Voltammetric Assay of Mercury Ion in Fish Kidneys

Voltammetric analysis of mercury ions was developed using paste electrodes (PEs) with DNA and carbon nanotube mixed electrodes. The optimized analytical results of the cyclic voltammetry (CV) of the 1∼14 ng L−1 Hg(II) concentration and the square wave (SW) stripping voltammetry of the 1∼12 ng L−1 Hg(II) working range within an accumulation time of 400 seconds were obtained in 0.1 M NH4H2PO4 electrolyte solutions of pH 4.0. For the relative standard deviations of the 1 ng L−1 Hg(II), which were observed at 0.078% (n = 15) at the optimum conditions, the low detection limit (S/N) was pegged at 0.2 ng L−1 (7.37 × 10−13M) for Hg(II). The results can be applied to assays in biological fish kidneys and wastewater samples.


INTRODUCTION
In the aquatic ecosystem, mercury ions come from art paint, coal, food, pharmaceuticals, and others (Sang et al., 2004;Eun et al., 2003;Seung et al., 2003), which create trace metals that cause food or environmental toxicology (Joanna and Michael, 2000;Eisie and Gail, 2000;Nathalie et al., 2005). Their assay is particularly important in biological analyses (Janos and Erzsebet, 2005) such as for paresthesia, ataxia, dysarthyria, hearing defects, visual disturbances, and other medical fields (Carrington et al., 2004;Ertas and Tezel, 2004). Thus, various common analytical methods of mercury ion assay have been developed, such as gas chromatography -mass spectrometry (Giuseppe et al., 2004;Petru and Freddy, 2004), HPLC with atomic fluorescence spectrometry (Li et al., 2003(Li et al., , 2004, zeeman atomic absorption spectrometry (Sholupov et al., 2004), atomic absorption spectrometry (Cizdziel and Shawn, 2004;Shigehiro et al., 2004;Claudia et al., 2005), and others (Van Staden and Taljaard, 2004). These methods achieve very low detection ranges and are composed with various processing systems that involve separation, sampie injection, temperature control, and detection, which depend on spectroscopic or voltammetric detection systems. Of late, more compact, sim-pie, and sensitive analytical techniques are being required. Stripping voltammetry is sensitive in trace analysis, which depends on working electrode systems and is commonly used with drop mercury and mercury film electrodes (Clinio and Giancarlo, 2001;Tsai et al., 2001;Sonia, 1998;Percio et al., 2003;Joseph et al., 2001), glassy carbon electrodes (Joseph et al., 2000;Suw et al., 2002Suw et al., , 2004, paste electrodes (Tesfaye et al., 1999;Jahan et al., 2001;Korbut et al., 2001), and other modified electrodes. Despite this, few studies have been conducted on mercury analysis and, and these studies have achieved low detection limits. For example, the gold disk electrode arrived at 5 ng L-1 after 10 min of deposition time (Ricardo et al., 2000); the electrochemical quartz crystal microbalance methods arrived at the 1 ppb level (Nieis et al., 1996); the borondoped diamond film electrode method arrived at the 0.005-50 ppb working range (Manivannan et al., 2005); and the nafion-coated glassy carbon electrode method reached a 10 nM detection limit (Zuliang et al., 1999). Some of these methods are used, however, with long accumulation times and have arrived at poor detection limits. Thus, in this study, new and more sensitive methods are investigated using paste electrodes (PEs) with DNA and carbon nanotube mixed electrodes. A former researcher has researched on the DNA-coated carbon paste electrode for drug detection (Radi, 1999), and the DNA-modified electrode responded distinctly with other molecules and ions (Yuan et al., 1999;. On the other hand, carbon nanotube properties s. Y. Ly are useful in eleetroeatalysis and sensor applications [35] (Bailure et al., 2003) and high eleetrieal eonduetivity , and can perform eleetron transfer with biomoleeules (Gang et al., 2002). In this study, the DNA and nanotube properties were combined in mereury ion analysis, whieh aehieved optimum eonditions at lower deteetion limits, thus registering better performance than other eommon voltammetrie methods.

EXPERIMENTAL PROCEDURE
All the voltammetrie measurements in this study were made using a CHI660A instrument electroehemical workstation (from CH Instruments, Ine., Cordova, TN, USA). A three-eleetrode system was used to monitor the voltammogram. The PE was used as the working eleetrode and was saturated with Ag/AgCI/KCI as the reference, and a platinum wire was used as the auxiliary eleetrode. The eleetrolyte solutions were used with doubledistilled water (18 M ohm em-1 ), and the double-stranded calf thymus DNA (dsDNA) and the other reagents were obtained from Aldrieh. The multi-walled earbon nanotubes (15-40 nm in diameter) for the CVD method were obtained from Nanoteeh in Korea. Following this, the DNA and graphite nanotube mixing weight ratios of 5 : 1, 4 : 1, 3 : 1, 2 : 1, 1 : 1, 1 : 2, 1 : 3, 1 : 4, and 1 : 5 9 were examined. At 1 : 2, the maximum inereased peak eurrent of 0.2 V Hg(lI) appeared in eyclie voltammetry. Thus, the 1 : 2 ratio was used for all the experimental mixing weights. The paste eleetrode was made by mixing 70% powder with 30% mineral oil. This mixture was homogenized in a mortar for 30 minutes. The mixed paste was inserted into a plastie syringe with a diameter of 3.0 mm, and a eopper wire was eonneeted to the eleetrical system. The three-electrode system was immersed in a 15 m; cell that eontained eleetrolyte solutions of 0.1 M ammonium phosphate solution, whereas the other parameters were maintained at optimal eonditions. All the experiments were performed at a room temperature of 24 ± 0.5°C, without removing oxygen, and it was found that all the experiments eould be performed in an open cireuit. Various acid and base eleetrolyte solutions (all 0.1 M solutions) were initially examined in seareh of a possible supporting eleetrolyte. Ammonium phosphate solution was found to have been the most suitable.
V switehing potentials and a scan rate of 0.5 V/so Only oxidation peak signals were obtained at these eonditions; reduetion signals did not appear. Thus, various working ranges were examined, and very low eoncentration ranges were obtained at ppt levels of 10,20,30,40,50,60,70,80,90, and 100 ng L-1 Hg(lI) (not shown here). At these ranges, the linear equation y = 0.3168x + 4.893 and an R 2 = 9879 precision were sensitively obtained, the peaks of whieh sharply responded. More sensitive working ranges were examined using a faraday cage via environmental noise exclusion. Fig. 1 shows the very low ranges of 1-14 ng L-1 Hg(lI) with PE in an eleetrolyte solution, the very smooth blank solutions, and the non-appearance of noise peak signals; and that from 1 to 14 ng L-\ sharply and linearly inereasing signals were obtained. At these eonditions, the slope ratio of M/>,.y = 5.50, the precision of R 2 = 0.9947, and the maximum peak height of 76.86 x 10-7 A appeared, and the peak width narrowly responded. These results are usable in any environment and biological applieation. More sensitive working ranges were searehed for using square wave stripping voltammetry.
Stripping voltammetry optimization. Fig. 2(a) shows the voltammetrie peak eurrent in the 100 mg L-1 Hg(lI) eoncentration as a funetion of varying square wave frequeneies in the 50-500 Hz range, dswithin an aeeumulation time of 20 seeonds. All the optimized examinations were used with short aeeumulation times for fast results at these eonditions. Mereury ions were dispersed at the two peak potentials of 0.2 and 0.4 V, and the 0.2 V peak potential very quickly increased from 50 to 150 Hz, and then to 312 Hz after the peak current did not increase. From 357 to 500 Hz, the 0.2 V peak current very quickly increased. Thus, all the other frequency conditions were used at 500 Hz; and at the maximum conditions, the 0.2 V peak current appeared at a 151.2 x 10-6 A high, whereas the 0.4 V peaks were smaller and later disappeared. Thus, all the other experimental conditions used these results. Fig. 2(b) shows the various incremental potentials of 1-10 mV when other parameters were used. In this voltammogram, two separate peaks appeared and the peak width broadly increased. Finally, the maximum response was obtained at the 7 mV incremental potential, at which a 181.86 x 10-6 A peak high appeared more sensitively than did the frequency results. Thus, this potential was used for all the other experimental conditions. In Fig. 2(c), the accumulated potentials were examined for the native potential range of -1--2 V using the conditions in Fig. 2(b). At -1.3 V, maximum peak currents of 446.7 x 10-6 A were obtained, which responded better than the frequency or incremental potential peak high, and -1.3 V was used for all the other experiments, the peak sharps of which were not separated. Also, cathodic stripping was performed at the optimized conditions, but no signals were obtained. Fig. 3(a) shows the results of various square wave amplitudes that were examined for the 0.01-0.35 V range of the peak potentials that appeared at 0.2 V, for other small peak potentials that responded positively at 0.5 V, and for wh oie peak currents that increased to 0.35 V, then did not increase. The peak width also broadly appeared. Thus, all the other conditions were used at the 0.35 amplitude potentials. At this state, an 80.23 x 10-6 A peak high was obtained, which was more poorly influenced than were the other stripping parameters. Finally, the accumulation times were examined. Fig. 3(b) shows the results, which responded very sensitively and better than the other experimental parameters. Thus, lower concentrations of 0.01 fl9 L-1 Hg(lI) were examined from 20 to 400 s accumulation times at the range of 0-20 sec. The oxidation peaks slowly appeared and the peak width broadly increased, whereas at the 50-400 s range, the peak high quickly increased and the peak half-width sharply decreased. At these conditions, the maximum current of 78.57 x 10-6 A appeared, no separate peak appeared, and only a simple peak was obtained. Increasing accumulation times were not used for the experimental time consumption at the fixed conditions, however, and the electrode-usable times were examined for several weeks, during which time the electrode surface was cleaned with the weigh-ing paper and a much longer reproducibility was obtained. Finally, the analytical application was examined in the biological and toxicological waste sampies.

Statistics and application.
At the optimized conditions, various working concentrations ranges were examined. Fig. 4(a) shows the results of a 400 s deposition time at these conditions. One to 12 ng L-1 Hg(lI) appeared linearly, and the peak width was very sharply obtained, whereas at the more increased concentrations of 13-15 ng L-1 , the mercury ion responded nonlinearly. All the raw and calibrated equations show this.  2,3,4,5,6,7,8,9,10,11, and 12 ng L-1 on PE with an electrolyte solution of 0.1 M NH4H2P04 (pH 5.0). The deposition potential was -1.3 V for a 400 s accumulation time, an SW frequency of 500 Hz, a 0.35 V amplitude, and a 7 mV incremental potential. (b) Application of the blank solution, the 0.03 m L-1 fish kidney, and the 0.1, 0.2, and 0.3 mg L-1 standard Hg(lI) spikes. The other experimental parameters in Fig. 3 were held constant. results yielded 0.2 ng L· 1 for Hg(lI) (7.37 x 10-13 M). The experimental results were more sensitive than those derived fram other eommon voltammetrie methods, and fram sueh results, various possible interference metal ions were examined by adding several other ions using 0.1 mg L-1 Hg(II), and the existenee of 1 mg L-1 of Pb(II), Ba(II), Ca(II), Bi(II), Co(III), Fe(II), Cr(lll), and Pt(l) resulted in 29.3%, 330.0%, 106.2%, 66.6%, -78.8%, 97.2%, -23.6%, and -61.0%. The presenee of other ions was also effeetively removed using standard addition methods. At the optimum eonditions, the analytieal applications were performed with river fish kidney solutions prepared using fish with 80 9 body weights and extraeted kidneys with 0.53 9 weights, whieh were diluted in a 0.1 M HCI solution and examined during the application. In Fig. 4(b), however, the first peak shows the eleetralyte solution results, after whieh the 0.03 ml sampie solutions were spiked and a small 0.2 V mereury ion potential appeared, whieh manifested no noise signal.
Thus, another standard for 0.1, 0.3, and 0.4 mg L-1 Hg (lI) was spiked at these results, caleulated for the standard addition methods, and yielded y = 0.0345x + 4.783 and R 2 = 0.9818 and 0.139 mg/go Other known waste solutions were examined and a mean 95% eonfidence limit was obtained via tri pie analysis. Quantitation of known laboratory waste sampies was prepared with 1.0 fl9 L-1 Hg(lI) using an industrial solution. In a 10 ml 0.1 M NH4H2 P04 eleetralyte solution, a spiked 1.5 ml waste sampie showed a distinet peak eurrent of Hg(II), after whieh 6, 12, and 18 fl9 L-1 Hg(lI) standards were added to the eleetralyte eell systems. Three repeated determinations were made on eaeh cell system. The plot of the peak eurrent against the mereury eoncentration was linear (R 2 = 0.998). The eoncentration of mereury in the sampie was found to have been 0.96 ± 0.06 flg L-1 (n = 3). This agrees weil with the value fram the known eoncentration.

CONCLUSION
A newly prepared DNA and carbon nanotube paste eleetrade was developed to deteet mereury ions at nanogram levels with stripping and eyclie voltammetry. The method offers attraetive praperties eompared to other voltammetrie methods, sueh as a lower deteetion limit, simple electrode preparation, and long stability. Optimized analytical eonditions were researehed on and applied to the deteetion of mereury ion eoncentrations in low eoneentration ranges, and various interferenee ions were searehed for. The analytical applications of the fish tissue were also examined. This study can also be applied in other fields that require mereury ion analysis of food or enviranmental toxieology.