Anal. Chem. 2000, 72, 3476-3479
Determination of Selenium(IV) by a Photooxidized 3,3′-Diaminobenzidine/Perfluorinated Polymer Mercury Film Electrode Hao-Yun Yang and I-Wen Sun*
Department of Chemistry, National Cheng-Kung University, Tainan, 70101, Taiwan
A new, and easily fabricated, chemically modified electrode for the determination of selenium(IV) was examined by cathodic square-wave stripping voltammetry. This new electrode consisted of an anion-exchange perfluorinated polymer (Tosflex) coated thin mercury film electrode containing photooxidized 3,3′-diaminobenzidine (ODAB). The coating solution of Tosflex and ODAB was spin-coated on a glassy carbon electrode followed by electroplating of a thin film of mercury. During the preconcentration, ODAB was reduced electrochemically and selenium was accumulated simultaneously onto the electrode by interacting with the reduced ODAB. After a 5-min preconcentration period, linear response was observed from 0.5 to 50 ppb selenium, and the detection limit was 0.1 ppb. The proposed method does not require a darkened room, which was required in many of the previous methods involving 3,3′-diaminobenzidine. In addition, the resistance to interference from surface-active compounds was improved by incorporating Tosflex in the film. The determination of selenium is of interest because of the biological significance of this element. Electrochemical methods based on stripping voltammetry have been developed for this task. Early study indicated that anodic stripping voltammetric determinations of selenium on bare solid electrodes such as gold could be complicated by peak splitting.1 It was reported recently that cathodic stripping determination of selenium at a silver electrode was not encountered with peak splitting. However, a preconcentration time as long as 30 min and a medium exchange step were required in order to achieve the desired sensitivity.2 Adsorptive stripping methods also have been developed for the determination of selenium.3-6 Many of the reported adsorptive stripping determinations of selenium relied on the reaction between selenium and 3,3′-diaminobenzidine (DAB) to form an adsorptive complex compound which accumulated on the mercury electrode surface. Because the reaction rate between DAB and selenium is slow in the bulk solution, analysis time exceeding several hours was required for a single determination. Recently, the analysis time was greatly shortened by using a gold electrode modified with (1) Andrews, R. W.; Johnson, D. C. Anal. Chem. 1975, 47, 294. (2) Lshiyama, T.; Tanaka, T. Anal. Chem. 1996, 68, 3789. (3) van den Berg, C. M. G.; Khan, S. H. Anal. Chim. Acta 1990, 231, 221. (4) Breyer, P. H.; Gilbert, B. P. Anal. Chim. Acta 1987, 201, 23. (5) Breyer, P. H.; Gilbert, B. P. Anal. Chim. Acta 1987, 201, 33. (6) Stava, V.; Kopanica, M. Anal. Chim. Acta 1988, 208, 231.
3476 Analytical Chemistry, Vol. 72, No. 15, August 1, 2000
electropolymerized DAB.7 This method involved several steps including preconcentration, reduction of the accumulated selenium in a different medium, and a final stripping step. Like the other methods that use DAB, the preparation of this polymer-modified electrode needs to be performed in a darkened room, as the DAB is susceptible to air oxidation in the presence of light. Nevertheless, the prepared polymer-modified electrode was useable for at least one week. This work examines a new approach to applying DAB to the determination of selenium. Instead of DAB, in this present approach, ODAB, the product of the photooxidation of DAB, was coated together with a perfluorinated-anion-exchange polymer named Tosflex8-10as well as with mercury film onto a glassy carbon electrode. This modified electrode was used to preconcentrate and detect selenium. Previous reports5 have shown that ODAB can be electrochemically reduced at a potential dependent on the solution pH. Thus, in this work, a potential was applied to reduce the ODAB in the film, and selenium was accumulated onto the electrode surface by forming complex with the reduced ODAB. The accumulated selenium was then determined by cathodic square-wave stripping voltammetry (SWSV). The new approach greatly reduces the analysis time compared with many of the previous methods and eliminates the necessity of a darkened room. The anion-exchange polymer, Tosflex, not only provided a simple way to immobilize ODAB but also offered a better resistance toward interference from surface-active compounds. EXPERIMENTAL SECTION Apparatus. All electrochemical experiments were performed with a Bioanalytical Systems BAS CV-50W electrochemical analyzer. The three-electrode system consists of a Ag/AgCl reference electrode (BAS), a platinum-wire counter electrode, and the glassy carbon working electrode chemically modified or unmodified. Reagents. Deionized water was used to prepare all solutions. All chemicals were of analytical grade unless otherwise stated. Glassware and polyethylene bottles were soaked in diluted nitric acid and rinsed with deionized water before use. Tosflex membrane, denoted IE-SA 48, was from Tosoh Soda, Japan. 3,3′(7) Cai, Q.; Khoo, S. B. Anal. Chem. 1994, 66, 4543. (8) Dunsch, L.; Kavan, L.; Weber, J. J. Electroanal. Chem. 1990, 280, 313. (9) Ugo, P.; Moretto, L. M.; Mazzocchin, G. A. Anal. Chim. Acta 1995, 305, 74. (10) Lu, T.-H.; Sun, I.-W. Electroanalysis 1998, 10, 1052. 10.1021/ac991145v CCC: $19.00
© 2000 American Chemical Society Published on Web 07/07/2000
Diaminobenzindine (DAB) was obtained from Fluka. Standard solutions (1000 ppm) of selenium(IV), molybdenum(VI), and bismuth(III) were from Mallinckrodt. Standard solutions (1000 ppm) of lead(II), copper(II), and cadmium(II) were from Fisher. Triton X-100 was from Lancaster. Cetyltrimethylammonium bromide (CTAB) was from Arcos Organics. Sodium dodecyl sulfate (SDS) was received from Reidel-de-Haen. Seawater was collected from the beach near Tainan, Taiwan. Determinations of these water samples were performed after the samples were passed through a filter with 0.45-µm pore size. Preparation of the Electrode. The Tosflex coating solutions were prepared by dissolution of finely cut dry membrane into a water-methanol-2-propanol solution according to the literature procedure.8 A typical ODAB/Tosflex solution was prepared by saturating a 0.5 wt % Tosflex solution with DAB in a transparent sample bottle and irradiating with a 100 W tungsten light bulb for 8 h. Results from mass spectrometry experiments indicate that, during this period, dimerization of DAB occurred to form ODAB. The resulting coating solution appeared to be stable against further exposure to the room light and could be used for at least a week. In the preparation of a ODAB/Tosflex mercury electrode (ODTMFE), 4 µL of the ODAB/Tosflex coating solution was spin-coated onto a GCE at a spin rate of 3000 rpm. A uniform thin film was formed after 3 min of spinning. Mercury was then electroplated onto the ODAB/Tosflex coated GCE from 5 mL of 30 ppm mercury(II) solution containing 0.1 M sodium perchlorate at a potential of -0.8 V vs Ag/AgCl for 4 min with stirring. Procedure. Unless otherwise stated, in this study, 1.0 M nitric acid was used to adjust the pH of the medium and sodium chloride was used as the supporting electrolyte in the electrochemical experiments. In each experiment, the analyte solution was deaerated with argon for 3 min. Following this, a freshly prepared ODTMFE was dipped into the analyte solution and selenium(IV) was preconcentrated at a proper potential with stirring of the solution. After the preconcentration period, the stirring was stopped, and quantitative determination was performed in the SWSV mode. The potential was scanned in a negative direction from -0.3 to -0.9 V vs Ag/AgCl. After each determination, the selenium species produced was anionic Se(-II) and, therefore, would be retained in the polymer film due to the anion-exchange nature of Tosflex. As a result, a new modified electrode was used for each determination. RESULTS AND DISCUSSION Parts a and b of Figure 1 show the linear-sweep voltammograms (LSV) for a pH 2.1 blank solution (containing 0.1 M KCl as electrolyte) recorded at a Tosflex mercury-film-coated GCE (TMFE) and at a ODTMFE, respectively. As can be seen, no reduction waves were observed on the TMFE, whereas a reduction wave at ∼-0.35 V was observed at the ODTMFE. Similar to that reported by Breyer and Gilbert,5 the reduction wave observed at ODTMFE was due to the reduction of ODAB in the film. LSV at various electrodes in solution containing 10 ppm Se(IV) are shown in parts c, d, and e of Figure 1. At the bare GCE (Figure 1c), no reduction waves due to Se(IV) could be observed at this Se(IV) concentration level. When a TMFE was used, two reduction waves at ∼-0.22 and -0.61 V were obvious (Figure 1d). These two waves are typical for Se(IV) at a mercury electrode in acidic aqueous solutions11 and could be attributed to the reduction of
Figure 1. Linear-sweep voltammograms (LSV) of a blank solution at: (a) TMFE, and (b) ODTMFE, and LSV of 10 ppm Se(IV) at: (c) bare GCE, (d) TMFE, and (e) ODTMFE. Sweep rate: 50 mV s-1, pH ) 2.0, [KCl] ) 0.1 M.
Se(IV) to form HgSe with mercury and further reduction of HgSe to give Se(-II). In comparison with that observed at the bare GCE, the better sensitivity observed at the TMFE resulted from the HgSe formation capability of mercury and the anion-exchange feature of the Tosflex polymer. Compared with Figure 1d, Figure 1e shows that, for the same experiment performed with the ODTMFE, the LSV demonstrated an even higher response for Se(IV), revealing that the presence of ODAB in the film further enhanced the accumulation of Se(IV). If the ODTMFE was preconditioned at -0.35 V in a blank solution containing no Se(IV), to reduce the ODAB in the film, the resulted electrode was able to accumulate Se(IV) at an opencircuit potential when the electrode was transferred into a Se(IV) solution. This was investigated with linear-sweep voltammetry for 10 ppb Se(IV) solutions. Linear-sweep voltammograms obtained for a Se(IV) solution after accumulation at open-circuit potential on an unconditioned ODTMFE showed only the cathodic wave due to reduction of ODAB, and no appreciable stripping peak of selenium could be seen. This indicates that the accumulation efficiency of the unconditioned ODTMFE for Se(IV) was low. On the other hand, when the same experiment was performed at an ODTMFE, which had been preconditioned at -0.35 V for 60 s in a pH 2.0 blank solution containing 0.1 M KCl before being immersed in the Se(IV) sample solution, a distinguishable selenium stripping peak was obtained, indicating that the accumulation of selenium on this modified electrode relied on the reduced ODAB. However, if the preconditioning, medium exchange accumulation, Se(IV) reduction, and stripping steps were performed separately, the analysis time would be long. To shorten the analysis time, in the following study, the steps of reduction of ODAB and accumulation and reduction of Se(IV) were conducted simultaneously in the Se(IV) sample solution at a proper preconcentration potential prior to the final stripping step. The dependence of the selenium stripping peak current on the preconcentration potential was examined over the -0.2 to -0.45 V range with the pH and accumulation time fixed at 2.0 and 4 min, respectively. Figure 2 shows that the peak current increased as the preconcentration potential approached the (11) Adeloju, S. B.; Bond, A. M.; Briggs, M. H.; Hughes, H. C. Anal. Chem. 1983, 55, 2076.
Analytical Chemistry, Vol. 72, No. 15, August 1, 2000
3477
Figure 2. Dependence of SWSV peak height 10 ppb Se(IV) on the preconcentration potential, Ep. Accumulation time, tp ) 4 min, pH ) 2.0, supporting electrolyte, 0.1 M KCl. SWSV parameters: modulation amplitude, 50 mV; modulation frequency, 50 Hz.
Figure 3. Effect of pH on the SWSV peak height of 10 ppb Se(IV). Ep ) -0.35 V, tp ) 4 min, [KCl] ) 0.1 M. SWSV parameters were the same as in Figure 2.
reduction potential of ODAB, reached a maximum at -0.4 V, and then decreased sharply as the preconcentration potential moved closer to the stripping peak potential. The effect of different pHs on the selenium stripping peak current was investigated by accumulating Se(IV) at -0.35 V for 4 min. Figure 3 shows that the stripping peak current increased with increasing pH until about pH 2.0. Further increase in pH lowered the stripping peak current. As described by Breyer and Gilbert,4 the reaction rate for Se/DAB complex formation depends on the distribution of undissociated selenious acid as well as on the number of protonated DAB molecules. The relative amounts of these two species in the solution are determined by their dissociation constants and the solution pH. A maximum Se/DAB complex formation rate at pH 1.5 was observed by Breyer and Gilbert. Higher or lower pH resulted in either decreased protonation of DAB or increased dissociation of selenious acid, which in turn reduced the Se/DAB complex formation rate and, thus, the stripping peak current. The pH which gave highest peak current in our study is slightly larger than that reported by Breyer and Gilbert. This difference could be due to the difference in the reaction medium (bulk aqueous to Tosflex polymer film). A change in the dependence of the Se/DAB complex formation rate on pH was also observed in the study which involved using a electropolymerized DAB film-modified gold electrode to determine Se(IV).7 3478 Analytical Chemistry, Vol. 72, No. 15, August 1, 2000
Figure 4. Dependence of the SWSV peak height of (a) 3 ppb and (b) 10 ppb Se(IV) on the preconcentration time. Ep ) -0.35 V. pH ) 2.0. [KCl] ) 0.1 M. SWSV parameters were the same as in Figure 2.
The detection of selenium with ODTMFE also relies on the effective formation of anionic [SeCl6]2-, which can penetrate to the electrode surface by way of the anion-exchange feature of Tosflex. The effect of changing the chloride-ion concentration on the selenium stripping peak current was studied with 10 ppb Se(IV) solutions. The results showed that selenium response increased with increasing chloride concentration from 0 to 0.1 M and started to level off at higher chloride concentrations. Therefore, a chloride concentration of 0.1 M was employed in the subsequent experiments. The dependence of the selenium stripping peak current on the preconcentration time, tp, was studied from 1 to 8 min, for two different initial Se(IV) concentrations of 3 and 10 ppb. Figure 4 shows plots of the stripping peak current as a function of the preconcentration time for each of the selenium concentrations studied. It is notable that for both Se(IV) concentrations, the stripping peak current increased slowly with tp until about 200 s, after which the stripping current increased more rapidly with increasing tp, and eventually it leveled off. This observation reflected that the ODAB in the film had to be reduced first before it interacted with selenium. Also indicated in Figure 4 is that a higher Se(IV) concentration gave a larger stripping peak current and a shorter tp for the stripping peak current to level off. Apparently, the accumulation rate of Se(IV) was also affected by the diffusion of Se(IV) ion into the film, and a higher concentration gradient of the Se(IV) ion across the film means a faster accumulation rate to reach the equilibrium. The SWSV peak height was found to vary linearly with Se(IV) concentration for a 5-min preconcentration at -0.4 V. Under the chosen conditions, for Se(IV) concentration from 0.5 to 50 ppb, the calibration plot gave a slope of 1.5 µA ppb-1 with a correlation coefficient of 0.997. The detection limit was estimated to be 0.1 ppb. Trace metals can interfere if they can form a complex with DAB under the condition employed and if these ions have redox processes overlapping with the stripping peak for Se(IV). However, such interference can be largely suppressed by adding ethylenediaminetetraacetate (EDTA) in the sample solution. For the determination of 10 ppb Se(IV) in a solution saturated with EDTA and a 5-min preconcentration time under the chosen
conditions, the presence of 5 ppm (500-fold wt. excess concentration) of Bi(III), Pb(II), Cd(II), and Mo(VI) gave no observable interferences. However, the presence of Cu(II) at a concentration level higher than 1 ppm interfered with the determination. Interferences from surface-active compounds were observed in many of the previous methods reported for the determination of selenium. Such interferences were reduced by including Tosflex into the ODTMFE. In this study, several surfactants, including Triton X-100, SDS, CTAB, were used to exemplify the effects of typical surfactants on the measurement of 5 ppb Se(IV) using both the bare-mercury-film electrode (MFE) and the ODTMFE. As shown in Figure 5, the ODTMFE, in comparison with the MFE, displayed higher resistance to the surfactant interference. The analytical utility of the proposed procedure using the ODTMFE was assessed by applying it to the determination of Se(IV) in seawater samples. No selenium was detected in the original water samples, so they were spiked with 3 ppb of Se(IV). The detected value for the spiked seawater samples was 2.92 ppb, indicating that the recovery (97%) of spiked Se(IV) is good for the seawater sample. Note that the amount of Se(IV) in natural seawater is typically very low,3 and this is indeed the case in this study. In conclusion, the method described provides a sensitive and yet a simple approach to the determination of selenium. Fewer steps are involved and a relatively short preconcentration time can be employed for measuring selenium, thus greatly reducing the total analysis time. The use of ODAB eliminates the need of a darkened room for conducting the analysis. Tosflex film
Figure 5. Effect of the surfactants: (O) Triton X-100, SDS ([), and CTAB (4) on the stripping response for 10 ppb Se(IV) on the ODTMFE (solid line) and the bare MFE (dashed line). [KCl] ) 0.1 M, pH 2.0, tp ) 4 min, Ep ) -0.35 V.
offers also a good resistance to interferences from surface-active compounds. ACKNOWLEDGMENT The authors wish to acknowledge gratefully the financial support from the National Science Council of the Republic of China, Taiwan under Grant NSC 88-2113M006002. Received for review October 4, 1999. Accepted May 15, 2000. AC991145V
Analytical Chemistry, Vol. 72, No. 15, August 1, 2000
3479
