Anal Bioanal Chem (2015) 407:983–990 DOI 10.1007/s00216-014-8259-9

RESEARCH PAPER

Ti metal electrode as an unconventional amperometric sensor for determination of Au(III) species Fabio Terzi & Barbara Zanfrognini & Stefano Ruggeri & Giulio Maccaferri & Laura Pigani & Chiara Zanardi & Renato Seeber

Received: 21 July 2014 / Revised: 6 October 2014 / Accepted: 8 October 2014 / Published online: 19 October 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract The control of the noble metal concentration is crucial in order to increase the efficiency of hydrometallurgic processes in mining and in the recovery of precious materials from electronic waste. The present study is devoted to the development of an effective procedure for the quantification of Au(III) species dissolved in aqueous solutions, similar real complex matrices included. In particular, a novel electrode system based on Ti has been studied. This electrode material is still poorly investigated in the framework of electroanalysis, despite its lack of sensitivity to common interfering species, such as oxygen; hence, the determination of metal species can be carried out without performing deaeration of the solution. In addition, the interfering effect due to the presence of other heavy metal ions, such as Ag, Fe and Pb, has been minimised by a proper choice of the conditions adopted for the amperometric measurements. Ti electrodes exhibit reproducible electrochemical responses, even in the presence of high concentration of organic fouling species typical of biosorption processes. Keywords Electroanalysis . Amperometric sensor . Titanium . Gold . Heavy metals determination

Introduction The recovery of heavy metals from aqueous solutions is of outmost importance in a number of industrial and environmental frames, from the mining industry to the wastewater Published in the topical collection celebrating ABCs 13th Anniversary.

F. Terzi : B. Zanfrognini : S. Ruggeri : G. Maccaferri : L. Pigani : C. Zanardi : R. Seeber (*) Department of Chemical and Geological Sciences, University of Modena and Reggio Emilia, 41121 Modena, Italy e-mail: [email protected]

treatment and from the environmental remediation to the waste recycling [1–3]. A number of different processes have been developed to these purposes [1–3]. In particular, the procedures based on the use of biomasses have proven to offer very interesting effectiveness [4–6]: Some biomasses, such as those from food industries and forest activities, are produced in large quantities, are relatively inexpensive and are suitable to recover or recycle, by biosorption, a large number of heavy metals. Among these, Au is, for obvious reasons such as the lack of abundant natural sources and the high economic value, a particularly important one. Au is recovered from fluids deriving from hydrometallurgical processes typical of mining and carried out in plants for recycling of scrapped electronic components [2, 7, 8]. Recovery should be obviously coupled to monitoring of the initial and of the residual content; currently, Au can be determined in laboratory through, for instance, inductively coupled plasma systems, even hyphenated with mass spectrometry for best sensitivity. Despite the effectiveness, these analytical methods suffer from a number of drawbacks, such as the complexity of the instrumentation, the need for dedicated laboratories and for compressed inert gases, as well as for skilled work force. In any case, they require sampling, transport and eventual pre-treatment of the sample, viz. all the disadvantages, in terms of costs and of low frequency of measurements that are connected to off-line analyses [9]. The determination of dissolved Au species has been also carried out through amperometric techniques, employing conventional electrodes based on, e.g., Pt, glassy carbon and carbon paste [10–13]. Electroanalytical methods possess a number of advantages over other analytical techniques, such as simplicity and low cost. The analysis has been carried out by anodic stripping voltammetry after cathodic deposition of Au(0) onto the electrode surface [12, 13]. The deposition is often performed at a constant potential, more negative than

984

that of oxygen reduction, which generates species of extreme reactivity. Consequently, the determination of heavy metals often requires the deaeration of the solution by an inert gas, hampering the wide adoption of similar analytical methods as on-line or even at-line procedures. As a further drawback, the stripping process requires the oxidation of Au(0) at potential values high enough to electro-oxidise also many organic species commonly present in a number of matrices. The present study is part of our research on electrode systems different from the most usual ones, suitable to give answers to analytical issues that have been unsatisfactorily approached or that are even unsolved. When working on matrices presenting potential drawbacks to organic electrode coatings, we choose to couple, to organic and to composite electrode systems [14], an analysis of the capabilities of metals very scarcely employed as electrode surfaces. Among unusual metals, Ti was chosen on the basis of its characteristics, such as inertness and proven (electro)catalytic behaviour [15]. In addition, Ti offers the notable advantage to exhibit particularly high overvoltage with respect to oxygen reduction [15], which occurs at potentials either coincident to or even more negative than the background cathodic reduction. This allows the determination of a number of reducible species otherwise suffering from interference of oxygen not only at the level of the relevant cathodic response but also of reactive species that are intermediate or final products of the reduction. It has been shown that a few nanometre thick layer of TiOx is spontaneously formed on Ti surface [16]. This layer consists of Ti(II), Ti(III) and Ti(IV) species and is partially amorphous in nature; the density of oxygen atoms progressively decreases at increasing the distance from the interface with air. It should be evidenced that the nature of crystalline TiO2 [17], widely employed as electrode material in electroanalysis and mainly under the form of particles and nanoparticles, is quite different with respect to Ti oxides spontaneously grown on Ti, constituting the surface of our electrode. TiO2 electrode coatings have been often employed as a part of multicomponent materials, metal nanoparticles or enzymes anchored to the oxide surface and representing the portion of the electrode system actually in charge of interaction with the analytes [18]. The present study focuses the attention on the determination of Au(III) species dissolved in aqueous media by using Ti electrodes. Calibration curves with respect to Au(III) content have been drawn both in pure pH-buffered aqueous solution and in real matrices, namely solutions containing dispersed biomasses. The aim was to mimic the environment in which recovery and monitoring of Au salt do represent a particularly urgent practical problem. Repeatability and reproducibility of the voltammetric signals as well as the linear dependence on the concentration in the range of interest have been assessed: The results obtained demonstrate the effectiveness of the proposed procedure and system.

F. Terzi et al.

Experimental Chemicals were from Sigma. Solutions containing the different metals were prepared from solid AgNO 3 , CuCl 2 , FeNH 4(SO 4) 2, NiCl 2·2H 2O, Zn(CH 3COO)2·2H 2O and Pb(NO3)2. Au solutions were prepared from a 1000 ppm standard solution (from CPI). As to biomasses, 10 g/l wheat bran and sawdust from Picea abies (from Norway spruce) dispersed in 0.1 M HClO4 have been employed. All solutions were prepared using ultrapure water (18 MΩ cm). The electrochemical measurements were performed with an Autolab PGSTAT12 (Ecochemie) potentiostat/galvanostat in a single-compartment three-electrode cell at room temperature. Two-millimetre diameter Ti disks (grade 1) constituted the working electrodes obtained from different stocks of Ti rods by cutting it with cutting machine. They were polished subsequently with 1200 mesh emerit paper; 6, 3 and 1 μm diamond spray (Remet); and 50 nm Al2O3 aqueous dispersion on polishing cloth, then they were sonicated three times for 1 min, changing the solution in the bath before each sonication step. An aqueous saturated Ag/AgCl, KCl sat. (Amel), was the reference electrode; all the potential values given are referred to it. A graphite rod was the auxiliary electrode. Voltammetric traces have been recorded at 50 mV s−1 potential scan rate. Calibration curves have been constructed using data from amperometry at constant potential in quiescent solution and performing the regression by the method of linear least squares. The method of random additions of analyte was adopted, carrying out a gentle polishing of the electrode surface using 50 nm Al2O3 aqueous dispersion on a polishing cloth for 30 s before each addition and rinsing with water. The solution was stirred for 1 min between two subsequent measurements, the working electrode being at open circuit potential. Scanning electron microscope (SEM) images were acquired using a FEI Quanta 200 equipped with Energy Dispersive Spectrometer (EDS, INCA system, Oxford Instruments) and a FEI Nova NanoSEM 450 electron microscopes. Raman spectra were recorded using a Jobin-Yvon, Model LABRAM spectrophotometer equipped with a He–Ne laser (632.8 nm) and a confocal microscope supplied by Olympus (BX 40). Raman investigations have been carried out both in air and in an electrochemical cell coupled to the spectrometer.

Results and discussion Au(III) species, under our experimental conditions, are present in the form of chloro-hydroxyl complexes with different formulations, i.e., AuClxOHy−, where x and y can assume any integer value from 0 to 4. The relative abundances of these species depend on pH of the solution [19]; pH also determines the stoichiometry of the relevant reduction processes [20].

Ti metal electrode as an unconventional amperometric sensor

985

surface. This process is not completely reversible and leads to detrimental effects in terms of repeatability of the Au(III) reduction signal. Not surprisingly, the cathodic limit is mostly conditioned by the solution pH. The results show that a significant decrease of the current intensity due to the reduction of Au(III) species occurs at increasing pH values. Hence, a proper choice of the solvent medium is crucial in the development of effective Ti electrodes: 0.1 M HClO4 has been demonstrated to constitute the best choice among those tested by us. The Au(III) solutions that are most frequently found in processes devoted to recovery of Au may contain significant quantities of additional heavy metals, which constitute potential interfering species. The influence on Au(III) reduction of most common metal species, namely Ag(I), Cu(II), Fe(III), Ni(II), Zn(II) and Pb(II), has been investigated by cyclic voltammetry and amperometry at constant potential. As to cyclic voltammetry, the presence of Fe(III), Ni(II), Cu(II) and Zn(II) leads to forward potential sweep traces similar to that registered in the background solution (Fig. 2); only Ag(I) and Pb(II) seem to constitute potential interfering ions in Au(III) determination. As a further step of the interference study, binary mixtures containing the same concentration (1.00×10−5 M) of Au(III) and of the supposedly interfering metal have been tested. The

On the other hand, both acid and basic functionalities, whose formulation and surface concentration are in turn pH-dependent, have been reported to be present on Ti surface [21]. As an example, at the solution/metal interface, Ti(II), Ti(III) and Ti(IV) species, bridged by oxygen atoms and by –OH moieties, have been reported. It is obvious that the presence of similar species affects the electrochemical behaviour of Ti as an electrode. Furthermore, the electrochemical responses are conditioned by the presence of high concentrations of anions in solution, as it is the case of buffered aqueous solutions; anions, in fact, can be adsorbed on the Ti surface [22]. The analysis of the behaviour of a Ti electrode in the specific solutions to study may lead to quite different results as a function of pH: Voltammetric investigations have been carried out at pH 1, 4, 7 and 10 realised in 0.1 M perchloric acid, citrate, phosphate and borate buffer, respectively. The voltammetric responses of a Ti electrode in these solutions are shown in Fig. 1. The addition of Au(III) at a concentration of 1.00×10−5 M leads to a significant increase of the cathodic currents, no well-defined current peak being however detectable. Steady-state voltammograms, shown in the figure, are achieved at the third potential scan. The potential range has been selected taking into account that a significant increase of the current is recorded at values higher than +0.30 V due to the oxidation of the electrode

Background Au(III)

-0.4

-0.2

0.0

0 -500

I / nA

500

0 500 -2000

I / nA

B

-1000

A

0.2

Background Au(III)

-0.4

-0.2

E/V

0.0

0.2

E/V

D

Background Au(III)

-0.4

-0.2

0.0 E/V

0.2

-200 -400

I / nA

-400

I / nA

0

200

0 200

C

-800

Fig. 1 Voltammetric curves on 1.00×10−5 M Au(III) aqueous solutions containing 0.1 M HClO4 (A), citrate (B), phosphate (C) and borate (D) buffers and H2SO4 (E). Third scans are shown. The bullet points indicate the starting open circuit potential value. The initial potential scan is toward more negative potential values

Background Au(III)

-0.4

-0.2

0.0 E/V

0.2

986

F. Terzi et al.

Fig. 2 Voltammetric traces recorded in solutions containing 1.00×10−5 M Zn(II) (A), Cu(II) (B), Fe(III) (C), Ni(II) (D), Ag(I) (E) and Pb(II) (F) in 0.1 M HClO4. The trace on 1.00× 10−5 M Au(III) is also reported for comparison. Third scans are shown. The bullet points indicate the starting, open circuit potential value. The initial potential scan is toward more negative potential values

presence of Fe(III), Ni(II) and Zn(II) leads to voltammograms similar to that obtained on a solution containing only Au(III) species; on the contrary, solutions containing Ag(I), Cu(II) and Pb(II) exhibit potential interference. The case of Pb(II) represents the most critical situation, the current values being comparable to that of Au(III) alone; in addition, trace crossing of forward and backward scan currents is observed, as an indication of the occurrence of complex surface processes. A careful choice of the potential at which to polarise the electrode in amperometric tests at constant potential is necessary in order not to suffer from metal interference.

Amperometry at constant potential was actually preferred to cyclic voltammetry in order to define the electroanalytical procedure for the determination of Au(III) with Ti electrode in the presence of additional metal ions. On the basis of the voltammetric responses, the suitable working potential for the amperometric measurements was chosen not to be more negative than −0.10 V: Meaningful signal due to Ag(I), Cu(II) and Pb(II) reduction is observed past such a value. On the other hand, a significant current due to the oxidation of the electrode surface is observed in background solution operating at a constant potential value higher than +0.10 V. As a consequence, suitable potential values range from −0.10 to +0.10 V.

Ti metal electrode as an unconventional amperometric sensor

987

Within this potential range, three values have been tested, namely −0.10, 0.00 and +0.10 V in order to estimate the effect of the interference of the considered metals on the Au(III) reduction signal. We verified that repeatability and reproducibility of the measured current can be obtained at a polarisation time of 7 s. Moreover, best repeatability of the current values is gained when three similar polarisation steps are subsequently carried out by stirring of the solution for 1 min at open circuit potential between two subsequent measurements. This procedure has been adopted for the study of the possible interferences, as well as in the construction of the calibration curve discussed below. Fe(III), Cu(II), Ni(II), Pb(II) and Zn(II), when separately tested, do not affect the background signal at any of the indicated potentials. 1.40×10−5 M Ag(I) leads to a signal comparable to that of 1.00×10−5 M Au(III). However, when solutions containing the same concentration (1.00×10−5 M) of Au(III) and of the supposedly interfering metal where tested, interference of Ag(I) is not evident. A quantitative estimation of the interference effect can be given on the basis of the selectivity coefficient K proposed by Raluca-Ioana et al. for amperometric tests [23]. The results reported in Table 1 show that Ti electrodes polarised at +0.10 V exhibit high selectivity coefficients for Au(III) determination, with respect to all the considered interfering metals. This potential value has been selected for the construction of the calibration curves. As it was specified in the experimental section, quiescent solutions have been employed for amperometry at constant potential. In principle, the use of a rotating disk electrode could lead to better performance in terms of sensitivity and limit of detection. Actually, we ascertained that, as a counterpart of only slight higher electrochemical responses, lower reproducibility is gained, presumably due to interference of the hydrodynamic conditions on the reduction/deposition of Au on the Ti electrode substrate, which will be discussed below. Table 1 Effect of the interfering species Interfering metal

|K|a −0.10 V

0.00 V

+0.10 V

Ag Cu

0.26 0.19