Analytica Chimica Acta 421 (2000) 189–197

Determination of lead(II) on a copper/mercury-plated screen-printed electrode Jyh-Myng Zen∗ , Hsieh-Hsun Chung, Annamalai Senthil Kumar Department of Chemistry, National Chung-Hsing University, Taichung 402, Taiwan Received 7 February 2000; received in revised form 6 June 2000; accepted 21 June 2000

Abstract A Cu/Hg-plated screen-printed electrode (CMSPE) was developed for trace Pb(II) determination by square-wave anodic stripping voltammetry. The intermetallic property of lead with copper in mercury was utilized to increase the sensitivity of Pb(II) detection. The calculated transfer coefficient value for the Pb(II)/Pb redox process on the CMSPE is ∼0.5 representing the ideality of the system for Pb(II) detection. Systematic investigations were made to optimize the Cu–Hg plating condition on SPE towards the Pb(II) signals in terms of Cu/Hg ratio, plating potential, and plating time. The optimum conditions for electrochemical and square-wave parameters were also evaluated. The calibration curve was linear in the range of 0–100 ppb Pb(II) with a detection limit of 0.81 ppb (S/N = 3) for 60 s preconcentration time. The analytical utility of the CMPSE was demonstrated by applying to various water samples. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Lead(II); Screen-printed electrode; Square-wave voltammetry

1. Introduction Determination and detection assays of Pb(II) are a continuous research interest due to its toxicity to the environment and biological livings [1]. The most often used way to get sensitive Pb(II) signal is to detect it by anodic stripping voltammetry (ASV) on a Hg-based electrode through the formation of Pb-amalgam. As reported earlier, the formation of simple binary mixtures, so called intermetallic compounds, have presumable influence in altering the stripping signals [2,3]. In general, the suppression of the Pb(II) signal through the formation of M–Pb (where M = Cu, Au, Ag, etc.) intermetallic compounds appears to be the most commonly reported observation [4]. On the other hand, Dong et al. noticed 50% increase in the Pb(II) ASV ∗ Corresponding author. Fax: +886-4-2862547. E-mail address: [email protected] (J.-M. Zen).

signal in the presence of Cu(II) above 10−6 M concentrations [5]. Gunasingham et al. studied the necessary influencing parameters to achieve maximum Pb(II) signals by using ASV coupled with wall-jet detecting technique on a Cu–Hg film electrode [6]. However, the interference effect of surface-active compounds and common ions was not mentioned. Our group has further demonstrated the excellent sensitivity and selectivity towards the Pb(II) detection using the intermetallic property of Cu–Pb binary mixtures on a Nafion/copper–mercury film electrode [7]. Yet, this method is not suitable for routine analysis due to the necessity of forming fresh Cu–Hg plating in each determination. To overcome this drawback, we developed a new class of copper/mercury-based disposable screen-printed electrode (designated as CMSPE) for the sensitive detection of Pb(II). Disposable SPEs serve versatile commercial applications with low cost and easy preparation for analytical assays [8–12]. In

0003-2670/00/$ – see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 3 - 2 6 7 0 ( 0 0 ) 0 1 0 5 2 - 7

190

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

this paper, the preparation of the CMSPE and the necessary electrochemical and instrumental parameters were systematically optimized. Square-wave voltammetry (SWV) was employed in this study because it provides a fast response and high sensitivity. Additionally, the SWV enables the simultaneous inspection of both the reduction and oxidation processes, which provides considerable insight to the mechanism of the electrode reaction. Finally, The analytical utility of the CMSPE was demonstrated for the Pb(II) detection in various water samples.

2. Experimental 2.1. Chemicals and apparatus Standard solutions of Cu(II), Hg(II), and Pb(II) (1000 mg/l, AAS grade) were bought from Merck. The other standard metal solutions used in the interference studies were also obtained from Merck. A solution of 0.02 M, pH 2 acetate buffer was prepared from Merck Suprapur reagents and used in all electrochemical experiments. All the other compounds (ACS-certified reagent grade) were used without any further purification. Aqueous solutions were prepared with doubly distilled deionized water. Voltammetric measurements were carried out with a CHI Model 660 electrochemical workstation (Austin, TX, USA). The three-electrode system consists of either a SPE or CMSPE working electrode, an Ag/AgCl reference (Model RE-5, BAS), and a platinum auxiliary electrode. 2.2. Electrode fabrication A manual screen-printer was used to produce the disposable SPEs. A stencil with structure of 16 continuous electrodes was used to screen-print the conducting carbon on a flexible polypropylene film (50 mm × 70 mm). After the coating of carbon layers, the unit was cured under UV irradiation at intensity of 1.85 mW/cm2 for 2 h. After drying, an insulator layer was spread manually over the SPE leaving the working area of 0.2 cm2 with a conductive track dimension of 8 mm × 2.5 mm. The resistance value of the bare film tracks (Rf ) were determined using two-point probe digital multimeter and the average Rf

for the 16 strips is 166.3 ± 8.9 /cm, which is ∼six times lower than that of a recent report [13]. 2.3. Analytical procedure The SPE was first equilibrated in acetate buffer solution for about 10 min before electrochemical experiments. It was then pretreated by continuous scans in the window of −0.5 to 0.5 V at a scan rate 50 mV/s until a stable background current obtained. The Cu–Hg film was electro-deposited on the SPE in an appropriate plating solution and potential to prepare the CMSPE. The amount of Pb(II) was detected quantitatively using SWV at the CMSPE. The potential range was set form −1.0 to −0.3 V in the anodic direction for most cases. The standard addition method was adopted to evaluate the Pb(II) content in the water samples.

3. Results and discussion 3.1. Electrode optimization Linear scan voltammetric (LSV) responses of a bare SPE, Cu-plated SPE, Hg-plated SPE, and the CMSPE in 0.02 M, pH 2 acetate buffer were first studied. No signal was noticed on the bare SPE indicating the absence of any Faradaic processes in the operating condition; whereas, discrete anodic half wave potentials (E1/2 ) of 0.15 and 0.60 V were noticed on the Cu-plated SPE and Hg-plated SPE, respectively. The obtained E1/2 were close to the standard redox potential of Cu(II)/Cu and Hg(II)/Hg, which is 0.12 and 0.57 V, respectively [14]. Obviously, both Cu and Hg were properly plated on the SPE under the present experimental conditions. Further evidence can be seen form the LSV response of the CMSPE in the same experimental condition, where three peaks at E1/2 around 0.15, 0.60, and 0.75 V were observed over a huge background current. The first two peaks match with the Cu(II)/Cu and Hg(II)/Hg redox transitions and the E1/2 at 0.75 V is the Cu–Hg amalgam peak response. The [Cu(II)/Hg(II)] ratio in preparing the CMSPE is key to get effective Pb(II) detection signals. Fig. 1A shows the square-wave anodic stripping voltammetry (SWASV) responses of 50 ppb Pb(II) at the CMSPE prepared in various [Cu(II)/Hg(II)] ratio based on xCu(II) + yHg(II), where x = 0–250 ppm and

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

191

Fig. 1. (A) SWASV responses of 50 ppb Pb(II) in 0.02 M, pH 4.22 acetate buffer on the CMSPE at different [Cu/Hg] ratios. Cu concentration was varied over fixed Hg concentration of 50 ppm. Plating potential = –1.0 V; plating time = 200 s. SWV conditions: f = 50 Hz, E amp = 50 mV, and E step = 4 mV. (B) Typical ipa and Epa response against the [Cu/Hg] ratio.

y = 50 ppm. The stripping peak current (ipa ) and peak potential (Epa ) obtained was plotted against the [Cu(II)/Hg(II)] ratio as shown in Fig. 1B. As can be seen, the SWASV responses for Pb(II) increase as the [Cu(II)/Hg(II)] ratio increases until a maximum at the [Cu(II)/Hg(II)] ratio of 2. The obtained Epa was about 200 mV more negative than the standard E0 of Pb(II)/Pb, which again evidenced the formation of complex Pb(II) matrixes in the present case [14]. The shift of Epa in anodic direction with increase in [Cu(II)/Hg(II)] ratio indicates the proper formation of the amalgamated Cu–Hg. When [Cu(II)/Hg(II)] > 2, the abrupt shift in Epa indicates the Hg-based electrode has switched to a Cu-based electrode which in turn leads to the lowering in ipa . The broadening Pb(II) signal at higher [Cu(II)/Pb(II)] ratios can also be attributed to the coexistence of the Cu–Pb intermetallic solid solution on the CMSPE [7]. The maximum ipa measured is about three times higher than that at a pure Hg-plated SPE. Obviously, the presence of Cu can indeed increase the anodic stripping signal of Pb(II).

Under the optimum [Cu(II)/Hg(II)] ratio of 2, we expect ∼50% of the total Cu were converted to Cu–Hg amalgam on the CMSPE. During the Pb(II) accumulation step, the free Cu present in the CMSPE involved in the intermetallic formation as follows (Cu–Hg) · mCu + nPb2+ + 2e → (Cu–Hg) · Cum Pbn where (Cu–Hg)Cum Pbn is the Cu–Pb intermetallic compound formed on the CMSPE. This (Cu–Hg)Cum Pbn compound in turn yields the enhanced ipa signal during the stripping process in SWASV. The three-fold increased signal over the Hg-plated SPE clearly indicates the favorable and feasible stripping of Pb(II) from the (Cu–Hg)Cum Pbn intermetallic compound than the Hg–Pb amalgamated one. This result indirectly evidenced the formation of somewhat weaker Cu–Pb bond on the (Cu–Hg)Cum Pbn than the Hg–Pb. Overall, a platting solution with the [Cu(II)/Hg(II)] ratio of 2, i.e. 50 ppm Hg(II) and 100 ppm Cu(II) in 0.02 M, pH 2 acetate buffer, was used in all subsequent experiments.

192

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

Fig. 2. (A) SWASV responses of 50 ppb Pb(II) in 0.02 M, pH 4.22 acetate buffer on the CMSPE at different plating potentials. Solution [Cu/ Hg] = 2 and plating time = 200 s. SWV parameters are the same as in Fig. 1A. (B) Typical ipa and Epa response against the plating potential.

Following the optimization of plating concentration of Cu(II) and Hg(II), we further investigated the effect of plating potential and plating time on the detection of Pb(II). Fig. 2 shows the typical SWASV response of 50 ppb Pb(II) at the CMSPE prepared at different plating potential. As can be seen, the maximum stripping ipa was observed at −0.8 V and this value was hence used in all subsequent experiments. Interestingly, the obtained Epa also follows almost the same trend against the plating potential as that observed in ipa . The accumulation forces and surface interaction parameters controlled by the applied potential may lead to these alternations. The effect of plating time was studied towards the detection of 50 ppb Pb(II). As shown in Fig. 3, by fixing the optimal plating solution composition as 50 ppm Hg(II) and 100 ppm Cu(II), continuous increase in the SWASV signals against the plating time were observed up to 100 s. Meanwhile, the Epa and the measured background currents were found to shift and increase simultaneously with the plating time. These results clearly indicate the continuous growth of the Cu–Hg on the SPE. Around

six-fold increase in ipa were noticed on the relatively thick CMSPE, which is coincident with the earlier observation of more significant peaks in thick films than thin films [6,15]. 3.2. Optimal conditions in detection The effect of preconcentration potential (Pp ) on the SWASV response of Pb(II) at the CMSPE was shown in Fig. 4A. The SWASV signals against Pp started to increase from −0.7 to −1.0 V and displayed an S-shape result. Fig. 4B is the effect of preconcentration time (tp ) for Pb(II) concentration of 10, 30, and 50 ppb in 0.02 M, pH 2 acetate buffer solution. As can be seen, the effect of tp is very significant on the SWASV signals. For 50 ppb Pb(II), a sharp increase in ip was noticed up to 200 s; whereas, for lower concentrations of Pb(II), longer tp is required for ip to level off. For convenience, P p = −1.0 V and t p = 60 s were chosen in most of the subsequent studies. The investigated SWV parameters are frequency (f, Hz), step height (Estep , mV), and amplitude (Eamp ,

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

193

Fig. 3. (A) SWASV responses of 50 ppb Pb(II) on the CMSPE at different plating time. Plating potential = −1.0 V and other conditions are the same as in Fig. 2A. (B) Typical ipa and Epa response against the plating time.

mV). These parameters are interrelated and have a combined effect on the response towards Pb(II) signal as shown in Fig. 5. The response for Pb(II) was found to increase with SW frequency, but at frequencies higher than 60 Hz slopping background current renders the measurement difficult. Similarly, the increase in amplitude causes an increase in the Pb(II) signal up to 60 mV. The scan increment together with the frequency defines the effective scan rate; hence, increase in either the frequency or the pulse increment results in an increase in the effective scan rate. Overall, the best signal to background current characteristics are as follows: Eamp , 60 mV; f, 60 Hz; effective scan rate, 360 mV/s. Before going into the analytical characterization, the ideality of the (Cu–Hg)Cum Pbn film towards the Pb(II) detection was evaluated. In SWASV, the peak potential of an irreversible oxidation reaction linearly depends on the logarithm of the SW frequency with the slope ∂Epa /∂log(f ) value of 2.303RT/βnF, for which β is the transfer coefficient of oxidation [16,17]. Note that in the present case the peak current ratio (if /ir ) between the forward (if ) and reverse (ir ) scans

is close to 2. As shown in the inset of Fig. 5A, for 50 ppb Pb(II) with E amp = 50 and E step = 4 mV, the slope of ∂Epa /∂log(f ) was 52.8 mV/Hz corresponding to βn = 1.12. Since n = 2, β is therefore equal to 0.56. Meanwhile, β can also be calculated from the SWV half-peak width (1E1/2 ), based on that the 1E1/2 for the irreversible oxidation reaction is equal to 63.5/βn [16,17]. The 1E1/2 value for the 50 ppb Pb(II) at f = 50 Hz, E amp = 50, and E step = 4 mV was found to be 64 mV corresponding to the βn value of 0.99. This calculated β value of 0.50 is fairly close to the previous one. Furthermore, the slope of ∂Epa /∂E amp = −1.3 (Fig. 5C), is also comparable with the theoretically assigned value of −1 for the irreversible reactions [16]. These results evidenced that the potential energy barrier of the Pb(II)/Pb redox couple is symmetry, and so the (Cu–Hg)Cum Pbn film is ideal for the detection of Pb(II) by SWASV. 3.3. Analytical characterization Under the optimized experimental conditions, calibration data were obtained for Pb(II) on the

194

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

Fig. 4. The Pp (A) and tp (B) effect to the SWASV response of 50 ppb Pb(II) in 0.02 M, pH 4.22 acetate buffer on the CMSPE. (A) t p = 200 s and (B) P p = −1.0 V. Plating potential = −0.8 V and plating time = 200 s. SW conditions: f = 60 Hz, E amp = 60 mV and E step = 6 mV.

Fig. 5. Optimization of SWV parameters: (A) f, (B) Estep , and (C) Eamp for 50 ppb Pb(II) in 0.02 M, pH 4.22 acetate buffer. Insert plots in (A) & (C) are (Epa vs. log f) and (Epa vs. Estep ), respectively.

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

CMSPE. In all cases, an anodic stripping response was noticed at a potential near −0.56 V for SW voltamograms on the CMSPE for 0, 10, 20, 40, 60, 80, and 100 ppb Pb(II), respectively. The observed peak currents were then used for the construction of the calibration plot with slope (␮A/ppb), intercept (␮A), and correlation coefficient of 0.26, 0.15, and 0.9984, respectively. The linear range is ranging from 0 to 100 ppb with a detection limit of 0.81 ppb (S/N = 3) for 60 s of preconcentration time. Note that the detection limit could be even lower if the tp is longer than 60 s as indicated in Fig. 5B. Moreover, this detection limit was comparable with earlier result on the Nafion/copper-mercury film electrode (0.08 ppb with t p = 150 s), poly(4-vinylpyridine)/mercury film electrode (0.3 ppb with t p = 30 s), and electrochemically activated glassy carbon electrode (0.7 ppb with t p = 100 s) [7,18–20]. In the present case, since Pb(II) detection assay is based on disposable screen-printed electrode, we believe it is more variable in application for environmental and clinical samples. Metal ions of various groups were examined concerning their interference in the determination of Pb(II) as shown in Table 1. For 20 ppb Pb(II) with t p = 60 s, the results showed that over 10-fold excess concentration of Ca(II) showed totally nil interfering effect; while, the d-transition elements, such as, Ti(II), Mn(II), Ni(II), Zn(II), and Hg(II) shows acceptable tolerance effect. On the other hand, the main group (III–V) elements, such as, Tl(III), Sn(III), and Bi(III) showed considerable influencing effect, especially Tl(III) and Bi(III) in the [Xe]6s2 4f14 5d10 6pn (n = 1–3 for Tl, Pb, and Bi, respectively) series have major effect in the detection signal. Since Pb(II) is from main group IV origin, it is reasonable that the neighboring metals from the same origin have more influencing effect. Nevertheless, since the amount of Tl(III) and Bi(III) that occurs in natural waters is extremely low, they are not supposed to cause any interference in real sample analysis and this is indeed the case as will be discussed later. SWASV responses recorded for continuous and successive 20 measurements for 20 and 50 ppb Pb(II) on the CMSPE under the optimized condition showed nearly unchanged in the detecting signals with the R.S.D. of 4.81 and 1.11%, respectively. Such performance of the CMSPE indicates effective regeneration of a ‘lead-free’ surface (after each stripping scan),

195

Table 1 Influence of interfering metal ions on the response of Pb(II) at the CMSPE Metal ions origin

Ions

Concentration excess over Pb(II)a

Main group II

Ca(II)

10× 50×

0 0

Ti(IV)

10× 50× 10× 50× 10× 50×

−1 −2 19 30 −1 −2

10× 50× 10× 50× 10× 50× 10× 50×

−4 −43 9 12 −1 −2 −2 0

Mn(II) Ni(II) d-Transition elements

Pd(II) Cu(II) Zn(II) Hg(II)

Contribution (%)

Main group III

Tl(III)

10× 50×

−47 −100

Main group IV

Sn(IV)

10× 50×

−15 −42

Main group V

Bi(III)

10× 50×

−52 −67

a

[Pb(II)] = 20 ppb; t p = 60 s.

which is one of the valuable point to extend it for routine analysis. This efficient surface cleaning effect was further demonstrated from alternate spiking of different Pb(II) concentration as shown in Fig. 6. As can be seen, sequential exposures to Pb(II) solutions (20 and 50 ppb) at different time interval yield constant and reproducible signals with a R.S.D. of 3.75 and 0.63% for 20 and 50 ppb Pb(II), respectively. 3.4. Real sample analysis The analytical utility of the CMSPE was finally demonstrated by applying it to the determination of Pb(II) in groundwater, tap water, and seawater and the results obtained are shown in Fig. 7 and Table 2. These results provide a sufficient evidence for the feasibility of using the CMSPE for the determination of Pb(II) in real water samples. The recovery was good in all three different water samples. The detection mechanism

196

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

Fig. 6. Electrode renewal of 20 and 50 ppb Pb(II) at different time intervals under optimized conditions.

Fig. 7. Real samples analysis for Pb(II) in (A) ground water, (B) tap water, and (C) seawater under optimized conditions. The tp was 1, 3 and 3 min for ground water, tap water, and seawater, respectively. Table 2 Pb(II) assay in three water samples using the CMSPE under the optimized conditionsa

Linear equation R Detected value, original/ppb Spike/ppb After spike/ppb Recovery/% a

Groundwater

Tap water

Seawater

i pa = 12.32 + 0.24[Pb(II)] 0.9998 51.33 ± 1.04 10 62.1 ± 0.12 101.2

i pa = 2.34 + 0.70[Pb(II)] 0.9958 3.34 ± 0.02 10 13.61 ± 0.18 101.9

i pa = 2.25 + 1.66[Pb(II)] 0.9971 1.36 ± 0.08 4 5.43 ± 0.11 101.3

The number of samples assayed were three.

J.-M. Zen et al. / Analytica Chimica Acta 421 (2000) 189–197

of Pb(II) based on the intermetallic property of Cu–Pb was ideal in nature on the CMSPE. The excellent sensitivity, precision, and reproducibility of the CMPSE towards Pb(II) detection offer an opportunity for routine analysis.

Acknowledgements The authors gratefully acknowledge financial support from the National Science Council of the Republic of China under Grant NSC 89-2113-M-005-019. References [1] J. Breen, C. Stroup (Eds.), Lead Poisoning, Lewis Publishers, CRC Press, Boca Raton, FL, 1995. [2] D.A. Roston, E.E. Brooks, W.R. Heineman, Anal. Chem. 51 (1979) 1728. [3] E.Ya. Neiman, L.G. Petrova, V.I. Ignatov, G.M. Dolgopolova, Anal. Chim. Acta 113 (1980) 277. [4] O.L. Kabanova, S.M. Beniaminova, J. Anal. Chem. USSR 26 (1970) 94.

197

[5] S. Dong, Y. Wang, Talanta 35 (1988) 819. [6] H. Gunasingham, R.R. Dalangin, Anal. Chim. Acta 246 (1991) 309. [7] J.-M. Zen, Y.-S. Ting, Anal. Chim. Acta 332 (1996) 59. [8] M. Alvarez-Icaza, U. Bilitewski, Anal. Chem. 65 (1993) 525A. [9] J. Hart, S. Wring, Electroanalysis 6 (1994) 617. [10] C. Galan-Vidal, J. Munoz, C. Domingez, S. Algeret, Trends Anal. Chem. 14 (1995) 225. [11] M. Green, P. Hilditch, Anal. Proc. 28 (1991) 374. [12] J. Wang, Analyst 119 (1994) 763. [13] O. Bagel, B. Limoges, B. Schollhorn, C. Degrand, Anal. Chem. 69 (1997) 4688. [14] A.J. Bard, R. Parsons, J. Jorden (Eds.), Standard Potential in Aqueous Solutions, IUPAC, Marcel Decker, New York, 1985. [15] C.C. Ngo, Ph.D. thesis, National University of Singapore, 1985. [16] M. Lovric, S.K. Lovric, A.M. Bond, J. Electroanal. Chem. 319 (1991) 1. [17] S.K. Lovric, M. Lovric, A.M. Bond, Anal. Chim. Acta 258 (1992) 299. [18] J.-M. Zen, J.-W. Wu, Anal. Chem. 68 (1996) 3966. [19] J.-M. Zen, H.-H. Chung, G. Ilangovan, Electroanalysis 11 (1999) 108. [20] J.-M. Zen, H.-H. Chung, G. Ilangovan, A. Senthil Kumar, Analyst 125 (2000) 1139.