Structure and Electrochemical Properties of Carbon Films Prepared by

National Institute of Advanced Industrial Science and Technology, 1-1-1 Higashi, Tsukuba, ... Japan, Department of Applied Chemistry, Keio University,...
0 downloads 0 Views 220KB Size
Anal. Chem. 2007, 79, 98-105

Structure and Electrochemical Properties of Carbon Films Prepared by a Electron Cyclotron Resonance Sputtering Method Jianbo Jia,† Dai Kato,‡ Ryoji Kurita,† Yukari Sato,† Kenichi Maruyama,‡,§ Koji Suzuki,‡,§ Shigeru Hirono,| Toshihiro Ando,⊥ and Osamu Niwa*,†,‡

National Institute of Advanced Industrial Science and Technology, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan, Core Research for Evolutional Science and Technology (CREST) of Japan Science and Technology (JST), 4-1-8, Honcho, Kawaguchi, Saitama 332-0012, Japan, Department of Applied Chemistry, Keio University, 3-14-1 Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan, NTT-Afty Corporation, 2-35-2 Hyoe, Hachioji, Tokyo 192-0918, Japan, and National Institute for Material Science, 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japanu

This paper describes the characterization, electrochemical properties, and applications of carbon films prepared by the electron cyclotron resonance (ECR) sputtering method. The ECR-sputtered carbon film was deposited within several minutes at room temperature. The optimized sputtering conditions significantly change the film structure, which includes many more sp3 bonds (sp3/sp2 ) 0.702) than previously reported film (sp3/sp2 ) 0.274)1 with an extremely flat surface (0.7 Å). The ECR-sputtered carbon films exhibit excellent electrochemical properties. For example, they have nearly the same potential window in the positive direction as that of high-quality, borondoped diamond (moderately doped, 1019-1020 boron atoms/cm3)2 and an even wider potential window in the negative direction with a low background current, high stability, and suppression of fouling by electroactive species without pretreatment. The electron-transfer rates at ECR-sputtered carbon films are similar to those of glassy carbon (GC) for Ru(NH3)62+/3+ and Fe(CN)63-/4-, whereas they are much slower than those of GC for Fe2+/3+, dopamine oxidation, and O2 reduction due to weak interactions between electroactive species and the ECR-sputtered carbon film surface. Such a response can be attributed to the ultraflat surface and low surface O/C ratios of ECR-sputtered carbon films. ECR-sputtered carbon film is advantageous for measuring biochemicals with high oxidation potentials because of its wide potential window and high stability. Highly reproducible and welldefined cyclic voltammograms were obtained for histamine and azide ions with a peak potential at 1.25 and 1.12 V vs Ag/AgCl, respectively. The film is very stable for continuous voltammetry measurements in 10 µM * Corresponding author. Tel: +81-29-861-6158. Fax: +81-29-861-6177. Email: [email protected]. † National Institute of Advanced Industrial Science and Technology. ‡ JST-CREST. § Keio University. | NTT-Afty Corp. ⊥ National Institute for Material Science.

98 Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

bisphenol A, which usually fouls the electrode surface with oxidation products. Carbon materials have been widely used in electroanalysis, especially for biological sensing, because they have a number of advantageous properties.3-8 Recently, several methods have been reported for the preparation of new carbon-based thin films. These include the plasma chemical vapor deposition (CVD) of organic gases,6-10 sputtering,1,11-14 electron beam evaporation,15-17 pyrolysis of polymeric thin films,18-20 filtered cathodic vacuum arc (1) You, T.; Niwa, O.; Tomita, M.; Ichino, T.; Hirono, S. J. Electrochem. Soc. 2002, 149, E479-E484. (2) Granger, M. C.; Xu, J. S.; Strojek, J. W.; Swain, G. M. Anal. Chim. Acta 1999, 397, 145-161. (3) McCreery, R. L. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1991; Vol. 17, pp 221-374. (4) McCreery, R. L. In Interfacial Electrochemistry; Wieckowski, A., Ed.; Marcel Dekker: New York, 1999; Chapter 35. (5) Friedrich, J. M.; Ponce-de-Leon, C.; Reade, G. W.; Walsh, F. C. J. Electroanal. Chem. 2004, 561, 203-217. (6) Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.; Show, Y.; Swain, G. M. Diamond Relat. Mater. 2003, 12, 1940-1949. (7) Compton, R. G.; Foord, J. S.; Marken, F. Electroanalysis 2003, 15, 13491363. (8) Fujishima, A.; Einaga, Y.; Rao, T. N.; Tryk, D. A. Diamond Electrochemistry, Elsevier B. V., Amsterdam, 2005. (9) Eriksson, A.; Norekrans, A.-S.; Carlsson, J.-O. J. Electroanal. Chem. 1992, 324, 291-305. (10) Mani, R. C.; Sunkara, M. K.; Baldwin, R. P.; Gullapalli, J.; Chaney, J. A.; Bhimarasetti, G.; Cowley, J. M.; Rao, A. M.; Rao, R. J. Electrochem. Soc. 2005, 152, E154-E159. (11) Schlesinger, R.; Bruns, M.; Ache, H.-J. J. Electrochem. Soc. 1997, 144, 6-15. (12) Zeng, A.; Liu, E.; Tan, S. N.; Zhang, S.; Gao, J. Electroanalysis 2002, 14, 1110-1115. (13) Hirono, S.; Umemura, S.; Kaneko, R. Appl. Phys. Lett. 2002, 80, 425-427. (14) Benlahsen, M.; Cachet, H.; Charvet, S.; Debiemme-Chouvy, C.; Deslouis, C.; Lagrini, A.; Vivier, V. Electrochem. Commun. 2005, 7, 496-499. (15) Blackstock, J. J.; Rostami, A. A.; Nowak, A. M.; McCreery, R. L.; Freeman, M. R.; McDermott, M. T. Anal. Chem. 2004, 76, 2544-2552. (16) Kiema, G. K.; Brett, M. J. J. Electrochem. Soc. 2003, 150, E342-E347. (17) Kiema, G. K.; Brett, M. J. J. Electrochem. Soc. 2004, 151, E194-E198. (18) Niwa, O.; Tabei, H. Anal. Chem. 1994, 66, 285-289. (19) Ranganathan, S.; McCreery, R. L.; Majji, S. M.; Madou, M. J. Electrochem. Soc. 2000, 147, 277-282. (20) Ranganathan, S.; McCreery, R. L. Anal. Chem. 2001, 73, 893-900. 10.1021/ac0610558 CCC: $37.00

© 2007 American Chemical Society Published on Web 11/29/2006

deposition,21 and ion beam deposition.22 Results from these studies show that carbon films differ significantly as regards such electrochemical properties as potential window, capacitance, conductivity, and stability as a result of the different deposition procedures. The development of new carbon-based, thin-film electrodes with wide potential windows is very important if we are to extend the range of available applied potentials and thus detect analytes with high oxidation or low reduction potentials. Boron-doped diamond (BDD) has attracted considerable interest for such applications because of its excellent electrochemical properties.6-8 The macroscopic surface of BDD has micrometer-order roughness because the film consists of microcrystalline diamonds and their associated grain boundaries. Although the microcrystalline diamonds at BDD contribute to the wide potential window and low capacitive current, it is difficult to fabricate the film into the microor nanosized electrodes now required for detecting small-volume biomolecules. The flat surface is not only useful for such applications but also important when using the films as substrates for studying adsorbed molecules using scanning probe microscopes with a high resolution and as narrow gap electrodes for molecular electronics. In addition, a high temperature is usually required when depositing BDD (700-1000 °C),2 and so there are limitations as regards applications such as the available substrate materials. Recently, nanocrystalline diamonds with a much flatter surface have been reported.23 However, the rms surface roughness of nanocrystalline diamond is currently 34 nm, which is still much rougher than that of the graphite-like carbon films with nearatomic flatness obtained by the pyrolysis of photoresist film or the electron beam evaporation method reported recently.15,20 These carbon films show similar electrochemical properties to a very flat version of glassy carbon (GC).15,20 The potential window of these ultraflat carbon films is not as wide as that of BDD since the film mainly has a graphite structure. Another class of carbon materials is tetrahedrally bonded nitrogen-doped amorphous carbon prepared by ion beam deposition or sputtering.12,21,24 It has low resistivity and has achieved a wide potential window and active charge-transfer properties.24 Various kinds of sputtering methods are very often employed to form carbon thin films. The sputtering method is widely used because of its versatility, its ability to sputter many materials, and the ease with which it can produce carbon film on a large scale.25 Carbon film prepared by the electron cyclotron resonance (ECR) sputtering method forms a unique carbon film that is different from amorphous carbon.13 The film can be deposited at room temperature with an ultraflat surface and controllable sp3/sp2 ratios in a relatively short time.26 Growth at a moderate temperature on many kinds of substrates with ultraflat surfaces offers important advantages over the difficult nucleation and high-temperature growth technique required for BDD film, especially for application (21) Yee, N. C.; Shi, Q. F., Cai, W. B.; Scherson, D. A.; Miller, B. Electrochem. Solid-State Lett. 2001, 4, E42-E44. (22) Pleskov, Y. V.; Krotova, M. D.; Polyakov, V. I.; Khomich, A. V.; Rukovishnikov, A. I.; Druz, B. L.; Zaritskiy, I. J. Electroanal. Chem. 2002, 519, 6064. (23) Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15, 879-888. (24) Yoo, K. S.; Miller, B.; Kalish, R.; Shi, X. Electrochem. Solid-State Lett. 1999, 2, 233-235. (25) Robertson, J. Mater. Sci. Eng. R 2002, 37, 129-281. (26) Niwa, O. Bull. Chem. Soc. Jpn. 2005, 78, 555-571.

to the fabrication of micro- and nanosized electrodes. The ECR sputtering method is clearly inexpensive as it has already used for the past decade for fabricating optical coating film for semiconductor laser devices, passivation film for SAW devices, and insulating film for magnetic heads. The electrical conductivity of ECR-sputtered carbon film is 19 orders of magnitude greater than that of diamond without doping, and its hardness is comparable to that of diamond.13 In our preliminary study, an ECRsputtered carbon film electrode (sp3/sp2 ) 0.274) exhibited significant advantages over a GC electrode for the electrochemical oxidation of alkylphenols.1,26 More recently, we applied an ECRsputtered carbon film electrode to detect oligonucleotides (4 mer) without pretreatment of the electrode surface.27 In this paper, we report the structure and basic electrochemical properties of ECR-sputtered carbon films with different sp3 content, which we expect to extend the potential window and film stability of adsorptive electroactive species. The properties were also compared with results for GC and BDD. The surface structure of ECR-sputtered carbon film was characterized by atomic force microscopy (AFM) and X-ray photoelectron spectroscopy (XPS). The basic electrochemical properties of the ECR-sputtered carbon films were studied by measuring the cyclic voltammetry of several redox species with different electron-transfer properties. We also studied the use of ECR-sputtered carbon film for measuring certain biochemicals with high oxidation potentials such as histamine and azide ions. The stability of ECR-sputtered carbon film was also studied by measuring bisphenol A (BPA), which usually fouls the conventional carbon and metal electrode surfaces after electrochemical oxidation. EXPERIMENTAL SECTION Carbon Film Preparation. Carbon films were deposited on highly doped silicon (100) substrates with ECR sputtering equipment at room temperature. The advantage of this method is that the ion irradiation of the growing surface during deposition is highly controllable. ECR sputter equipment is now commercially available28 and commonly used for preparing various kinds of thin films with which to fabricate electronic and optical devices. The microwave power and dc voltage applied to the carbon target were 500 W and 500 V, respectively. The Ar gas pressure was 5.0 × 10-2 Pa. We used ion acceleration voltages of 20 and 75 V to form carbon films with different structures, and abbreviated as ECR20 and ECR-75, respectively. The ECR-sputtered carbon films, which were obtained in ∼8 min, were typically 40 nm thick. The wafer was cut to a rectangular shape. We then fixed a plastic tape with a 2-mm-diameter hole in it onto the carbon film to form a disk electrode. BDD was grown on Si (100) by the CVD method using CH4/ H2/B2H6 source gas.29 In brief, the deposition was carried out with a gaseous mixture of methane 1.0 vol % and hydrogen gas at a microwave power of 800 W, a substrate temperature of 800 °C, and a total gas pressure of 50 Torr. The film was doped with boron by introducing diborane gas to the methane-hydrogen mixture with a 10 ppm atomic ratio B/C in the gas phase. A film thickness of ∼2 µm was achieved after 4 h of deposition. Under these (27) Niwa, O.; Jia, J.; Sato, Y.; Kato, D.; Kurita, R.; Maruyama, K.; Suzuki, K.; Hirono, S. J. Am. Chem. Soc. 2006, 128, 7144-7145. (28) http://www.ntt-afty.co.jp/en/products/film_business/film01.html

Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

99

Figure 1. High-resolution decomposition XPS spectra of ECR-sputtered carbon films. (A) ECR-75; (B) ECR-20.

conditions, the narrow and sharp peak assigned to high-quality diamond can be observed in Raman spectra. Carbon Film Characterization. AFM measurements were performed with an Olympus NV-2000 instrument in air at room temperature. Measurements were made with triangular SiN microcantilever tips. The scan rate was 1 µm/s with a 10-nN load. XPS was conducted with a Kratos AXIS Ultra (Al KR 1486.6 eV) spectrometer with a resolution of 0.65 eV to determine the elemental composition of the film surface. Electrochemical Measurements. Cyclic voltammetry was performed using an HZ-3000 automatic polarization system (Hokuto Denko). All experiments were carried out using a conventional three-electrode system with a platinum wire as the auxiliary electrode and an Ag/AgCl (3 M NaCl) reference electrode. ECR-sputtered carbon film, BDD, or freshly polished GC was used as the working electrode. The GC (BAS, Tokyo, Japan) was polished in Al2O3/water slurry and ultrasonicated in ultrapure water for 5 min before use. The electrolyte solutions were purged with pure Ar for 20 min prior to the measurement and blanketed with Ar during the measurement in the deoxygenated experiments. For an oxygen reduction reaction (ORR), the solutions were purged with pure O2 for 20 min prior to the measurement and blanketed with O2 during the measurement. The electrochemical cell was located in a Faraday cage at room temperature for all the measurements. The heterogeneous electrontransfer rate constants were determined from the anodic/cathodic peak separation using the method developed by Nicholson30 by assuming R ) 0.5 and employing the following diffusion coefficients: Ru(NH3)63+/2+, DO ) DR ) 5.5 × 10-6 cm2/s, and Fe(CN)63-/4- DO ) 7.63 × 10-6 cm2/s, DR ) 6.32 × 10-6 cm2/ s.31,32 Chemicals. All chemicals were analytical grade, or better, and were used as received. Solutions of 0.10 mM hexaammineruthenium(III) chloride (Aldrich, Milwaukee, WI) and potassium ferrocyanide (Nacalai Tesque, Inc., Kyoto, Japan) in 1.0 M potassium chloride, 1.0 mM ammonium iron sulfate (SigmaAldrich, St. Louis, MO), and 0.50 mM dopamine (Nacalai Tesque, Inc.) in 0.10 M perchloric acid (Kanto, Tokyo, Japan) were 100

Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

prepared daily. The electrolyte for histamine (Wako, Osaka, Japan), sodium azide (Sigma, St. Louis, MO), and for BPA (Wako) was a mixture of 0.1 M KH2PO4 + 0.1 M Na2HPO4 (pH 7.2 PB). BPA (1 mM) in 0.1 M sodium sulfate solution was used as a stock solution, which was diluted to a suitable concentration with the electrolyte solution. Ultrapure water (Milli-Q) was used in all the experiments. RESULTS AND DISCUSSION Surface Characterization. An AFM image of the ECRsputtered carbon film determined from an image size of 500 × 500 nm shows that the mean surface roughness is only 0.7 Å.27 This is much flatter than the value of 34 nm for nanocrystalline diamond film23 and almost the same as that of the carbon films prepared by electron beam evaporation or the pyrolysis of photoresist reported by McCreery et al.15,20 Such flatness should be very useful in terms of fabricating micrometer- or nanometersized electrodes for use in electroanalysis and for molecular electronics such as nanogap electrodes, and the observation of adsorbed molecules by using a high-resolution scanning probe microscope. C 1s spectrum analysis is commonly employed for determining the relative concentration of sp2 and sp3 hybrid carbons because it is a very useful method for determining the relative concentration of sp2 and sp3 hybrid carbons.33,34 Figure 1 shows high-resolution C 1s XPS spectra of ECR-75 and ECR-20. These C 1s XPS spectra can be decomposed into six peaks (C1C6). The peaks that appeared at 284.3 (C1) and 285.3 eV (C2) are assigned to sp2 and sp3 hybrids, respectively.33 The sp3/sp2 ratios of ECR-75 and ECR-20 are 0.702 and 0.190, respectively. The improved deposition conditions (75-V acceleration voltage) (29) Ushizawa, K.; Gamo, M. N.; Watanabe, K.; Sakaguchi, I.; Sato, Y.; Ando, T. J. Raman Spectrosc. 1999, 30, 957-961. (30) Nicholson, R. S. Anal. Chem. 1965, 37, 1351-1355. (31) Gerhardt, G.; Adams, R. N. Anal. Chem. 1982, 54, 2618-2620. (32) Kovach, P. M.; Deakin, M. R.; Wightman, R. M. J. Phys. Chem. 1986, 90, 4612-4617. (33) Diaz, J.; Paolicelli, G.; Ferrer, S.; Comin, F. Phys. Rev. B 1996, 54, 80648069. (34) Chu, P. K.; Li, L. Mater. Chem. Phys. 2006, 96, 253-277.

Figure 2. Voltammograms of different electrodes in 0.05 M H2SO4 deoxygenated with Ar; scan rate, 0.1 V/s.

can greatly increase the sp3/sp2 ratio compared with previously reported film (sp3/sp2 ) 0.274) used for electrochemical measurements.1 The C3 and C4 peaks are assigned to the CsO bond,35 which is caused by the oxygen-containing groups at the ECRsputtered carbon film surface. The surface O/C ratio of ECR-75 from XPS was 3.3% two weeks after deposition. This is much less than the value of 14% for GC polished in Al2O3/water slurry.36 One of the present authors has studied the structure of ECRsputtered carbon films with TEM.13 The ECR-sputtered carbon films consist of nanocrystallites whose lattice fringes are parallel or curved and fine-closed structures. (TEM image of ECR film is provided as Supporting Information.) A parallel and curved graphene region is typically observed in graphite-like carbon. When the sp3/sp2 ratio was increased, the area containing parallel or curved structures decreased and the fine-closed region increased, indicating that the nanostructure of ECR-sputtered carbon film prepared with a higher ion acceleration voltage is different from a typical graphite structure. In contrast, fine graphene sheets were vertically oriented to the film surface in the ECR-sputtered carbon film. At a higher ion acceleration voltage, the sp2 nanocrystallites are connected with adjacent crystallites by sp3 bonding; therefore, the film has greater hardness than graphite-like carbon. At the same time, the conductivity of the ECR-sputtered carbon film (typically from 20 to 38 S/cm) is much higher than that of diamond1,13 since sp2 bonds account for at least more than 50% of the total number of bonds. Potential Window and Capacitive Current. As shown in Figure 2, we obtained voltammograms for different electrodes in 0.05 M H2SO4 so that we could compare their potential windows. The potential window is defined as the potential range between the current limits that do not exceed (500 µA/cm2, as previously reported by Swain et al.2 The potential windows for ECR-75 and ECR-20 are 3.7 (-1.7 to 2.0 V vs Ag/AgCl in 0.05 M H2SO4) and 3.4 V (-1.5 to 1.9 V), respectively. In contrast, the potential window for BDD (10 ppm) is 3.1 V (-1.0 to 2.1 V), and for GC is 2.6 V (-0.9 to 1.7 V) under the same conditions. ECR-75 (sp3/sp2 ) 0.702) exhibits a wider potential window than that of ECR-20 (35) Tillborg, T.; Nillson, A.; Martensson, N. Surf. Sci. 1992, 273, 47-60. (36) Chen, P.; McCreery, R. L. Anal. Chem. 1996, 68, 3958-3965.

(sp3/sp2 ) 0.190). Both ECR-sputtered carbon films also show much wider potential windows than GC, suggesting that the sp3 carbon ratio greatly affects the potential window even when the sp3 carbon ratio is low. ECR-75 exhibits a nearly equivalent potential window as that of BDD on the positive side of the potential region in spite of the relatively large sp2 bond content. In the negative region, the high overpotential for hydrogen evolution indicates that the adsorption of protons or hydronium ions is weak at the ECR-sputtered carbon film surfaces. The wider overpotential of the ECR-sputtered carbon film could be due to its undoped state. In contrast, the potential window of BDD becomes narrower with increasing doping level.37 The electrode capacitance value (C°) was obtained from CVs at 0.25 V versus Ag/AgCl in 1.0 M KCl. The C° values for ECR75 and ECR- 20 are 15.7 and 22.2 µF/cm2, which means they are 5.3 and 3.8 times lower than that of GC (84.2 µF/cm2), respectively. At the same time, ECR-75 shows a 3.1 times higher capacitance than BDD (5.1 µF/cm2), which has a clear Raman peak and also a low sp2 region.29 It is reported that the ultraflat surface could account for the fact that the capacitance is 2-4 times lower that of GC.20 This indicates that other factors besides surface ultraflatness contribute to the low C° value of the ECR-75. As discussed earlier and in a previous paper,13 the low surface O/C ratio may contribute to the low C° value particularly for ECR-75. Swain’s group reported polycrystalline BDD films with various levels of sp2-bonded non-diamond carbon.38 The sp2-bonded carbon impurity content occurred as the methane-to-hydrogen (C/H) source gas mixture ratio increased. Films prepared with a C/H ratio of less than 3% exhibited a low C° and a wide potential window. In contrast, a significantly reduced potential window and an enlarged background current that resembles that of freshly polished GC can be observed when the C/H ratio is 5%. Raman spectroscopy results indicate that the non-diamond carbon impurity for the film (C/H ) 5%) is largely sp2-bonded, much like the turbostratic microstructure of GC. The nondiamond carbon contains a mixed sp2/sp3 microstructure with a limited number of sites where redox-active carbon-oxygen functionalities form. With ECR-sputtered carbon, the much lower capacitance than that of GC (but still a higher capacitance than that of high-quality BDD) could be due to a mixed sp2/sp3 microstructure with a limited number of sites where redox-active carbon-oxygen functionalities form in addition to an extremely flat surface. Therefore, the fact that ECR-75 has a lower C° than ECR-20 in spite of a similar surface flatness could be due to the stability of the sp2/sp3 mixed structure, which is a fine-closed nanocrystalline structure, unlike that of graphite. Electrochemical Behavior for Different Redox Species. Several redox systems exhibiting different electrode kinetics36 were used to study the electrochemical properties of ECRsputtered carbon films. Figure 3 shows the voltammograms of various redox species at different electrodes, and Table 1 summarizes the kinetic parameters obtained from Figure 3. Table 1 also includes values obtained from other previously reported ultraflat carbon films and high quality BDD.15,20,43 It is wellestablished that the electron-transfer rate constant of Ru(NH3)63+/2+ (37) Ndao, A. N.; Zenia, F.; Deneuville, A.; Bernard, M.; Levy-Clement, C. Diamond Relat. Mater. 2000, 9, 1175-1180. (38) Bennett, J. A.; Wang, J.; Show, Y.; Swain, G. M. J. Electrochem. Soc. 2004, 151, E306-E313.

Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

101

Figure 3. Voltammograms of four kinds of redox systems at different electrodes deoxygenenated with Ar: (A) 0.10 mM Ru(NH3)63+/2+ in 1.0 M KCl, scan rate, 0.1 V/s. (B) 0.10 mM Fe(CN)64-/3- in 1.0 M KCl, scan rate, 0.01 V/s. (C) 1.0 mM Fe2+/3+ in 0.1 M HClO4, scan rate, 0.1 V/s. (D) 0.50 mM DA in 0.1 M HClO4, scan rate, 0.1 V/s.

Table 1. Comparison of Electron-Transfer Kinetics at Different Electrodes

electrode ECR-75 ECR-20 BDD 10 ppm BDD (NRL) 43 BDD (USU) 43 GC Py 7-nm ECF 15 7-nm ECF 15 Py 200-nm ECF 15 200-nm ECF 15 PPF (1-2 µm) 20

k° for k° for ∆Ep ∆Ep of Ru(NH3)63+/2+ Fe(CN)63-/4- of DA Fe2+/3+ (× 10-2 cm/s) (× 10-2 cm/s) (mV) (mV) reva reva 3.0 1.2 1.7 5.73 4.6 1.9 2.7 4.3 2.0

1.22 0.24 reva 1.7 1.9 1.97 1.4 0.27 2.9 0.57 1.2

472 464 734 480 509 58 227 119 243 130 287

587 571 686 581 837 206

654

a Indicates that ∆E is very close to the reversible limit and no p attempt was made to quantify k°.

is insensitive to surface chemistry or an adsorbed monolayer and thus is considered to be a simple outer-sphere system. The most important factor affecting the reaction rate is the electronic properties of the electrode, specifically the density of the electronic states near the formal potential of the redox system.36,39 The redox (39) Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 74, 2553-2560.

102 Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

reaction of Ru(NH3)63+/2+ at ECR-sputtered carbon films is reversible, indicating that the density of the electronic states is sufficient to support the rapid electron transfer of Ru(NH3)63+/2+. k° for Fe(CN)64-/3- at GC is different with various pretreatment methods; for example, k° is 0.5 cm/s for laser activated GC.40,41 The k° value does not appear to depend on surface oxides except for a fairly small Frumkin effect.42 The ECR-sputtered carbon films show electrode kinetics for Fe(CN)63-/4- similar to that observed at BDD and polished GC as shown in Figure 3B. In both cases, the peak current is linear with v1/2 between scan rates 0.01 and 2.0 V/s for both kinds of ECR-sputtered carbon films, which implies that both Ru(NH3)63+/2+ and Fe(CN)63-/4- are behaving as a semiinfinite linear diffusion-controlled system. ECR-sputtered carbon films exhibit active responses for Ru(NH3)63+/2+ and Fe(CN)63-/4similar to that of BDD without pretreatment. However, timeconsuming pretreatment (such as polishing or high potential application) is very often required to activate conventional sp2 carbon and metal electrodes to improve electron transfer.39,41 (40) Rice, R. J.; Pontikos, N. M.; McCreery, R. L. J. Am. Chem. Soc. 1990, 112, 4617-4622. (41) Ranganathan, S.; Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71, 35743580. (42) Deakin, M. R.; Stutts, K. J.; Wightman, R. M. J. Electroanal. Chem. 1985, 182, 113-122.

As also shown in Figure 3 and Table 1, the electron-transfer rates for Fe3+/2+ and dopamine (DA) at ECR-sputtered carbon films are several orders of magnitude lower than that at GC, whereas they are comparable to the values at BDD. Fe3+/2+ is an inner-sphere type, and its k° value is very sensitive to the surface carbon-oxygen functionalities, specifically carbonyl groups, for both GC and BDD.36,38,39,41,43 The properties of the ECR-sputtered carbon films are similar to those of hydrogenated glassy carbon (HGC) in the slow electron-transfer kinetics of Fe3+/2+.44,45 These can be attributed to low carbon-oxygen functional groups at both kinds of material surfaces. However, the electron-transfer kinetics of DA is significantly different for ECR-sputtered carbon film and HGC; namely, it is slower at the former than the latter. The ∆Ep of DA for HGC is 220 mV, while these values are more than 460 mV for both kinds of ECR-sputtered carbon films.44 The oxidation of DA is a multistep process and known to involve reactant adsorption at the electrode surface. Previous results show that the adsorption of DA at the electrode surface is necessary for fast electron transfer.43,46 DA has fast kinetics and increased adsorption on highly oxidized GC, which contains many surface anionic sites.46 The ∆Ep of DA for HGC is similar to that for a polished GC.45 The large peak separations for DA at ECR-sputtered carbon films compared to that of GC and HGC indicate weak interactions between DA and the film surfaces. The electrode reaction kinetics for Fe3+/2+ and DA become faster after ECR-sputtered carbon film surfaces are oxidized (for example, after electrochemical pretreatment) since the interaction between the electrode surface and this redox system changed after surface oxidation. This is similar to that of the pyrolyzed photoresist films (PPFs) reported by McCreery’s group.20 However, the stability of ECR-sputtered carbon film could be much higher than that of PPF reported as the electrochemical cycling of ECR-sputtered carbon film, which was conducted in 0.1 M HClO4 between the potential limits -0.5 and +1.5 V, especially for ECR-75. ORR was also measured in both 0.1 M HClO4 and 0.1 M KOH solutions with ECR-sputtered carbon films. The reduction peak potential is about 0.62∼0.65 V at a scan rate of 0.01 V/s at both ECR-75 and ECR-20. ORR is a surface-sensitive reaction, and the kinetic overpotential is dependent on the sites available. The absence of a significant pH effect for ORR at ECR-sputtered carbon films indicates that the rate-determining step may be the 1ereduction of O2 to form superoxide, which is a pH-independent reaction.38 The absence of a significant pH effect for ORR at BDD has also been reported.38,47,48 In contrast, ORR is pH dependent on clean and unmodified GC.49 The slower electron-transfer kinetics for O2 reduction at ECR-sputtered carbon films also shows that the films interact more weakly with O2 than with GC, resulting in a reduction in the catalytic effect of the carbon surface. Such a response can be attributed to surface flatness and a low surface O/C ratio at ECR-sputtered carbon films. (43) Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang, M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M. Anal. Chem. 2000, 72, 3793-3804. (44) Chen, Q.; Swain G. M. Langmuir 1998, 14, 7017-7026. (45) Kuo, T. C.; McCreery, R. L. Anal. Chem. 1999, 71, 1553-1560. (46) DuVall, S. H.; McCreery, R. L. J. Am. Chem. Soc. 2000, 122, 6759-6764. (47) Yano, T.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. J. Electrochem. Soc. 1998, 145, 1870-1875. (48) Yano, T.; Popa, E.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. J. Electrochem. Soc. 1999, 146, 1081-1087.

Figure 4. Voltammograms in 0.1 M pH 7.2 PB with (solid line) and without (dotted line) 50 µM histamine at ECR-75, scan rate, 0.1 V/s. Inset, GC under the same conditions.

Analytical Applications. The wide potential window and low background current of ECR-75 mean it should be used to measure chemicals with high oxidation potentials. Histamine is an important biogenic amine present in many food products and which acts as a chemical messenger in biological systems. Although electrochemical detection methods achieve high sensitivity, there have been few studies using BDD or carbon fiber microdisk bundle electrodes for the direct oxidation of histamine by electrochemical methods.50,51 This is due to the high oxidation potential and electrode surface fouling by the oxidized products of histamine. The oxidation peak of histamine is observed at a GC electrode with a high background current as shown in Figure 4. This large background current in the vicinity of the histamine oxidation potential results from a combination of oxygen evolution and electrode surface oxidation. In contrast, highly reproducible CVs were obtained with a peak potential at 1.25 V versus Ag/ AgCl at ECR-75 when the potential was continuously scanned. In contrast, the peak current decreased rapidly at GC under the same conditions. The peak current was linear in relation to histamine concentration in the examined range (1.0-100 µM) in 0.1 M pH 7.2 PB at ECR-75, where r ) 0.997, and the detection limit was 0.2 µM at S/N ) 3. Such a detection limit is even more sensitive than the value previously reported with BDD (1 µM).50 Azide ions are highly toxic and cause health hazards at relatively modest levels. In addition, azide ions undergo irreversible electrooxidation to nitrogen gas at carbon electrodes.52-54 Figure 5 shows CVs for ECR-75 and GC in 0.1 M pH 7.2 PB with and without 50 µM azide ions. An oxidation peak is observed at both electrodes, but the background current at GC is significantly larger than that at ECR-75. In contrast, well-defined CVs were obtained with their peak potential at 1.12 V versus Ag/AgCl at ECR-75 for azide ions. The peak current was linear in relation to (49) Yang, H.-H., McCreery, R. L. J. Electrochem. Soc. 2000, 147, 3420-3428. (50) Sarada, B. V.; Rao, T. N.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2000, 72, 1632-1638. (51) Weng, Q. F.; Xia, F. Q.; Jin, W. R. J. Chromatogr., B 2002, 779, 347-352. (52) Dalmia, A.; Wasmus, S.; Savinell, R. F.; Liu, C. C. J. Electrochem. Soc. 1996, 143, 556-560. (53) Xu, J. S.; Swain, G. M. Anal. Chem. 1999, 71, 4603-4608. (54) Xu, J. S.; Swain, G. M. Anal. Chem. 1998, 70, 1502-1510.

Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

103

Figure 5. Voltammograms in 0.1 M pH 7.2 PB with (solid line) and without (dotted line) 50 µM azide ions at ECR-75, scan rate, 0.01 V/s. Inset, GC under the same conditions.

Figure 6. Variation in the peak current of CVs for 10 µM BPA in 0.1 M pH 7.2 PB at GC and ECR-75 without pretreatment, respectively, scan rate, 0.01 V/s. The interval of each CV is 5 min.

the azide ion concentration in the examined range (1.0-100 µM) in 0.1 M PB at ECR-75 with r ) 0.998. The detection limit is 0.4 µM at S/N ) 3. The ECR-75 provides a low detection limit to azide ions that almost equals that at BDD, while it is more sensitive than that at GC (19 µM).54 These results suggest that ECR-75 can provide performance comparable to that of BDD for measuring analytes with high oxidation potentials of over 1.0 V. In addition, the fabrication of ECR-sputtered film takes only few minutes (e.g., ∼8 min with a thickness of 40 nm), and it easy to microfabricate the film into any electrode shape with conventional lithographic and etching techniques because the film can be obtained with a thickness of several tens of nanometers and a flat surface. In contrast, BDD is usually several to tens of micrometers thick and requires a longer deposition time. Moreover, the BDD microfabrication process is more complicated, as previously reported.55,56 Weak interactions between an ECR-sputtered carbon film surface and analytes should prove very useful for the detection of chemicals that are easily adsorbed on conventional electrode materials. Phenols can be used as model compounds for examining electrode stability because of their well-known electrodefouling properties for GC.1,57 BPA exhibits estrogenic activity, thereby serving as an environmental endocrine disrupter.58,59 However, it is difficult to determine BPA with an electrochemical method because electropolymerized BPA film is tightly adsorbed at the electrode surface even at very low concentrations.60 We compared the reproducibility of oxidation peak currents of 10 µM BPA at ECR-75 and at GC. The peak current changed very little during the measurement period at the ECR-75 at intervals of 5 min, as shown in Figure 6. In contrast, the peak current of BPA

at GC is initially high due to its greater roughness and stronger interaction with BPA than ECR-75, while the current decreased quickly and disappeared after 10 cycles of the slow potential scan (10 mV/s). This indicates that the ECR-75 surface is much less easily passivated by the oxidized products of phenols than that of GC. Such a response can be attributed to the surface flatness and low concentration of oxygen-containing groups of ECR-75.1 Since there are various kinds of chemicals for adsorption at the electrode surface in the biological and environmental samples, the less severe fouling characteristic of ECR-sputtered carbon film is greatly advantageous in terms of measuring such samples. The film should find various applications as an electrochemical detector coupled to liquid chromatography and capillary electrophoresis for analytes with high oxidation potential or where the electrode surface is easily fouled by oxidation products. ECR-sputtered carbon film could also be integrated in microflow channels to produce miniaturized electrochemical detectors since the film is more easily miniaturized into micro- and nanosized electrochemical detectors than BDD because of its flatter surface and thinner film thickness.

(55) Cvacka, J.; Quaiserova, V.; Park, J.; Show, Y.; Muck, A.; Swain, G. M. Anal. Chem. 2003, 75, 2678-2687. (56) Pagels, M.; Hall, C. E.; Lawrence, N. S.; Meredith, A.; Jones, T. G. J.; Godfried, H. P.; Pickles, C. S. J.; Wilman, J.; Banks, C. E.; Compton, R. G.; Jiang, L. Anal. Chem. 2005, 77, 3705-3708. (57) Gattrell, M.; Kirk, D. W. Canadian J. Chem. Eng. 1990, 68, 997-1003. (58) Krishnan, A. V.; Stathis, P.; Permuth, S. F.; Tokes, L.; Freldman, D. Endocrinology 1993, 132, 2279-2286. (59) Brotons, J. A.; Oleaserrano, M. F.; Villalobos, M.; Pedraza, V.; Olea, N. Environ. Health Perspect. 1995, 103, 608-612. (60) Kuramitz, H.; Nakata, Y.; Kawasaki, M.; Tanaka, S. Chemosphere 2001, 45, 37-43.

104 Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

CONCLUSION We developed ECR-sputtered carbon film deposited under improved conditions with a higher sp3 content (sp3/sp2 ) 0.702). The film has an ultraflat surface and exhibits excellent electrochemical properties including a wider potential window and a lower background current than that of film with a lower sp3 content (sp3/sp2 ) 0.190) and a GC electrode and is nearly comparable to that of high-quality, moderately doped diamond. The films exhibit comparable electron-transfer rates to those of GC for Ru(NH3)62+/3+ and Fe(CN)63-/4-, while they are much slower than those of GC for Fe2+/3+, DA, and O2 because of the ultraflat surface and low surface O/C ratios of the films. This results from the weak interaction between the film surface and the redox systems. The film exhibits excellent advantages over a conventional GC electrode when used to detect analytes such as histamine, azide ions, and BPA, which have a high oxidation potential and whose oxidized products foul the electrode surface. In addition, the ECR sputtering method is particularly advantageous in terms of its short

production time and cost-effectiveness. Applications for measuring other biomolecules with higher molecular weights such as oligonucleotides at ECR-sputtered carbon films are being studied. Besides electroanalytical applications, the films have the potential to be used to fabricate nanometer-sized electrodes for molecular electronics and for use in scanning probe microscope studies because of the ease with which they can be mass-produced.

SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review June 9, 2006. Accepted October 18, 2006. AC0610558

Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

105