis time and requires exchange of solution before determination. Only in two works, the results of Ni(II) and Co(II) determinations on the BiBEs were presented, these were the bismuth microelectrode arrays [15] and the bismuth vibrating electrode [20]. In this work we present fabrication of our new renewable bismuth bulk annular band working electrode (RBiABE) and its application in adsorptive stripping voltammetric (AdSV) determination of Ni(II) and Co(II) ions. Construction of the presented RBiABE is very similar to those of the renovated silver ring electrode (RSRE) [37–39] and renewable ceramic electrodes described in our previous works [40,41]. The idea of the presented solutions consists in fast, few seconds lasting electrochemical activation of the working electrode (before each measurement) in the electrolyte solution filling the sensor body. Optimal conditions for the determination of Ni(II) and Co(II) in the form of DMG complexes at the RBiABE, were selected, as well as the possible interferences were investigated. The method was validated using the certified reference waters (SPS-SW1 and SPS-SW2) and was applied to analysis of natural water samples of Rudawa river (Cracow, Poland). 2. Experimental 2.1. Reagents and materials All chemicals were of analytical-reagent grade and used as received. All solutions were prepared with quadruple distilled water (two last stages from quartz). Atomic absorption standard solutions of Ni(II) and Co(II) at concentrations 1000 mg L 1 were obtained from Merck and diluted as required. 1 mol L 1 ammonia buffer (pH 8.2) was prepared by mixing the corresponding amounts of NH4Cl and ammonia solution (both Suprapur1, Merck). 1 mol L 1 acetate buffer (pH 5.0) was prepared by mixing the corresponding amounts of 96% acetic acid and 25% ammonia solution (both Suprapur1, Merck). 0.01 mol L 1 dimethylglyoxime (DMG, Sigma–Aldrich) solution was prepared by dissolving an appropriate amount of the reagent in 96% ethanol. The addition of NH4OH or HCl was used to obtain the required pH of the sample solution. Certified reference materials: Surface Waters SPSSW1 Batch 114 and Surface Waters SPS-SW2 Batch 118 were obtained from the National Institute of Standards and Technology USA. Triton X-100 (Fluka) and humic acid (sodium salt, Sigma– Aldrich) were used. 0.1% solution of humic acid was prepared by dissolving 50 mg of the reagent in 40 mL of water containing one drop of 25% NH3 and made up to 50 mL in a volumetric flask. The resin was TRANSLUX D180 (AKSON, France). The RBiABE was prepared from the bismuth foil 2.0 mm (0.08 in) thick, (ALFA AESAR, Germany, 99.999%, no. 41637). From the bismuth foil rectangular bar of size 2  2  10 mm was cut, and the final cylindrical shape (diameter 1.8 mm) was obtained using a turning lathe.

46

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

2.2. Apparatus

e) Moving back the BiABE into the sensor body with internal

electrolyte. Differential pulse adsorptive stripping voltammetry (DP AdSV) was applied using the M161 electrochemical system (mtm-anko, Poland). A three-electrode configuration consisting of a homemade electrochemical sensor with the bismuth bulk annular band working electrode (RBiABE, surface: 0.18 cm2), a double junction reference electrode Ag/AgCl/3 M KCl with replaceable outer junction (2.5 M KNO3), and a platinum wire as a counter electrode was used. All experiments were carried out in the onecompartment electrochemical cell at room temperature. A magnetic stirrer (ca. 300 rpm) was used for mixing solution during the accumulation period. Data were processed with the EALab 2.1 software [42,43]. 2.3. DP AdSV procedure The stripping analysis of Ni(II) and Co(II) was performed without the removal of oxygen. Appropriate volume of standards Ni and Co or sample solution was placed into the voltammetric cell, next 0.500 mL of 1 mol L 1 ammonia buffer (pH 8.2), 0.025 mL of 0.01 mol L 1 dimethylglyoxime (DMG) and 0.500 mL of 4 mol L 1 nitrite (NaNO2) were added, and the total volume of the test solution was finally made up to 5 mL with distilled water. The sensor body was filled with the internal electrolyte containing 0.05 mol L 1 acetate buffer (pH 5.0) and 0.01 mol L 1 KCl. The DP AdSV procedure was performed in an uninterrupted sequence with the following steps: a) Short electrochemical cleaning of the BiABE surface from traces

of adsorbed ions, complexes, surfactants etc., application of the activation potential Eact = 2.0 V vs. silver counter/quasi-reference electrode (AgQRE) for activation time tact = 5 s in the internal electrolyte filling the sensor body. b) Moving of the BiABE into the measurement cell with the sample solution. c) Adsorptive pre-concentration of Ni(II)–DMG and Co(II)–DMG complexes on the BiABE at accumulation potential Eacc = 0.70 V vs. Ag/AgCl/3 M KCl for accumulation time tacc = 30 s with stirring. d) After a rest period (equilibration time) of teq = 5 s, the DP voltammogram was recorded by applying a negative going potential scan from 0.70 to 1.20 V with a step potential, Es = 3 mV; pulse amplitude, dE = 30 mV and pulse period, timp = 20 ms.

For all experiments, at least two independent assays were carried out; minimum three repetitive scans were made per measurement. Quantification was performed by standard additions of Ni(II) and Co(II) method. 2.4. Determination of Ni(II) and Co(II) in real samples The developed RBiABE sensor performance was assessed by applying it for determination of Ni(II) and Co(II) in the certified surface water samples (CRMs) and also for the determination of both ions in river water. The CRM samples were analyzed without pre-treatment. River water samples collected from Rudawa river (Cracow, Poland), in clean polypropylene bottles, were analyzed after a minor pre-treatment, consisting of successive filtrations through Whatman1 filter No 1 (Sigma–Aldrich) and a 0.45 mm cellulose acetate membrane filter in the Sartorius device. The analyzed sample was pipetted into the electrochemical cell and quadruply distilled water was added to the volume 3.975 mL. Then, 0.500 mL of 1 mol L 1 ammonia buffer, 0.025 mL of 0.01 mol L 1 DMG and 0.500 mL of 4 mol L 1 NaNO2 were added. If pH of the solution was different from 8.2  0.05, NH4OH, it was added to obtain the required value. 2.5. Construction of the RBiABE sensor and the principle of the bismuth electrode activation Construction of the presented herein electrochemical sensor with the renewable bismuth bulk annular band working electrode (RBiABE) is basically similar to those of our other sensor with the renovated silver ring electrode (RSRE) [37,38]. In both examples, perfect sensor’s surface renovation was achieved by means of its electrochemical activation. To do this, to the working electrode (SRE or BiABE) constant activation potential (Eact) equivalent to hydrogen evolution reaction (HER) is connected for a few seconds. What is important, the electrode activation is not performed in the analyzed sample solution, but in the electrolyte solution which fills the sensor body. The composition and concentration of the internal electrolyte as well as the activation potential of the electrode are optimized experimentally, depending on the used electrode material and composition of the sample solution. The activation potential is set in reference to the AgQRE mounted in the sensor body. From the point of view of SV, important is the fact, that the

Fig. 1. (A) Construction of the BiABE: (a) bismuth tube, (b) stainless steel wire, (c, d) resin. (B) The sensor with RBiABE used in all experiments: (1) bismuth annular band working electrode (BiABE); (2) silver wire covered with AgCl layer as a counter/quasi-reference electrode (AgQRE); (3) O-ring; (4) sensor body; (5) internal electrolyte; (6) PTFE centering element. (C) One full cycle of Ni and Co determination and the BiABE activation.

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

time of the working electrode contact with the sample solution is precisely controlled. As a result, the analyte pre-concentration time on the working electrode and the time of interfering agents adsorption on its surface can be as precisely controlled as in case of the mercury drop electrodes. During movement of the MeABE through the silicon O-ring, hydrogen bubbles and also all other solid impurities not chemically bounded to the MeABE, are removed from its surface. Even in case of manually operated sensor, the total time of the working electrode regeneration is no longer than few seconds. The main difference in construction of the RSRE and RBiABE consists in shape of the working electrode. Namely, the SRE electrode is formed in the shape of a straight tube of very small surface of contact between silver and resin [38]. In the effect, mechanical polishing of the electrode as well as repeated pushing it through the O-ring and a dozen or so temperature variations lead to loosening of the contact and permanent damage of the electrode. Also, the use in the RSRE construction insulation in the form of thermos-shrinkable plastic tube, turned out to be less effective. Thus, in the construction of the RBiABE other shape of the working electrode (Fig. 1A), of expanded, two-step contact surface between bismuth and resin, was applied. Moreover, the stainless steel wire with the embedded Bi electrode was, along the whole length, covered with resin. In this way, chemical stability, durability and mechanical resistance of the RBiABE comparing to the RSRE, were substantially improved. While not in use, the BiABE is stored in distilled water filling the sensor body. Fig. 1A shows preparation of the main RBiABE sensor component, i.e., the bismuth annular band working electrode (BiABE). The polycrystalline bismuth tube of a special shape (a) (’o.d = 1.8 mm, ’i.d = 0.9 mm, l = 3.5 mm) was slid over and mechanically tightened on the stainless steel wire (b) (’ = 0.9 mm). The steel wire below and above the bismuth tube was covered with resin (c). The excess of resin was then mechanically removed (d). After mounting, the electrode surface (resin and the bismuth electrode) was grounded by emery papers of decreasing roughness and finally polished with 0.05 mm alumina slurry (Buehler), thoroughly washed with distilled water and dried. Fig. 1B shows the construction of complete, ready-for-measurement sensor configuration: (1) the bismuth annular band working electrode (BiABE); (2) the silver wire (0.5 mm) covered with AgCl layer as a counter/quasi-reference electrode (AgQRE); (3) O-ring (’i.d = 1.5 mm) made from 2 mm thick silicon rubber; (4) polypropylene sensor body; (5) internal electrolyte; (6) PTFE centering element. Fig. 1C shows a scheme of a single cycle of Ni(II) and Co(II) determination by means of adsorptive stripping voltammetry on the RBiABE, as well as regeneration/activation of its surface.

47

internal electrolyte (filling the sensor body) containing 0.05 mol L 1 of acetate buffer (pH 5.0) and 0.01 mol L 1 of KCl. The influence of the constant activation potentials (Eact) for stripping peak current Ni(II)–DMG and Co(II)–DMG complexes was investigated for the supporting electrolyte containing 5  10 7 mol L 1 Ni(II) and 5  10 8 mol L 1 Co(II). The activation potential was changed in the range from 1.3 to 2.2 V. The results of BiABE activation obtained for voltammograms of Co(II)–DMG complex are presented in Fig. 2. As it can be seen, the Co(II) peak is not observed when Eact < 1.4 V, next it increases in quasi-linear mode as the activation potential decreases to 1.8 V. Below 2.0 V the Co (II) peak current does not change significantly. Similar results were obtained for Ni(II)–DMG complex (not presented data). Despite the activating potential, the time of activation step (tact) was also selected experimentally in the range from 1 to 30 s. According to our procedure, the RBiABE is fast and effectively cleaned and activated before each measurement by applying the constant HER potential Eact = 2.0 V vs. AgQRE for tact = 5 s. Such operation guaranties very good reproducibility of the Ni and Co peak currents and the optimal signal-to-background current ratio. Moreover, during chemical activation of the RBiABE, AgCl is continuously build-up on the AgQRE what guarantees its potential stability during cleaning (activation) steps. 3.2. Optimization of conditions for DP AdSV of Ni and Co determination at the RBiABE To establish the most adequate experimental conditions for DP AdSV protocol to determine Ni(II) and Co(II) using the RBiABE, a univariate optimization study was carried out with the parameters that may affect voltammetric signal of the target species as variables. 3.2.1. Effect of ammonia buffer, ligand (DMG) and nitrite ions concentration The mixture of ammonia buffer, DMG and NaNO2 was selected, because basing on the opinion of many scientists, it is an excellent supporting electrolyte for the simultaneous determination of Ni(II) and Co(II) ions [20,28–36].

3. Results and discussion 3.1. Cleaning procedure Voltammetric stripping is known to be affected by surfactants. Adsorption of surface active compounds leads to obtaining lower or broader peaks and shifts in the peak potential. The other reason of the signal decrease and low precision of measurements are chemical reactions (for example oxidation) and physical changes (morphology–roughness) of the electrode surface. Therefore, often and even before each measurement renovation of the solid-state electrode is necessary. It is especially easy in the case of the renovated ring (annular band) electrodes [37–41,44,45]. Problems, like electrochemical cleaning and regeneration of the BiABE and formation of solid bismuth hydroxy-compounds in ammonia and acetate buffers at various pH, were intensely investigated in our team. Thus, we have found that the best effect of cleaning and regeneration of the BiABE is achieved in the

Fig. 2. DP AdSV voltammograms of 5  10 8 mol L 1 Co(II) following different activation potential of the BiABE, from bottom to top: 1.30; 1.35; 1.40; 1.45; 1.50; 1.80; 2.00 and 2.20 V. The supporting electrolyte: 0.1 mol L 1 ammonia buffer (pH 8.2), 5  10 5 mol L 1 DMG and 0.4 mol L 1 NaNO2. Activation time: tact = 5 s. Accumulation: Eacc = 0.7 V, tacc = 30 s with stirring. DP instrumental parameters: teq = 5 s, Es = 3 mV, dE = 30 mV, timp = 20 ms. Inset: the resulting plot of Co(II)–DMG stripping peak current vs. activation potential of the BiABE.

48

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

The effect of concentration of ammonia buffer on the stripping responses (peak currents and peaks separation (EpNi–EpCo)) in the solution containing 30  10 8 mol L 1 Ni(II) and 3  10 8 mol L 1 Co(II) was examined in the range 0.025–0.2 mol L 1 and is illustrated in Fig. 3A. It can be seen, that the Ni peak height increased 3-times with increasing buffer concentration while the Co peak height increased slowly up to 0.15 mol L 1 concentration of buffer and remained constant for higher concentrations. This general dependence was also observed for AdSV on bismuth bulk and film electrodes [29,34–36]. Thus, buffer concentration of 0.1 mol L 1 was adopted for the subsequent experiments. The effect of the ligand (DMG) concentration was investigated in the range 5  10 6 to 2  10 4 mol L 1 and it is illustrated in Fig. 3B. At low DMG concentrations, both peaks increased with increasing DMG concentration. However, the Ni peak height essentially stabilized for DMG amount concentrations higher than 5  10 5 mol L 1. On the contrary, the Co peak height increased almost linearly up to DMG concentration of 5  10 5 mol L 1 and still continued to increase at higher DMG concentration but at lower rate. Thus, DMG concentration of 5  10 5 mol L 1 was adopted for the subsequent experiments.

Effect of nitrite (NaNO2) concentration was examined in the range 0.1–0.6 mol L 1 and is illustrated in Fig. 3C. As it can be seen, the Ni peak height increased slightly up to 0.4 mol L 1 of nitrite, next greatly decreased while Co peak height increased greatly with increasing NaNO2 concentration and resulted in improvement of the Ni(II) and Co(II) peaks separation. Thus, NaNO2 at concentration of 0.4 mol L 1 was adopted for the subsequent experiments. 3.2.2. Effect of pH The potential window of the RBiABE significantly depends on pH. The cathodic limit is given by the reduction of protons and therefore it is shifted with increasing pH. Unfortunately, in nonacidic media, the oxidized form of Bi(III) hydrolyses very easily. In turn, the anodic limit is more negative than in acidic media. 0.05 mol L 1 acetate buffer (pH 5.0) was chosen as the best internal electrolyte for activation of the RBiABE. For the strong acidic (pH < 2) electrolytes we have observed disadvantageous changes in the working electrode surface. The formation and stability of the Ni(II)–DMG and Co(II)–DMG complexes are strongly dependent on pH. The influence of pH on the Ni and Co striping responses was studied in pH range

Fig. 3. Effect of several chemical parameters: (A) ammonia buffer (pH 8.2); (B) dimethylglyoxime (DMG); (C) nitrite (NaNO2) and (D) pH on the voltammetric stripping responses (peak currents and peaks separation (EpNi–EpCo)) for 30  10 8 mol L 1 Ni(II) and 3  10 8 mol L 1 Co(II). Other conditions as in Fig. 2. Values are the average of three replicates of three independent experiments.

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

5.2–8.6 and is presented in Fig. 3D. It can be seen, that the Co peak height increased greatly in pH range 5.2–8.6 while the Ni peak height initially increased and decreased at pH values greater than 8.4, perhaps due to the formation of hydroxyl-compounds. The profile indicated that pH 8.2 offers the most favorable performance for Ni(II) and Co(II) determination at the RBiABE.

49

sampling time), tw = ts) from 10 to 60 ms (not presented data). The best result of the signal-to-background current ratio and peaks separation (EpNi–EpCo) was obtained for Es = 3 mV, dE = 30 mV and timp = 20 ms, therefore these value were adopted for the subsequent experiments. 3.3. Stability and reproducibility of the RBiABE

3.2.3. Effect of accumulation potential and time The influence of accumulation potential (Eacc) on the Ni and Co stripping responses was studied in the range from 0.55 to 0.95 V and is shown in Fig. 4A. The accumulation (adsorption) of both Me–DMG complexes at the RBiABE surface increased greatly as the potential became more negative than 0.4 V and the height of both peaks almost stabilized at potential more negative than 0.55 V. Because the Ni and Co peaks heights revealed strong decreasing tendency at more negative potentials than 0.8 V, therefore for further studies the accumulation potential of 0.7 V was chosen. Effect of accumulation time (tacc) on the Ni and Co stripping responses in the range 10–600 s is illustrated in Fig. 4B. The Ni and Co peaks currents increased initially almost linearly with the increasing accumulation time and at higher accumulation times (tacc > 100 s) the plots for both metals started to level-off as equilibrium of surface concentration of the adsorbed complexes was approached. As expected for the linearized part of the Langmuir adsorption isotherm, the peak currents corresponding to the saturation surface concentration for both complexes were lower at lower solution concentrations of metals. It was also interesting to note, that the plot for Co exhibited a greater slope indicating that the use of longer accumulation times caused a greater enhancement of the Co peak. Accordingly, the accumulation time of 30 s was selected as the optimal compromise between high sensitivity and short analysis time. 3.2.4. Effect of DP parameters To optimize the conditions for simultaneous determination of Ni(II) and Co(II) ions, the following instrumental parameters were systematically varied: step potential (Es) in the range 1–6 mV, pulse amplitude (dE) from 10 to 60 mV (both positive and negative modes) and pulse width (timp = tw (waiting time) + ts (current

Stability of the RBiABE was checked by recording successive DP curves for the supporting electrolyte. For the fifty cycles, no change was observed in the voltammetric profiles of the RBiABE. As shown in the selected examples, for short periods of time stability of the RBiABE is excellent. Using the RBiABE to detect 30  10 8 mol L 1 Ni(II) and 3  10 8 mol L 1 Co(II) by DP CAdSV, slight change (2–3%) was observed in twenty consecutive measurements. While using it to detect five different Ni and Co samples (same concentration: 10  10 8 mol L 1 Ni(II) and 1 10 8 mol L 1 Co(II)), the R.S.D. observed were equal to 5.4% and 2.5%, respectively. Over longer periods (two week), for 4–6 h of measurements per day, the signal sensitivity fluctuates by about 20%. The response of the RBiABE is stable when the electrode is used continuously what prevents from oxidation of the electrode’s surface. Mechanical polishing (0.05 mm Al2O3) of the BiABE is only necessary after its exposition to air for more than 1 h. According to this, every few days the electrode was polished to avoid oxidation problems related to the influence of air and oxygen dissolved in the examined solutions. The RBiABE sensor used in synthetic solutions, free of surfactants, could be used ca. a few hundred stripping runs without replacement of the internal electrolyte. However, if the RBiABE is used for the measurements of the real samples, the internal electrolyte must be changed every several dozen of stripping runs, once a week the AgQRE was covered with the new AgCl layer. 3.4. Tolerance to interfering species The analytical methodology proposed in this paper is developed for evaluating traces of Ni(II) and Co(II) in matrices of

Fig. 4. Effect of the accumulation potential (Eacc) (A) and the accumulation time (tacc) (B) on the voltammetric stripping response (peak currents and peaks separation (EpNi–EpCo)) for 30  10 8 mol L 1 Ni(II) and 3  10 8 mol L 1 Co(II). Activation potential Eact = 2.0 V. Other conditions as in Fig. 2.

50

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

environmental interest, where potentially interfering co-existing ionic species and surface active substances can be present. It was documented that determination of 30  10 8 mol L 1 Ni (II) and 3  10 8 mol L 1 Co(II) at the RBiABE is influenced neither by 200-fold excess of Zn(II) nor by 100-fold excess of Fe(II), Mn(II) and Cu(II). Triton X-100 was chosen to simulate the effect of typical nonionic surfactant on the DP AdSV determination of 30  10 8 mol L 1 Ni(II) and 3  10 8 mol L 1 Co(II) at the RBiABE. 40% and 16% reduction of the initial values of the Co and Ni peak currents was observed after addition of 1 mg L 1 Triton X-100. However, the addition of 5 mg L 1 of Triton X-100 caused complete elimination of the Co signal and 70% decrease of the Ni signal. Furthermore, the influence of organic matter was studied by the addition of humic acid of concentration up to 3 mg L 1. Also in this case decay of the Ni and Co peaks was observed. Humic acid added in the amount of 1 mg L 1 resulted in 55% and 42% decrease of the Co and Ni response, respectively. Addition of 2.5 mg L 1 of humic acid caused complete elimination of both stripping responses. Natural waters are thought to contain organic matter with surfactant effect similar to that of 0.05–0.5 mg L 1 of Triton-X-100 [19]. Therefore, determination of Ni and Co at the RBiABE is possible in natural water samples without pre-treatment. But in such case, the key role plays cleaning of the RBiABE before each measurement. As it was proven in the course of our research deaeration of the sample solution for voltammetric analysis of Ni(II) and Co(II) at the RBiABE is not necessary, what is in agreement with literature information. 3.5. Analytical performance Dependence of stripping peak currents on the concentration of nickel and cobalt is shown in Fig. 5, which presents the DP AdSV voltammograms for increasing Ni(II) and Co(II) concentrations. The voltammetric curves of Ni and Co, obtained after 30 s of accumulation, were well shaped and separated (EpNi  0.92 V; EpCo  1.07 V). The difference in Ni and Co peaks potential for the RBiABE equal to dEp  150 mV was higher than those reported for

the HMDE (dEp  140 mV), MFEs (dEp  100 mV) or BiFEs plated on glassy carbon (dEp  130 mV) but is lower than that for the lead screen-printed electrodes (dEp  200 mV) [35]. Polynomial of 3th degree function was chosen as optimal for baseline reconstruction and extraction of Ni and Co peaks. Background subtraction was preceded by manual choice of the points intervals which were included in the baseline calculation [46,47]. Sensitivity of the AdSV determination of Co was 88  1 nA nM 1 for Ni(II) and 2870  33 nA nM 1 for Co(II). The calibration plots under the optimized conditions were linear from 1 10 8 to 70  10 8 mol L 1 for Ni(II) and from 1 10 9 to 70  10 9 mol L 1 for Co(II). The determination coefficients (R) were equal to 0.9991 and 0.9993 for Ni and Co, respectively. The detections limits (for tacc = 30 s), defined as three times the standard deviation of blank, are estimated to be 5  10 9 mol L 1 (0.29 mg L 1) Ni(II) and 3  10 10 mol L 1 (0.018 mg L 1) Co(II). Use of the RBiABE and the DP AdSV technique results in good repeatability of the measured signal. Five successive measurement of the Ni and Co peaks currents showed relative standard deviation of 5.4% and 2.5% for 1 10 7 mol L 1 and 1 10 8 mol L 1 of Ni(II) and Co(II), respectively. For individual determination of Ni(II), the linear calibration curve was obtained for the tacc = 30 s in the range from 4  10 9 to 4  10 8 mol L 1 according to the equation y = (101  2)x + 70  3, where y and x are the peak current (nA) and Ni(II) concentration (nM), respectively. The linear correlation coefficient was R = 0.9983. The detection limit for the lowest studied Ni(II) concentration and the tacc = 30 s was about 3  10 9 mol L 1 (0.17 mg L 1). For Co(II), the peak shape is similar in individual and simultaneous determinations. The LODs were well below the guideline values of Ni and Co in drinking, surface, and ground water formulated by the WHO [48,49], and it was proved that the sensor can be used for the real samples analysis. Moreover, by adjusting the accumulation time it is possible to obtain calibration graphs of different sensitivity and various dynamic range. Many analytical methods have been reported for detecting Ni (II) and Co(II) in different environmental samples and in the presence of various interfering agents. Table 1 summarizes past works for which the best values of sensitivity, LOD and linear range were achieved.

Fig. 5. (A) DP AdSV voltammograms obtained at the RBiABE for non-deaerated supporting electrolyte containing increasing amount of Ni(II) and Co(II). Concentration of Ni(II) from bottom to top: blank; 34; 68; 101; 135; 167; 234; 300; 365; 429; 493; 557; 621 and 685  10 9 mol L 1. Concentration of Co(II) from bottom to top: blank; 3.4; 6.8; 10.1; 13.5; 16.8; 23.5; 30.1; 36.6; 43.1; 49.5; 55.9; 62.3 and 68.7  10 9 mol L 1. (B) Corresponding calibration plots with R of 0.9991 for Ni(II) and R of 0.9993 for Co(II). Activation potential Eact = 2.0 V. Other conditions as in Fig. 2.

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

51

Table 1 Summary of past efforts on detection of Ni(II) and Co(II) obtained by various analytical methods. Method

LOD Ni(II) (mg L 1)

LOD Co(II) (mg L 1)

Linearity Ni(II) (mg L 1)

Linearity Co(II) (mg L 1)

References

ICP-OES ICP-MS HR-ICP-MS FI-ICP-MS HF-SPME-ICP-MS ED-XRF WD-XRF Capillary electrophoresis F-AAS DLPME-GF-AAS Ad CSV SW AdSV DP AdSV

0.44 8.0  10 1.5  10 9.0  10 2.0  10 2.0 0.6 0.012 1.1 0.033 0.097 0.2 0.29

0.28 4.0  10 1.2  10 2.3  10 3.9  10 1.5 0.7 180 0.9 0.021 0.058 0.3 0.018

1–1000 0.5–100 – – 0.1–50 max. 400 max. 150 0.006–5900 1.1–100 – 5–35 0.6–2.9 0.59–41

1–1000 0.5–100 – – 0.01–50 max. 400 max. 150 59–5900 0.9–100 – 10–55 0.6–5.9 0.06–4.1

[50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [34] [35] Present work

3 3 4 2

4 4 5 4

3.6. Analysis of certified reference water Accuracy of the designed method employing the RBiABE sensor was assessed by analyses of two types of the certified water samples: Surface waters SPS-SW1 and SPS-SW2 with different concentrations of Ni(II) and Co(II) using the standard addition procedure. Three replications of analyses were performed for each sample. The samples were analyzed without pre-treatment. The results for Ni(II), 9.93  0.24 mg L 1 for SPS-SW1 and 49.76  0.27 mg L 1 for SPS-SW2, agreed well with the respective total certified values of 10.0  0.05 mg L 1 and 50.0  0.3 mg L 1. The results for Co(II), 1.88  0.09 mg L 1 for SPS-SW1 and 10.04  0.15 mg L 1 for SPS-SW2, also agreed well with the respective total certified values of 2.00  0.02 mg L 1 and 10.0  0.05 mg L 1.The results obtained indicate that the proposed procedure can be applied for the simultaneous determination of Ni(II) and Co(II) in natural water samples containing amounts of Ni(II) and Co(II) greater than 0.29 mg L 1 (5  10 9 mol L 1) and 0.018 mg L 1 (3  10 10 mol L 1), respectively. 3.7. Analytical application of the RBiABE in analysis of river water samples The proposed method of simultaneous determination of Ni(II) and Co(II) was tested by analysis of the water samples of Rudawa river collected from two different sites on the suburbs of Cracow (Poland). The applied procedure of samples pre-treatment is

described in Section 2.4. Before the analysis time the samples were stored at 4  C. The baseline (dashed line in Fig. 6) was recorded to verify cleanness of the glassware and purity of the reagents. Then, 1 mL of the sample was diluted with the supporting electrolyte and distilled water to the final volume of 5 mL (see: Section 2.3) and after the accumulation time of 30 s, DP voltammogram was recorded. The concentrations of Ni(II) and Co(II) in the water samples were calculated through the standard addition method, i.e., 3 additions of 1 mg L 1 Ni(II) and 0.1 mg L 1 Co(II), resulting in the values of 1.87  0.05 mg L 1 and 0.08  0.02 mg L 1 (n = 3), respectively. The determined values are in good agreement with the nickel and cobalt concentrations reported in surface water in Europe [48,49]. The inset of Fig. 6 also demonstrates that a satisfactory linear response was achieved with a correlation coefficients (R) of 0.998 and 0.999 for Ni and Co, respectively. The accuracy of the method was tested on the same river sample spiked with the known amounts of Ni(II) 2 mg L 1 and Co(II) 0.25 mg L 1. It was found that the recovery values of Ni(II) and Co(II) were satisfactory being 94.1% and 96,3%, respectively. 4. Conclusions In this work, for the first time the new electrochemical sensor with the renewable bismuth bulk annular band working electrode (RBiABE) for stripping voltammetry (SV) in non-deaerated solutions is presented. The simple, very fast cleaning and activation procedures assure complete and efficient removal of all electrochemical products and adsorbed species from the electrode surface before each measurement cycle. The usefulness of this sensor was tested for the simultaneous determination of Ni(II) and Co(II) by differential pulse adsorptive stripping voltammetry (DP AdSV) after complexation with dimethylglyoxime (DMG) in the ammonia buffer (pH 8.2). Under the optimized operational conditions, the detection limits of 5  10 9 mol L 1 (0.29 mg L 1) and 3  10 10 mol L 1 (0.018 mg L 1) were obtained for Ni(II) and Co(II), respectively. Good selectivity against several metal ions and surfactants was observed. By analyzing the CRMs and river water samples, the capability of the RBiABE for use in the environmental analysis was demonstrated. Based on the presented preliminary results, the advantageous analytical properties of the RBiABE, combined with the minimal toxicity of metallic bismuth, indicate that there is extended scope for the adoption of this type of sensor for alternative target analytes in the field of adsorptive stripping analysis. Acknowledgement

Fig. 6. DP AdSV determination of Ni(II) and Co(II) in Rudawa river (Cracow, Poland). The baseline (dashed line) and the sample before and after three successive standard additions of Ni and Co in 1 mg L 1 Ni(II) and 0.1 mg L 1 Co(II) steps. Other conditions as in Fig. 2. Inset: corresponding standard additions plot.

This work was supported by the National Science Centre of Poland (Project No. DEC-2011/03/B/ST5/02713).

52

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53

References [1] J. Wang, Analytical Electrochemistry, 3rd ed., John Wiley & Sons, Inc., Hoboken, New Jersey, 2006. [2] C.G. Zoski, Handbook of Electrochemistry, Elsevier B.V., The Netherlands, 2007. [3] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and Applications, 2nd ed., John Wiley & Sons, Inc., NY, 2001. [4] J.O.M. Bockris, A.K.N. Reddy, Modern electrochemistry, 2nd ed., Electronics in Chemistry, Engineering, Biology, and Environmental Science, 2B, Kluwer Academic/Plenum, New York, 2000. [5] J. Wang, J. Lu, S.B. Hocevar, P.A.M. Farias, B. Ogorevc, Bismuth-coated carbon electrodes for anodic stripping voltammetry, Anal. Chem. 72 (2000) 3218–3222. [6] A. Królicka, A. Bobrowski, Bismuth film electrode for adsorptive stripping voltammetry – electrochemical and microscopic study, Electrochem. Commun. 6 (2004) 99–104. [7] E.O. Jorgea, M.M. Rochab, I.T.E. Fonsecab, M.M.M. Neto, Studies on the stripping voltammetric determination and speciation of chromium at a rotating-disc bismuth film electrode, Talanta 81 (2010) 556–564. [8] J. Wang, Stripping analysis at bismuth electrodes: a review, Electroanalysis 17 (15–16) (2005) 1341–1346. [9] I. Švancara, L. Baldrianova, E. Tesarova, S.B. Ho9 cevar, S.A.A. Elsuccary, A. Economou, S. Sotiropoulos, B. Ogorevc, K. Vytras, Recent advances in anodic stripping voltammetry with bismuth-modified carbon paste electrodes, Electroanalysis 18 (2) (2006) 177–185. [10] G.H. Hwang, W.K. Han, J.S. Park, S.G. Kang, Determination of trace metals by anodic stripping voltammetry using a bismuth-modified carbon nanotube electrode, Talanta 76 (20) (2008) 301–308. [11] C. Chen, X. Niu, Y. Chai, H. Zhao, M. Lan, Bismuth-based porous screen-printed carbon electrode with enhanced sensitivity for trace heavy metal detection by stripping voltammetry, Sens. Actuators B Chem. 178 (2013) 339–342. [12] L.A. Piankova, N.A. Malakhova, N.Yu. Stozhko, Kh.Z. Brainina, A.M. Murzakaev, O.R. Timoshenkova, Bismuth nanoparticles in adsorptive stripping voltammetry of nickel, Electrochem. Commun. 13 (2011) 981–984. [13] A. Jänes, P. Miidla, E. Lust, Adsorption of 1-heptanol on bismuth single-crystal plane electrodes, J. Solid State Electrochem. 7 (4) (2003) 189–200. [14] C. Kokkinos, A. Economou, I. Raptis, T. Speliotis, Disposable lithographically fabricated bismuth microelectrode arrays for stripping voltammetric detection of trace metals, Electrochem. Commun. 13 (2011) 391–395. [15] C. Kokkinos, A. Economou, I. Raptis, Microfabricated disposable lab-on-a-chip sensors with integrated bismuth microelectrode arrays for voltammetric determination of trace metals, Anal. Chim. Acta 710 (2012) 1–8. [16] Z. Zou, A. Jang, E. MacKnight, P.-M. Wu, J. Do, P.L. Bishop, Ch.H. Ahn, Environmentally friendly disposable sensors with microfabricated on-chip planar bismuth electrode for in situ heavy metal ions measurement, Sens. Actuators B Chem. 134 (2008) 18–24. [17] C. Kokkinos, A. Economou, I. Raptis, T. Speliotis, Disposable mercury-free cellon-a-chip devices with integrated microfabricated electrodes for the determination of trace nickel(II) by adsorptive stripping voltammetry, Anal. Chim. Acta 622 (2008) 111–118. [18] N.G. Naseri, S.J. Baldock, A. Economou, N.J. Goddard, P.R. Fielden, Disposable electrochemical flow cells for catalytic adsorptive stripping voltammetry (CAdSV) at a bismuth film electrode (BiFE), Anal. Bioanal. Chem. 391 (4) (2008) 1283–1292. [19] Z. Bi, C.S. Chapman, P. Salaün, C.M.G. van den Berg, Determination of lead and cadmium in sea- and freshwater by anodic stripping voltammetry with a vibrating bismuth electrode, Electroanalysis 22 (24) (2010) 2897. [20] Georgina M.S. Alves, Júlia M.C.S. Magalhães, Helena M.V.M. Soares, Simultaneous determination of nickel and cobalt using a solid bismuth vibrating electrode by adsorptive cathodic stripping voltammetry, Electroanalysis 25 (5) (2013) 1247–1255. [21] R. Pauliukaite, S.B. Ho9 cevar, B. Ogorevc, J. Wang, Characterization and applications of a bismuth bulk electrode, Electroanalysis 16 (9) (2004) 719–723. [22] K.C. Armstrong, C.E. Tatum, R.N. Dansby-Sparks, J.Q. Chambers, Z.-L. Xue, Individual and simultaneous determination of lead, cadmium, and zinc by anodic stripping voltammetry at a bismuth bulk electrode, Talanta 82 (2010) 675–680. [23] M. de la Gala Morales, M.R. Palomo Marin, L. Calvo Blazquez, E. Pinilla Gil, Performance of a bismuth bulk rotating disk electrode for heavy metal analysis: Determination of lead in environmental samples, Electroanalysis 24 (5) (2012) 1170–1177. [24] W. Pimrote, R. Ratana-Ohpas, L. Renman, A combination bismuth bulk electrode and its use in adsorptive constant-current stripping chronopotentiometry for iron using solochrome violet RS, Electroanalysis 23 (7) (2011) 1607–1614. [25] O. El Tall, D. Beh, N. Jaffrezic-Renault, O. Vittori, Electroanalysis of some nitrocompounds using bulk bismuth electrode, Int. J. Environ Anal. Chem. 90 (1) (2010) 40–48. [26] M. Bu9 cková, P. Gründler, G.-Uwe Flechsig, Adsorptive stripping voltammetric detection of daunomycin at a bismuth bulk electrode, Electroanalysis 17 (5-6) (2005) 440–444. [27] E.J. Underwood, Trace Elements in Human and Animal Nutrition, 4th ed., Academic Press, New York, 1977.

[28] S.B. Adeloju, A.M. Bond, M.H. Briggs, Assessment of differential-pulse adsorption voltammetry for the simultaneous determination of nickel and cobalt in biological materials, Anal. Chim. Acta 164 (1984) 181–194. [29] J. Wang, J. Lu, Bismuth film electrodes for adsorptive stripping voltammetry of trace nickel, Electrochem. Commun. 2 (2000) 390–393. [30] M. Morfobos, A. Economou, A. Voulgaropoulos, Simultaneous determination of nickel(II) and cobalt(II) by square wave adsorptive stripping voltammetry on a rotating-disc bismuth-film electrode, Anal. Chim. Acta 519 (2004) 57–64. [31] P. Kapturski, A. Bobrowski, The silver amalgam film electrode in catalytic adsorptive stripping voltammetric determination of cobalt and nickel, J. Electroanal. Chem. 617 (2008) 1–6. [32] C. Kokkinos, A. Economou, M. Koupparis, Determination of trace cobalt(II) by adsorptive stripping voltammetry on disposable microfabricated electrochemical cells with integrated planar metal-film electrodes, Talanta 77 (2009) 1137–1142. [33] R. Segura, M. Pradena, D. Pinto, F. Godoy, E. Nagles, V. Arancibia, Adsorptive stripping voltammetry of nickel with 1-nitroso-2-napthol using a bismuth film electrode, Talanta 85 (2011) 2316–2319. [34] A. Mardegan, S. Dal Borgo, P. Scopece, L.M. Moretto, S.B. Ho9 cevar, P. Ugo, Simultaneous adsorptive cathodic stripping voltammetric determination of nickel(II) and cobalt(II) at an in situ bismuth-modified gold electrode, Electroanalysis 25 (11) (2013) 2471–2479. [35] A. Bobrowski, A. Królicka, M. Maczuga, J. Zare˛bski, A novel screen-printed electrode modified with lead film foradsorptive stripping voltammetric determination of cobalt and nickel, Sens. Actuators B Chem. 191 (2014) 291–297. [36] E.A. Hutton, S.B. Ho9cevar, B. Ogorevc, M.R. Smyth, Bismuth film electrode for simultaneous adsorptive stripping analysis of trace cobalt and nickel using constant current chronopotentiometric and voltammetric protocol, Electrochem. Commun. 5 (9) (2003) 765–769. [37] B. Ba
s, M. Jakubowska, Z. Kowalski, Rapid pretreatment of a solid silver electrode for routine analytical practice, Electroanalysis 18 (17) (2006) 1710–1717. [38] B. Ba
s, The renovated silver ring electrode, Electrochem. Commun. 10 (2008) 156–160. [39] B. Ba
s, R. Piech, E. Niewiara, M. Ziemnicka, L. Stobierski, W.W. Kubiak, TiC working electrode: voltammetric characteristics and application for determination of lead traces by stripping voltammetry, Electroanalysis 20 (15) (2008) 1655–1664.
ski, M. Robótka, Renewable ceramic [40] B. Ba
s, R. Piech, M. Ziemnicka, W. Reczyn (TiN) ring electrode in stripping voltammetry: determination of Pb(II) without removal of oxygen, Electroanalysis 21 (16) (2009) 1773–1780. [41] B. Ba
s, M. Jakubowska, F. Ciepiela, W.W. Kubiak, New multipurpose electrochemical analyzer for scientific and routine tasks, Instrum. Sci. Tech. 38 (6) (2010) 421–426. [42] M. Jakubowska, Signal processing in electrochemistry, Electroanalysis 23 (3) (2011) 553–572. _ F. Ciepiela, Novel renovated silver ring [43] B. Ba
s, M. Jakubowska, M. Jez, electrode for anodic stripping analysis of Pb(II) and Cd(II) traces in water samples without removal of oxygen and surfactants, J. Electroanal. Chem. 638 (2010) 3–8.
ski, F. Ciepiela, W.W. Kubiak, Rapidly [44] B. Ba
s, M. Jakubowska, W. Reczyn renewable silver and gold annular band electrodes, Electrochim. Acta 73 (2012) 98–104. [45] Ł. Górski, F. Ciepiela, M. Jakubowska, Automatic baseline correction in voltammetry, Electrochim. Acta 136 (2014) 195–203. [46] F. Ciepiela, G. Lisak, M. Jakubowska, Self-referencing background correction method for voltammetric investigation of reversible redox reaction, Electroanalysis 25 (9) (2013) 2054–2059. [47] Nickel in Drinking-water. WHO Guidelines for Drinking-water Quality. http:// www.who.int/water_sanitation_health/gdwqrevision/nickel2005.pdf. 28.12.2014. [48] Cobalt and inorganic cobalt compounds. Concise International Chemical Assessment Document 69. http://www.inchem.org/documents/cicads/cicads/ cicad69htm#6.1. 28.12.2014. [49] H. Sereshti, Y.E. Heravi, S. Samadi, Optimized ultrasound-assisted emulsification microextraction for simultaneous trace multielement determination of heavy metals in real water samples by ICP-OES, Talanta 97 (2012) 235–241. [50] S. Chen, M. Xiao, D. Lu, Z. Wang, The use of carbon nanofibers microcolumn preconcentration for inductively coupled plasma mass spectrometry determination of Mn, Co and Ni, Spectrochim. Acta B 62 (2007) 216–1221. [51] A. Milnea, W. Landing, M. Bizimis, P. Morton, Determination of Mn, Fe Co, Ni, Cu, Zn, Cd and Pb in seawater using high resolution magnetic sector inductively coupled mass spectrometry (HR-ICP-MS), Anal. Chim. Acta 665 (2010) 200–207. [52] S. Nakatsuka, K. Okamura, K. Norisuye, Y. Sohrin, Simultaneous determination of suspended particulate trace metals (Co, Ni Cu, Zn, Cd and Pb) in seawater with small volume filtration assisted by microwave digestion and flow injection inductively coupled plasma mass spectrometer, Anal. Chim. Acta 594 (2007) 52–60. [53] S. Su, B. Chen, M. He, B. Hu, Graphene oxide–silica composite coating hollow fiber solid phase microextraction online coupled with inductively coupled

B. Ba
s et al. / Analytica Chimica Acta 881 (2015) 44–53 plasma mass spectrometry for the determination of trace heavy metals in environmental water samples, Talanta 123 (2014) 1–9. [54] K. Kocot, B. Zawisza, R. Sitko, Dispersive liquid–liquid microextraction using diethyldithiocarbamate as a chelating agent and the dried-spot technique for the determination of Fe, Co Ni, Cu, Zn, Se and Pb by energy-dispersive X-ray fluorescence spectrometry, Spectrochim. Acta B 73 (2012) 79–83. [55] B. Zawisza, R. Skorek, G. Stankiewicz, R. Sitko, Carbon nanotubes as asolid sorbent for the preconcentration of Cr, Mn Fe, Co, Ni, Cu, Zn and Pb prior to wavelength-dispersive X-ray fluorescence spectrometry, Talanta 99 (2012) 918–923.

53

[56] S. Almeda, H.E. Gandolfi, L. Arce, M. Valcárcel, Potential of porphyrins as chromogenic reagents for determining metals in capillary electrophoresis, J. Chromatogr. A 1216 (2009) 6256–6258. [57] V.A. Lemos, R.S. da Franca, B.O. Moreira, Cloud point extraction for Co and Ni determination in water samples by flame atomic absorption spectrometry, Sep. Purif. Technol. 54 (2007) 349–354. [58] H. Jiang, Y. Qin, B. Hu, Dispersive liquid phase microextraction (DLPME) combined with graphite furnace atomic absorption spectrometry (GFAAS) for determination of trace Co and Ni in environmental water and rice samples, Talanta 74 (2008) 1160–1165.