505

Short Communication

Elimination of the Copper-Zinc Interference at Mercury Film Electrodes by a Na®on=Clay Modi®ed Layer Jyh-Myng Zen,* Hong-Ying Lin, and Hsueh-Hui Yang Department of Chemistry, National Chung-Hsing University, Taichung 402, Taiwan; e-mail: [email protected] Received: July 24, 2000 Final version: August 23, 2000 Abstract

The construction of a Na®on=clay-modi®ed mercury electrode was shown to successfully eliminate the interference of Cu2‡ in the stripping determination of Zn2‡. The modi®ed electrode discriminated towards deposition of Cu2‡ during the electrolytic deposition of Zn2‡ thus it avoided the formation of the intermetallic compounds at mercury. Square-wave stripping peak currents after 1 min of electrolysis at 1.2 V revealed that the Zn-stripping signal obtained with the modi®ed electrode was linear over two ranges of 0±5 and 6±12 mM for Zn2‡ in the presence of 10 mM Cu2‡. The modi®cation procedure is reproducible and the resulting attachment is stable. The analytical utility is illustrated by the measurement of Zn2‡ in real water samples. Keywords: Zn2‡, Stripping, Mercury electrode, Cu2‡, Clay, Na®on

Anodic stripping voltammetry is an attractive method for determining trace concentrations of a number of heavy metal ions in solution [1, 2]. It also allows metal species characterization, fully or partially, depending on the complexity of the system. However, a major limitation of stripping analysis is that the responses are often complicated by the formation of intermetallic compounds and the presence of overlapping peaks [3±19]. During the reductive electrolysis of the solutions containing multicomponents, the reduced analytes are generated in the mercury electrode. The intermetallic compounds are formed either by the mutual interaction of reduced=amalgamated analytes or by the combination of such analytes with mercury. Thus, reliable analytical results can not be derived due to the responses which may be depressed or the stripping potential may be shifted. The interference from Cu2‡ in the determination of Zn2‡ at a mercury-based electrode was reported in 1958 [3]. As shown in Figure 1, the peak current for a constant Zn2‡ concentration decreases upon addition of Cu2‡ at a thin mercury ®lm electrode and a new peak appears which is due to the formation of the Zn-Cu intermetallic compounds. A number of studies have been focused on the elimination of this interference since Cu2‡ is a common ion in variety of analytical samples. Typical approaches can be summarized as follows: ± the use of a hanging mercury drop electrode instead of a thin mercury ®lm electrode [18], ± the addition of elements or molecules to remove the response of Cu2‡ by the formation of a more stable intermetallic compound, metal salt or metal complex [4±11], ± the use of solvent extraction or ion exchange to decrease the concentration of Cu2‡ [12, 13], ± the use of dual electrodes in a thin-layer cell to physically separate Cu2‡ on different electrodes [14, 15], and ± the derivation of an analytical expression for the relationship between the response and concentration based on a hypothesized mechanism [16±19]. Although these methods are often successful, there are disadvantages such as the requirement of sample pretreatment, complicated instrument, knowledge of the stoichiometry of the intermetallic compound and longer analysis time. Electroanalysis 2001, 13, No. 6

Recently, Mogensen and Kryger successfully used chemically modi®ed electrodes to diminish the in¯uence of Cu2‡ on the stripping response of Zn2‡ [6]. The electrodes were constructed from a glassy carbon electrode, which was ®rst coated with a

Fig. 1. Square-wave stripping voltammogram after 1 min of electrolysis at 71.2 V for 10 mM Zn2‡ and a) 0 mM, b) 2 mM, c) 4 mM, and d) 6 mM Cu2‡ in 0.01 M KNO3 solution on a thin mercury ®lm electrode.

# WILEY-VCH Verlag GmbH, D-69469 Weinheim, 2001

1040-0397/01/0604±0505 $17.50‡.50=0

506

Ê zeolite 13X-containing polymer matrix then exposed to the 10 A electrolytic deposition of mercury. Due to the ion exchange Ê zeolite capacity of the zeolite and the different af®nity of the 10 A to different size and shape of hydrated Zn2‡ and Cu2‡, the modi®ed electrode could discriminate against Cu2‡ during the electrolytic deposition of Zn2‡. They reported that only small amount of the accumulated Zn2‡ was stripped as the intermetallic compound and the potential of which was more negative than the copper stripping potential. We report here the construction of a novel modi®cation by Na®on polymer and nontronite clay to eliminate the interference of Cu2‡ in the determination of Zn2‡ at mercury-based electrodes. Nontronite is a smectite clay with the chemical formula ‡ of [(Si7.25Al0.75)(Fe3‡ 2:75 Al0.85Mg0.33Ti0.05)O20(OH)4X0:98 )] [20]. X-ray diffraction patterns con®rmed that the d001 interplanar Ê, distances (i.e., the basal spacing) of the coated clay is 12.9 A Ê which is comparable to the 10 A zeolite 13X. Nontronite clay not only has the characteristics of ion exchange similar to zeolite; it also has intercalation properties and appreciable surface area. Thus, it was expected to discriminate against Cu2‡ during the electrolytic deposition of Zn2‡ better than zeolite does. Na®on has been used for electrode coatings in a variety of electrochemical studies, mostly in conjunction with the immobilization of positively charged redox couple within the ®lm due to the good cation exchange ability and mechanical stability [21, 22]. The Na®on=nontronite modi®ed mercury electrode, designated as NNMME, can then easily be prepared by the same procedure as that in the preparation of thin mercury ®lm electrodes. The NNMME is expected as an inexpensive and selective method for the determination of Zn2‡. The NNMMEs were prepared either by the sandwich method or composite method. These procedures are described in the experimental section. Figure 2 shows the stripping curves after 1 min of electrolysis at 1.2 V for 10 mM of Zn2‡ and 0±6 mM of Cu2‡ on the NNMME, prepared by composite method (Figure 2A) and sandwich method (Figure 2B), respectively. Compared

J. Zen et al.

to Figure 1, the addition of Cu2‡ did not change the peak current for a constant concentration of Zn2‡ and there was no new peak formation due to the intermetallic compounds on the sandwich NNMME. However, the composite NNMME did not eliminate the intermetallic interference as effective as the sandwich NNMME. The incomplete cover of the electrode by the clay in composite method may possibly make the electrode less size selective. Thus, both metal ions can undergo reduction and form intermetallic compound. Therefore, the sandwich NNMMEs were chosen for the following experiments. Further investigation was made to the transport characteristics of Zn2‡ on the NNMME. The linear sweep voltammograms of Zn2‡ were taken at different scan rates (50 mV=s to 450 mV=s). The linear relationship between peak current and scan rate indicated the transport characteristic of Zn2‡ on the NNMME is adsorptioncontrolled. To test the suitability of the NNMME for the quantitative determination of Zn2‡ in the presence of Cu2‡, the studies were carried in solution with a constant concentration of Cu2‡ and progressively increasing concentrations of Zn2‡ and vice versa for Zn2‡ and Cu2‡. The results are given in Figure 3. Two things can be pointed out. First, for a constant concentration of Cu2‡ and Zn2‡, the responses did not change upon addition of the other metal ion. Second, the responses of copper and zinc increase linearly with the concentration. Linear calibration plots are obtained over two ranges of 0±5 and 6±12 mM for Zn2‡ in the presence of 10 mM Cu2‡. These results imply no intermetallic compound formation on the NNMMEs. Thus, the NNMMEs are suitable for the quantitative determination of Zn2‡ in the presence of Cu2‡ or the quantitative determination of Cu2‡ in the presence of Zn2‡. The reproducibility of the NNMMEs, repetitive measurementregeneration cycles were carried out in 10 mM Zn2‡. The NNMMEs can be easily renewed by soaking the electrodes in 0.1 M HNO3 for 30 s. The results of 10 successive repetitive measurement-regeneration cycles showed a small coef®cient of variation of 1.8 % for the detection of 10 mM Zn2‡ using the NNMMEs.

Fig. 2. Square-wave stripping voltammogram after 1 min of electrolysis at 71.2 V for 10 mM Zn2‡ and a) 0 mM, b) 2 mM, c) 4 mM Cu2‡, and d) 6 mM in 0.01 M KNO3 solution on A) the composite NNMME and B) the sandwich NNMME. Electroanalysis 2001, 13, No. 6

Copper-Zinc Interference at Mercury Film Electrodes

507

Fig. 3. Square-wave stripping peak currents after 1 min of electrolysis at 71.2 V for Zn2‡ and Cu2‡ in 0.01 M KNO3 solution on the sandwich NNMME A) 10 mM Cu2‡ and 0±12 mM Zn2‡; B) 10 mM Zn2‡ and 0±6 mM Cu2‡.

The practical analytical utility of the NNMMEs was illustrated by selective measurements of Zn2‡ in three water samples without any preliminary treatment as shown in Table 1. The accuracy of the method was determined by its recovery on spiking. Note that the amount of Cu2‡ was 0.7 mM in groundwater as reported in our previous study [23]. The recoveries of Zn2‡ from water samples were satisfactory with values ranging from 97.1 to 101.8 %, con®rming that quantitative and reproducible results can be obtained with this method. In summary, the present study demonstrates the NNMMEs permit the analysis of mixtures of copper(II) and zinc(II) by avoiding the formation of the intermetallic compounds. The electrode can be regenerated easily by soaking in 0.1 M HNO3 for 30 s. The modi®cation procedure is reproducible and the resulting attachment is stable. The recoveries of spiked zinc(II) were observed to be good in water samples. Based on this work, it is evident that compared to the previous schemes the NNMMEs have the advantages such as the elimination the requirement of sample pretreatment, complicated instrument, knowledge of the stoichiometry of the intermetallic compound and longer analysis time.

Experimental Clay colloids were prepared in the sodium form generally following the procedures previously described [24±27]. In brief, the clay was stirred in 1 M NaCl for 48 h to convert to the sodium form. After centrifugation, it was repeatedly washed ®rst with 50 % alcohol then with 95 % alcohol until a negative chloride test was obtained. The fractions were then separated by centrifugation and freeze-dried. The NNMMEs were prepared either by the sandwich method or composite method. These procedures were described as following. In the sandwich method, the glassy carbon electrodes (GCEs) were modi®ed with clay and followed by Na®on. The clay-modi®ed GCEs were prepared by dropping 6 mL of a clay colloid (0.5 g=L) onto a clean GCE and dried under ambient conditions, usually ca. 0.5 to 1 h. Uniform ®lms could be cast reproducibly from clay. Following the clay modi®cation, 4 mL of Na®on solution (4 wt. %) was spin-coated on the surface of the clay-modi®ed GCE at 3000 r.p.m. A uniform thin ®lm was formed after about 3 min of spinning. As for the composite modi®cation, 4 mL of the coating solution, containing the mixture of Na®on and clay was spin-coated on the GCE surface at

Table 1. Zn(II) assay in three water samples using the NNMME. The number of samples assayed were three.

Linear equation R Detected value (mM) Recovery (%)

Groundwater [a]

Tap water

Lake water

ipa ˆ 2.65 ‡ 1.06[Zn(II)] 0.9948 2.48 101.8

ipa ˆ 0.59 ‡ 1.56[Zn(II)] 0.9992 0.38 97.1

ipa ˆ 1.11 ‡ 1.84[Zn(II)] 0.9918 0.60 99.5

[a] [Cu(II)] ˆ 0.7 mM [23]. Electroanalysis 2001, 13, No. 6

508

3000 r.p.m. These electrodes were subsequently rinsed with water and then coated with mercury by electrolysis ( 400 mV vs. Ag=AgCl) of a 10 mg=mL solution of Hg(NO3)2 in 0.01 M HCl. All voltammetry experiments were performed with the Bioanalytical systems (West Lafayette, IN) BAS-50W electrochemical analyzer. A BAS Model VC-2 electrochemical cell was employed in these experiments. The three-electrode system consisted of a thin ®lm mercury electrode or a NNMME as a working electrode, a Ag=AgCl reference electrode (Model RE-5, BAS), and a platinum wire auxiliary electrode. Prior to electrolysis, all solutions were deoxygenated with argon for 10 min. Electrolytic depositions of Zn2‡ and Cu2‡ were carried out for 60 s at 1200 mV (vs. Ag=AgCl). The square-wave voltammetry with a 5 mV step height, 50 mV amplitude, and 20 Hz frequency was used for the determination of Zn2‡ and Cu2‡.

Acknowledgement The authors gratefully acknowledgement ®nancial support from the National Science Council of the Republic of China.

References [1] J. Wang, Stripping Analysis: Principles, Instrumentation and Applications, VCH, Deer®eld Beach, FL 1985. [2] K. Brainina, E. Neyman, Electroanalytical Stripping Methods, in Chemical Analysis, Vol. 126 (Ed: J.D. Winefordner), Wiley, Chichester 1993. [3] W. Kemula, Z. Galus, Z. Kublik, Nature 1958, 182, 1228. [4] O.-W. Lau, O.-M. Cheng, Anal. Chim. Acta 1998, 376, 197.

Electroanalysis 2001, 13, No. 6

J. Zen et al. [5] E.Y. Neiman, L.G. Petrova, V. I. Ignatov, G.M. Dolgopolva, Anal. Chim. Acta 1980, 113, 277. [6] L. Mogensen, L. Kryger, Electroanalysis 1998, 10, 1285. [7] G. Schulze, W. Frenzel, Fresenius Z. Anal. Chem. 1983, 314, 459. [8] A. Izquierdo, M.D. Luque de Castro, M. Valcarcel, Electroanalysis 1994, 6, 894. [9] C.M.A. Brett, M.B.Q. Garcia, J.L.F.C. Lima, Anal. Chim. Acta 1997, 339, 167. [10] A.L.B. Marques, G.O. Chierice, Talanta 1991, 38, 735. [11] M.S. Shuman, G.P. Woodman, Anal. Chem. 1976, 48, 1979. [12] S.V. Psaroudakis, C.E. Efstathiou, Analyst 1988, 114, 25. [13] R.R. Dalangin, H. Gunasingham, Analyst 1994, 119, 2187. [14] J. Wang, H.D. Dewald, Anal. Chem. 1983, 55, 933. [15] E. Beinrohr, P. Tschopel, G. Tolg, M. Nemeth, Anal. Chim. Acta 1993, 273, 13. [16] E. Sahlin, D. Jagner, Electroanalysis 1998, 10, 532. [17] H. Chan, A. Butler, D.M. Falck, M.S. Freund, Anal. Chem. 1997, 69, 2373. [18] B. Lazar, A. Nishri, S. Ben-Yaakov, J. Electroanal. Chem. 1981, 125, 295. [19] R.W. Gerlach, B.R. Kowalski, Anal. Chim. Acta 1982, 134, 119. [20] W.F. Jaynes, J.M. Bigham, Clays Clay Miner. 1987, 35, 440. [21] J.-M. Zen, G. Ilangoven, J.-J. Jou, Anal. Chem. 1999, 71, 2797. [22] J.-M. Zen, S.-H. Jeng, H.-J. Chen, Anal. Chem. 1996, 68, 498. [23] J.-M. Zen, N.-Y. Chi, F.-S. Hsu, M.-J. Chung, Analyst 1995, 120, 511. [24] J.R. White, A.J. Bard, J. Electroanal. Chem. 1986, 197, 233. [25] C.M. Castro-Acuna, F.-R.R. Fan, A.J. Bard, J. Electroanal. Chem. 1987, 234, 347. [26] R.D. King, D.G. Nocera, T.J. Pinnavaia, J. Electroanal. Chem. 1987, 236, 43. [27] G. Villemure, A.J. Bard, J. Electroanal. Chem. 1990, 282, 107.