Electrochemical Oxidation of Oxalic Acid at Highly Boron-Doped

Anal. Chem. 2006, 78, 3467-3471

Technical Notes

Electrochemical Oxidation of Oxalic Acid at Highly Boron-Doped Diamond Electrodes Tribidasari A. Ivandini,†,‡ Tata N. Rao,§ Akira Fujishima,⊥ and Yasuaki Einaga*,†

Department of Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan, Department of Chemistry, Faculty of Mathematics and Science, University of Indonesia, Kampus Baru UI Depok, Jakarta 424-0041, Indonesia, International Advanced Research Centre for Powder Metallurgy and New Materials, India, and Kanagawa Academy of Science and Technology, KSP 3-2-1 Sakado, Kawasaki 213-0012, Japan

Electrochemical oxidation of oxalic acid has been investigated at bare, highly boron-doped diamond electrodes. Cyclic voltammetry and flow injection analysis with amperometric detection were used to study the electrochemical reaction. Hydrogen-terminated diamonds exhibited well-defined peaks of oxalic acid oxidation in a wide pH range. A good linear response was observed for a concentration range from 50 nM to 10 µM, with an estimated detection limit of ∼0.5 nM (S/N ) 3). In contrast, oxygenterminated diamonds showed no response for oxalic acid oxidation inside the potential window, indicating that surface termination contributed highly to the control of the oxidation reaction. An investigation with glassy carbon electrodes was conducted to confirm the surface termination effect on oxalic acid oxidation. Although a hydrogenterminated glassy carbon electrode showed an enhancement of signal-to-background ratio in comparison with untreated glassy carbon, less stability of the current responses was observed than that at hydrogen-terminated diamond. Oxalic acid naturally occurs in many plants (spinach, ginger, chocolate, etc) and combines with Ca, Fe, Na, Mg, or K to form less soluble salts known as oxalates. High levels of these salts in the diet can lead to irritation of the digestive system, particularly of the stomach and kidneys. It is also known to contribute to the formation of kidney stones. The urinary level of oxalic acid has long been recognized as an important indicator for the diagnosis of renal stone formation. Many analytical techniques, such as spectrophotometric,1 gas chromatographic,2 and enzymatic methods3 have been proposed to quantify oxalic acid in several real * To whom correspondence should be addressed. E-mail: einaga@ chem.keio.ac.jp. † Keio University. ‡ University of Indonesia. § International Advanced Research Centre for Powder Metallurgy and New Materials. ⊥ Kanagawa Academy of Science and Technology. (1) Laker, M. F.; Hofman, A. F.; Meeuse, B. J. D. Clin. Chem. 1980, 26, 827. (2) Jellum, E. J. J. Chromatogr., B 1997, 143, 427. (3) Reddy, S. M.; Higson, S. P.; Vadgama, P. M. Anal. Chim. Acta 1997, 343, 59. 10.1021/ac052029x CCC: $33.50 Published on Web 04/20/2006

© 2006 American Chemical Society

matrixes. In addition, liquid chromatography and capillary electrophoresis separation combined with UV-absorption4 and electrochemical detection mode5,6 are also successfully employed for the determination of oxalate species. However, each method has often suffered from diverse disadvantages concerning costineffectiveness, insufficient selectivity, essential derivatization for sensitive detection, and time-consuming sample detection. The electrochemistry of oxalic acid at some electrode materials has been studied.7-13 Most of the investigations have been carried out using platinum electrodes.7,8 Some interests were addressed concerning other metals, such as gold and palladium.9,10 However, to the best of our knowledge, there is no report of direct oxalic acid oxidation at bare carbon electrodes, although some metal modification at glassy carbon (GC) electrodes has been reported.11-13 The superiority of highly boron-doped diamond (BDD) electrodes in comparison with other conventional carbon electrodes is well established due to the wide potential window in aqueous solution, low background currents, long-term stability, and low sensitivity to dissolved oxygen.14,15 Direct electrochemical detection of some important chemicals and biochemicals using BDD electrodes in which BDD electrode shows some advantages over GC electrode has been reported.15-17 (4) Gaffney, M.; Morrice, N.; Cookee, M. Anal. Proc. 1984, 21, 434. (5) Santos, L. M.; Baldwin, R. P. Anal. Chem. 1986, 58, 848. (6) Fogg, A. G.; Alonso, R. M.; Fernandez-Archinieza, M. A. Analyst 1986, 111, 249. (7) Berna, A.; Rodes, A.; Feliu, J. M. J. Electroanal. Chem. 2004, 563, 49. (8) Pron’kin, S. N.; Petrii, O. A.; Tsirlina, G. A.; Schiffrin, D. J. J. Electroanal. Chem. 2000, 480, 112. (9) Martinez-Huitle, C. A.; Ferro, S.; De Bastiti, A. Electrochim. Acta 2004, 49, 4027. (10) Santos, L. M.; Baldwin, R. P. Anal. Chem. 1986, 58, 848. (11) Casella, I. G.; Electrochim. Acta 1999, 44, 3353. (12) Shaidarova, L. G.; Gedmina, A. V.; Cheinokova, I. A.; Budnikov, G. K. J. Anal. Chem. 2003, 58, 886. (13) Casella, I. G.; Zambonin, C. G.; Prete, F. J. Chromatogr., A 1999, 833, 75. (14) 14.Diamond Electrochemistry; Fujishima, A., Einaga, Y., Rao, T. N., Tryk, D. A., Eds.; Elsevier-BKC: Tokyo, 2004. (15) Yano, T.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. J. Electrochem. Soc. 1998, 145, 1870. (16) Manivannan, A.; Seehra, M. S.; Fujishima, A. Fuel Process. Technol. 2004, 85, 513. (17) Nekrassova, O.; Lawrence, N. S.; Compton, R. G. Analyst 2004, 129, 804.

Analytical Chemistry, Vol. 78, No. 10, May 15, 2006 3467

As-deposited BDD (ad-BDD) electrodes are initially hydrogenterminated as they are deposited in a hydrogen-plasma chemical vapor deposition (CVD) chamber.15 A hydrogen-terminated BDD electrodes can be altered to oxygen-terminated by a variety of methods, including anodic treatment,18 oxygen plasma treatment,19 boiling in strong acid,20 or long-term exposure to air.21 An oxygenterminated BDD electrode can be converted back to hydrogen termination by hydrogen-flame22 or hydrogen-plasma treatment.23 The advantages of oxygen-terminated BDD electrodes in comparison with hydrogen-terminated BDD electrodes have been reported and demonstrated in a much wider potentials window24,25 and higher surface stability from fouling.26 However, an oxygenterminated BDD electrode has a disadvantage for direct oxidation of negatively charge molecules, that is, DNA27 or AQDS,28 due to repulsion effects of the surface. This present work focused on the investigation of the electrochemical behavior of oxalic acid at bare highly boron-doped diamond electrodes by using cyclic voltammetry (CV) and flow injection analyisis (FIA). We demonstrate the advantage of hydrogen-terminated BDD electrodes over oxygen-terminated for oxalic acid oxidation. Some investigations at hydrogen-terminated GC electrodes were conducted to confirm the termination effect. Signal and background current, as well as the stability of both hydrogen-terminated BDD and GC electrodes, were also compared. This result clearly demonstrates that although a hydrogenterminated GC shows some advantages over an oxygen-terminated one, BDD electrodes are still superior in terms of stability and practicality. EXPERIMENTAL SECTION Preparation of the Electrodes. BDD electrodes were deposited on Si (100) wafers in a microwave plasma-assisted chemical vapor deposition system (ASTeX Corp.) at 5 kW with high purity hydrogen as the carrier gas. The details of the film preparation have been described previously.14,15 B2O3, the boron source, was dissolved in an acetone-methanol mixture (ratio of 9:1 v/v) at a B/C molar ratio of 1:100. A film thickness of ∼40 µm was achieved after 10 h of deposition. The film quality was confirmed by Raman spectroscopy. Before use as a working electrode, the BDD film was pretreated by ultrasonication in 2-propanol for ∼10 min, followed by rinsing with high-purity water to remove the organic impurities that may have remained or formed during the deposition of diamond in the CVD chamber. (18) Notsu, H.; Yagi, I.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. Electrochem. Solid-State Lett. 1999, 2, 522. (19) Yagi, I.; Notsu, H.; Kondo, T.; Tryk, D. A.; Fujishima, A. J. Electroanal. Chem. 1999, 472, 173. (20) Hayashi, K.; Yamanaka, S.; Watanabe, H.; Sekiguchi, T.; Okushi, H.; Kajimura, K. J. Appl. Phys. 1997, 81, 744. (21) Martin, H. B.; Argoitia, A.; Landau, U.; Anderson, A. B.; Angus, J. C. J. Electrochem. Soc. 1996, 143, L133. (22) Lee, J.; Park, S.-M. Anal. Chim. Acta 2005, 545, 27. (23) Granger, M. C.; Swain, G. M. J. Electrochem. Soc. 1999, 146, 4551. (24) Popa, E.; Notsu, H.; Miwa, T.; Tryk, D. A.; Fujishima, A. Electrochem. Solid State Lett. 1999, 2, 49-51. (25) Popa, E.; Kubota, Y.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2002, 72, 1724. (26) Terashima, C.; Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima A. Anal. Chem. 2002, 74, 895. (27) Ivandini, T. A.; Sarada, B. V.; Rao, T. N.; Fujishima, A. Analyst 2003, 128, 924. (28) Rao, T. N.; Ivandini, T. A.; Terashima, C.; Sarada, B. V.; Fujishima, A. New Diamond Front. Carbon Technol. 2003, 13, 79.

3468

Analytical Chemistry, Vol. 78, No. 10, May 15, 2006

As-deposited BDD films were altered to oxygen-terminated by applying a potential of +3.00 V vs SCE for 20 min in BrittonRobinson buffer (pH 2.1). The glassy carbon electrode (Tokai Carbon Co., Ltd.) was pretreated by polishing with alumina powder (Wako), then followed by ultrasonication in 2-propanol before the experiment. Hydrogen termination was introduced to GC by hydrogen-plasma treatment in the CVD chamber at 3 kW for 20 min. The oxygen-to-carbon (O/C) ratio of the film was confirmed by XPS. Electrochemical Measurement. Electrochemical measurements were carried out with a single-compartment cell. A Ag/ AgCl (saturated KCl) electrode was used as the reference electrode, and a Pt wire was used as the counter electrode. The planar working electrode was mounted on the bottom of the glass cell by use of a silicon O-ring. The geometric area of the working electrode was estimated to be 0.07 cm2. The supporting electrolyte was a 0.1 M phosphate buffer solution (PBS), pH 2.1. All measurements were made at room temperature (23 ( 2 °C). The electrical contact for the BDD electrode was made through the backside of the scratched silicon substrate by contacting the brass current-collecting back plate. Cyclic voltammograms were recorded using a potensiostat (Hokuto Denko, Hz-1000). Flow Injection Analysis Measurement. The flow injection analysis (FIA) system used in the present study consisted of a micro-LC pump (GL Sciences, Inc., PU611), an injection instrument for constant 20-µL injections, a thin-layer flow cell (GL Sciences, Inc.), an amperometric detector (Bioanalytical Systems LC-4C), and a data acquisition system (EZ Chrom Elite, Scientific Software, Inc.). The flow rate set for the pump was 1 mL min-1 and was confirmed before every experiment by measuring the volume of the buffer collected at the outlet for 10 min. The walljet-type flow cell consisted of the Ag/AgCl/1 M LiCl reference electrode and a stainless steel tube as the counter electrode, which also served as the tube for the solution outlet. A 0.5-mm-thick silicon rubber gasket was used as a spacer in the cell. The geometric area was estimated to be 0.15 cm2, and the cell volume was estimated to be 6 µL by assuming a 25% compression of the gasket. The mobile phase was 0.1 M PBS, pH 2.1. Hydrodynamic voltammograms were obtained prior to amperometric determination, and the detection potential was chosen in the half peak oxidation potential of the voltammograms. The chromatograms were obtained at an applied potential 1.1 V vs Ag/AgCl at the BDD electrode. The flow cell was thermostated at 26 °C. RESULTS AND DISCUSSION Figure 1 shows a series of CVs for the oxalic acid oxidation at various concentrations (10-100 µM, n ) 4) in 0.1 M PBS, pH 2.1, at an ad-BDD electrode. The inset shows good linearity (r2 ) 0.9931) of the current density observed with a linear equation of y ) 0.0072x + 0.0299. The CVs show a well-defined oxidation peak corresponding to the oxidation of oxalic acid at ∼1.32 V (vs Ag/ AgCl). It can be seen that oxalic acid is easily oxidized at adBDD electrodes. For each CV, the upper curve (bold line) represents the oxidation (forward) scan, whereas the lower curve (thin line) represents the reduction (reverse) scan. No peak was observed during the reverse scan within the working potential from 0 to 1.45 V (vs Ag/AgCl), which indicates that the oxalic acid oxidation is an irreversible reaction.

Figure 1. Cyclic voltammograms of oxalic acid in various concentrations in 0.1 M phosphate buffer solution, pH 2.1, at as-deposited diamond electrode. Scan rate was 100 mV/s. The bold and thin line show forward and reverse scans, respectively. Insets show (a) linear dependence of the current density on oxalic acid concentration and (b) repetitive cyclic voltammograms of 100 µM oxalic acid solution with a quiet time of 1 min between cycling.

During the past decade, much has been learned about the factors that affect carbon electrode reactivity. It is clear, depending on the redox analyte, that one or more of the following factors, (1) surface cleanliness, (2) surface microstructure, (3) hydrophobicity/hydrophilicity, (4) electronic structure (i.e., density states), and (5) surface carbon oxides, are influential.29 It is well-known that BDD electrodes are to be resistant to fouling due to the compact structure of the sp3 configuration, whereas conventional carbon electrodes with a porous sp2 configuration are fouled easily by the adsorption problem.30 An investigation of the scan rate effect of 0.1 µM oxalic acid at an ad-BDD electrode showed that the current responses of oxalic acid oxidation increased in a straight line (correlation coefficient >0.99), with an intercept of 0, relative to the square root of the scan rate in the range of 20200 mV/s (n ) 6), not linear to the scan rate (data not shown), which suggests that oxidation is under diffusion control in the interfacial area of the electrode, and the adsorption steps and specific surface reaction can be neglected. However, it appears that during potential cycling between 1 and 1.45 V (vs Ag/AgCl), a slow but continuous change of electrode activity was observed, and the oxidation peak pontential shifted to the more positive potentials. An observation of the CV of 100 µM oxalic acid at an ad-BDD electrode shows that the peak potential shifted ∼10 mV (from 1.32 to 1.33 V), and the current response increase ∼0.02 mA after 20 repetitive cycles in a stirring solution (Figure 1 Inset b). This finding means that the electrooxidation process relevant to the oxidation peak is controlled by a rate-determining step involving a surface-controlled reaction. Furthermore, although there is no significant difference of peak potentials and current responses observed at the CV from the 20th to the 50th cycles, the oxidation peak of oxalic acid totally disappeared when the electrode was oxidized to be oxygen-terminated. Comparison of the voltammetric responses at as-deposited and anodically oxidized BDD (ao-BDD) electrodes for 100 µM oxalic (29) DeClements, R.; Swain, G. M.; Dallas, T.; Holtz, M. W.; Herric, R. D., II; Stickney, J. L. Langmuir 1996, 12, 6578. (30) Shin, D.; Tryk, D. A.; Fujishima, A.; Merkoci, A.; Wang, J. Electroanal. 2005, 17, 305.

acid oxidation, PBS pH 2.1, is shown in Figure 2a and b, respectively. Whereas a well-defined oxidation peak is observed at the ad-BDD electrode, no peak is observed within the cycling potential from 0 to 2.2 V (vs Ag/AgCl) at the ao-BDD electrode. Although, it was also clarified prior to analysis, an ao-BDD electrode has much wider potential windows than ad-BDD electrodes, only a relatively small increasing current was observed at the potential range higher than 2 V (vs Ag/AgCl). The current response is not clear, since it is buried in competition with the oxygen evolution. It is known that an oxalic acid molecule is negatively charged due to its two carboxyl functional groups, which is the main reason for its repulsion from the negative electrode surface. Previously, it had been reported that by anodic treatment, the surface charge of a BDD electrode could be converted to be relatively negative due to the formation of surface carbon-oxygen functionalities.19 An increase in the significant amount of oxygen functional groups was confirmed as the change of O/C ratio from 0.03 to 0.12 after surface oxidation by XPS measurements, and oxygen coverage was ∼10% on the ao-BDD electrode surface, resulting in a negative charge of the surface.29 Our group has discussed the effect of surface termination of BDD electrodes for the oxidation of some compounds.26-28 Compounds with a positive charge, for example, glutathione, are easily oxidized at an ao-BDD electrode, and an anodic peak was shifted to the negative direction, as compared to that at an ad-BDD electrode, due to the electrostatic attraction force from carbon-oxygen functionalities.26 In comparison, anodic peaks for negative-charge compounds, such as DNA and 2,6-AQDS, were more clearly obtained at ad-BDD electrodes than those at ao-BDD electrodes due to the existence of the electrostatic repulsion.27,28 Some experiments were conducted to confirm whether the surface termination influences oxalic acid oxidation at GC electrodes. Hydrogen termination was introduced at the GC surface by annealing the film in a hydrogen-plasma CVD chamber for 20 min at 3 kW. GC is known as a composite consisting of amorphous carbon and one or more additional materials. Since it is formed by carbonizing phenolic resins, which are made by reacting Analytical Chemistry, Vol. 78, No. 10, May 15, 2006

3469

Figure 2. Cyclic voltammograms of 0.1 M PBS, pH 2.1, at (a) as-deposited BDD electrode, (b) anodically oxidized BDD electrode, (c) untreated glassy carbon, and (d) annealed glassy carbon with (bold line) and without (thin line) the presence of 100 µM oxalic acid. Scan rate was 100 mV/s. Insets show magnification of the cyclic voltammograms in the absence of oxalic acid.

celullosics, aldehydes, and ketones, a GC surface is considered to have mainly hydroxyl and carbonyl terminations. Many variations for fabricating GC results in O/C ratios of a GC surface are different from factory to factory. XPS investigation (data not shown) for untreated and annealed GC showed that O 1s and O KLL spectra decreased after the annealing treatment. That the O/C ratio of the GC electrode decreased from 0.30 to 0.03 after annealing treatment indicates that some part of the oxygen termination was changed to hydrogen termination. Previously, hydrogen-plasma treatment GC electrodes had been reported by DeClements and group.29 By exposing the GC to hydrogen plasma CVD for 12 h, a sp3 bonded “diamond-like” phase can be produced. This surface was relatively oxygen-free because hydrogen chemisorbed at the edge plane sites, replacing the oxygen functional groups. Formation of this surface could be followed by subsequent nucleation and growth of a diamond film when a carbon source was introduced.31 In this work, we have annealed the GC using an exposure time limit of ∼20 min to avoid the sp3 bonds formation. By so doing, although the GC was hydrogenated, as demonstrated by the change of the O/C ratio, sp3 bonds were not formed, as confirmed by amorphous peaks in Raman spectrometry (data not shown). Figure 2c and d show CV a comparison of oxalic acid oxidation at untreated and annealed GC, respectively. An ill-defined peak was observed at untreated GC, whereas a well-defined peak was observed at annealed GC. The oxidation peak potentials are ∼1.32 V (vs Ag/AgCl), with current responses of ∼0.63 and ∼0.47 mA/ cm2 observed at untreated and annealed GC, respectively. Interestingly, the background currents decreased drastically after annealment. For example, at the potential of ∼1.32 V (vs Ag/ (31) DeClements, R.; Hirsche, B. L.; Granger, M. C.; Xu, J.; Swain, G. M. J. Electrochem. Soc. 1996, 143, L150.

3470

Analytical Chemistry, Vol. 78, No. 10, May 15, 2006

Table 1. Summary of Oxidation Potential, Signal, and Background Current Responses as well as S/B Ratio of 100 µM Oxalic Acid in 0.1 M PBS, pH 2.1, at BDD and GC Electrodes

electrode ad-BDD anodically oxidized BDD untreated GC annealed GC

E background (V vs signal Ag/AgCl) (mA cm-2) (mA cm-2)

S/B

O/C ratio

1.32 NA

0.482

0.005

95.4

0.03 0.20

1.32 1.32

0.630 0.472

0.11 0.008

5.90 0.30 57.3 0.03

AgCl), the background current decreased from 0.11 to 0.08 mA/ cm2, enhancing the signal-to-background (S/B) ratio ∼1 order higher from 5.90 to 57.3. This enhancement is believed to translate into lower limits of detection and to increase the sensitivity of the electrodes. Table 1 summarizes the potential, signal, background current and S/B ratio of oxygen and hydrogen termination at BDD and GC electrodes. The table clearly shows that although oxygen-terminated GC has a higher O/C ratio than oxygenterminated BDD electrodes, an oxidation peak can be observed at oxygen-terminated GC, but it is absolutely not obtained at oxygen-terminated BDD electrodes. The probable reason is GC may have functional groups other than hydrogen which can promote interaction between oxalic acid and the electrode surface, in addition to an absorption control mechanism that is generally involved in the reaction at GC electrode.32 Scan rate effect was investigated at the GC electrode and showed that the current responses increased linearly to a scan rate range of 20-200 mV/s (32) Rao, T. N.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. J. Electrochem. Soc. 1999, 146, 680.

Figure 3. The first and the fifth cyclic voltammograms of 100 µM oxalic acid in 0.1 M PBS, pH 2.1, at an annealed GC electrode. Scan rate is 100 mV/s.

Figure 4. Cyclic voltammograms of 10 µM oxalic acid in 0.1 M buffer Robinson solution in the pH range of 2-10. Inset shows dependence of peak current response on the pH of the solution.

(n ) 6), which indicates that the oxalic acid oxidations were limited by adsorption control at both oxygen and hydrogen termination (data not shown). These data are in agreement with the oxidation current responses, which changed faster at hydrogenterminated GC than at hydrogen-terminated BDD electrodes, as shown in Figure 3. The figure shows that after five cycles, the peak current increased ∼0.02 mA cm-2, which suggests strong adsorption of the reactant on the surface of hydrogen-terminated GC. Moreover, the oxidation peak potential shifted at ∼0.08 V, which indicates a significant change at the active site of the surface due to the adsorption of the impurities, in addition to the oxidation of hydrogen to be oxygen termination. To the best of our knowledge, no reaction mechanism of oxalic acid at a carbon electrode has been reported yet. The pH dependence of anodic peak potentials for 10 µM oxalic acid at ad-BDD electrodes was examined in the pH range of 2-10 (Figure 4). The peak potential values linearly decreased with the increasing pH values below pH 4.5 with a slopes of ∼-60 mV/pH, which suggests that an equal number of protons and electrons are involved in the overall oxidation process. The electrochemical oxidation of oxalic acid is believed to occur involving a twoelectron, two-proton mechanism to produce CO2. This is in agreement with electrooxidation at palladium and platinum, which proposed the slow absorption step involved in the oxidation reaction to produce CO2 and hydrogen ion.7,11 The inflection point of the Ep versus pH plots at pH ∼4.5 is confirmed because it is known that the pKa of oxalic acid is 4.2. At higher pH values, the oxidation potentials were rather constant, indicating that the process was not pH-dependent. A typical dependence on pH was

also confirmed at an annealed GC electrode (data not shown), which suggests identical mechanisms occurred at ad-BDD and annealed GC electrodes. Limits of detection (LOD) and stability of the electrode were investigated by FIA. A solution of 0.1 M PBS, pH 2.1, was used as the carrier. From a hydrodynamic voltammetry i-E curve, a potential of 1.1 V (vs Ag/AgCl) was kept for applied potential by considering the half potential oxidation peak as well as the possibility of electrode surface oxidation if the electrode was applied at very high potential. In the potential, the background current response was ∼57.95 nA with a noise of ∼0.5 nA. The calibration curve was linear over the range of 50 nM to 10 mM oxalic acid concentration (n ) 6), with a slope of 110.31 nA µM-1 and a correlation coefficient of 0.995 The lowest LOD (S/N ) 3) was estimated to be ∼0.5 nM. Reproducibility was good, with an RSD of ∼2.05% for a concentration of 10 µM (n ) 10 injections). The effect of several potential interferences,13 such ascorbic acid, uric acid, sulfite, and thiocyanate, was evaluated. Although in CV the compounds were oxidized in much lower potential and, therefore, they did not disturb the peak, in FIA, the current response contributions were significant. However, a linear increase in the responses of a mixture compounds (10 µM solution of each compounds) with an increasing concentration of oxalic acid (100 nM to 10 µM, n ) 5), suggested no interference in the measurement. No deactivation of the electrodes occurred, as is shown by the excellent stability of the background current and injection current responses. With regular use, the electrode showed excellent stability for at least 3 months. The electrode is usually easily cleaned by ultrasonication in 2-propanol for 10 min without any disorder. In the case that the responses decreases very much, the electrode can be recovered completely by exposing it to hydrogen plasma for 10 min. The results indicate that BDD electrodes can be used as the best electrodes for oxalic acid detection by controlling the surface termination. CONCLUSIONS As-deposited diamond electrodes have been successfully demonstrated for the oxidation of oxalic acid over wide ranges of pH and concentration. The result shows the superiority of a hydrogen-terminated carbon electrode, in comparison with an oxygen-terminated one, for oxalic acid oxidation. A clear and distinct peak was obtained for oxalic acid oxidation at both hydrogen-terminated BDD and GC electrodes, although due to its amorphousness and the complexity of its surface, less stability of current response was observed at hydrogen-terminated GC. In conclusion, oxidation of oxalic acid at hydrogen-terminated BDD electrodes confirmed the importance of controlling surface termination for the use of BDD electrodes for some charged molecules, especially oxalic acid. ACKNOWLEDGMENT A grateful acknowledgment is made to the JSPS (Japan Society for Promoting Science) and COE-KEIO-LCC (Life Conjugated Chemistry) programs for supporting this research. Received for review November 15, 2005. Accepted March 24, 2006. AC052029X

Analytical Chemistry, Vol. 78, No. 10, May 15, 2006

3471