Anal. Chem. 1996, 68, 3338-3343
Fabrication, Characterization, and Potential Application of Carbon Fiber Cone Nanometer-Size Electrodes Xueji Zhang,†,‡ Wuming Zhang,† Xingyao Zhou,† and Bozˇidar Ogorevc*,‡
Department of Chemistry, Wuhan University, Wuhan 430072, China, and Analytical Chemistry Laboratory, National Institute of Chemistry, P.O. Box 3430, 1001 Ljubljana, Slovenia
A novel method has been developed for the fabrication of carbon fiber cone nanometer-size ultramicroelectrodes (nanoelectrodes) with overall tip dimensions as small as 50 nm in diameter. In this method, carbon fibers were initially etched by an argon ion beam thinner. Afterward, a single etched carbon fiber was inserted into a glass capillary, which was then sealed by heating the glass/fiber interface in a vacuum; thus, no epoxy resin is involved. The success rate of our fabrication route for the electrodes with overall tip diameters of up to 500 nm was about 80%; for those with tip diameters of up to 100 nm, it was about 50%. The fabricated carbon fiber cone nanoelectrodes (CFCNEs) were inspected by optical and scanning electron microscopy. Their electrochemical behavior was examined by cyclic and linear sweep voltammetric measurements of ferricyanide and ferrocene ions in aqueous and nonaqueous media. The potential analytical applicability of the CFCNEs was tested by differential pulse voltammetric measurements of two well-known neurotransmitters, dopamine (DA) and 5-hydroxytryptamine (5-HT), and the results achieved were highly satisfactory. The calibration plots obtained were linear over the ranges from 5.0 × 10-7 to 1.0 × 10-4 and from 2.0 × 10-6 to 1.0 × 10-4 mol/L, with limits of detection of 1.0 × 10-7 and 5.0 × 10-7 mol/L for DA and 5-HT, respectively. Some advantages and improvements of the proposed CFCNE fabrication method, especially with respect to smoothness of the fiber (electrode) surface, strength, and control over the fiber tip dimensions, are also discussed. Several attractive properties associated with ultramicroelectrodes (UMEs) have stimulated many research efforts that are well reflected in a growing number of research and review reports on UMEs.1-4 The main beneficial features of UMEs include ultrasmall size with consequent short diffusion time scale and low double-layer capacitance, which results in high spatial and temporal resolution. Moreover, steady- or quasi-steady-state currents can be obtained with these electrodes. Due to the low iR drop, high resistive media and a simple two-electrode config†
Wuhan University. National Institute of Chemistry. (1) Wightman, R. M.; Wipf, D. O. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1989; Vol. 15, pp 267-353. (2) Montenegro, M. I., Queiros, M. A., Daschbach, J. L., Eds. Microelectrodes: Theory and application; Kluwer Academic Press: Dordrecht, The Netherlands 1991. (3) Heinze, J. Angew. Chem. Int. Ed. Engl. 1993, 32, 1268-1288. (4) Bond, A. M. Analyst 1994, 119, R1-R21. ‡
3338 Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
uration can be applied. Furthermore, the very small dimensions of UMEs allow small volume samples to be analyzed and, for example, facilitate fast cyclic voltammetry measurements for studying fast electrochemical processes. Adams5 and Wightman6 initially exploited microelectrodes with diameters of about 10 µm for in vivo electrochemical measurements of brain neurotransmitters. Regarding the shape, cylindershaped UMEs were shown to be particularly advantageous with in vivo electrochemical detection, for they could be inserted into live tissues and be maintained at specific target locations without damaging the tissues.7,8 Due to the needs and challenges for the miniaturization of detection elements and devices in clinical and microanalytical chemistry,9,10 a growing interest in investigations on smaller and smaller electrodes has led to the development of nanometer-size UMEs (nanoelectrodes) and exploration of their exciting properties. These electrodes are of particular interest in various ex vivo and in vivo clinical applications, including monitoring of local chemical events in special regions of organs, and in analyses of samples of extremely small volumes, especially for detection in single cells. So far, only a few groups have reported successful attempts in fabrication of nanoelectrodes.11-15 Lewis et al.11 described the preparation of nanometer-size platinum disk electrodes and introduced the term “nanode”. In our earlier work,12 a method for fabrication of ultramicroband electrodes as wide as 5-100 nm was reported. Strein and Ewing14,15 pioneered the fabrication of carbon fiber nanoelectrodes by employing the flame etching route. Very recently, a submicrometer-size carbon disk electrode was made by pyrolyzing methane gas in quartz capillaries.16 Nevertheless, most of the reported UMEs, especially those with band- and disk-shaped working surfaces, were housed in (5) Adams, R. N. Anal. Chem. 1976, 48, 1126A-1138A. (6) Wightman, R. M. Anal. Chem. 1981, 53, 1125A-1134A. (7) Beiley, F.; Malinski, T.; Kiechle, F. Anal. Chem. 1991, 63, 395-398. (8) Malinski, T.; Taha, Z. Nature 1992, 358, 676-678. (9) Bindra, D. S.; Zhang, Y.; Wilson, G. S.; Sternberg, R.; Thevenot, D. R.; Moatti, D.; Reach, G. Anal. Chem. 1991, 63, 1692-1696. (10) Cronenberg, C.; van Groen, B.; de Beer, D.; van den Heuvel, H. Anal. Chim. Acta 1991, 242, 275-287. (11) Penner, R. M.; Heben, M. J.; Longin, T. L.; Lewis, N. S. Science 1990, 250, 1118-1121. (12) Zhang, X.; Zhang, W.; Zhou, X.; Wang, Z. Gaodeng Xuexiao Huaxue Xuebao (Chem. J. Chin. Univ.) 1993, 14, 927-930 (in Chinese); Chem. Abstr. 1994, 120, 94110g. (13) Zhang, X.; Zhou, X. Chin. Chem. Lett. 1993, 4, 995-998; Chem. Abstr. 1994, 120, 318623g. (14) Strein, T. G.; Ewing, A. G. Anal. Chem. 1992, 64, 1368-1373. (15) Strein, T. G.; Ewing, A. G. Anal. Chem. 1993, 65, 1203-1209. (16) Wong, D. K. Y.; Xu, L. Y. F. Anal. Chem. 1995, 67, 4086-4090. S0003-2700(96)00096-0 CCC: $12.00
© 1996 American Chemical Society
relatively large insulating and supporting electrode bodies. As a result, the overall physical tip dimensions were in the millimeter range and above. Such UMEs are therefore appropriate mainly for kinetic17 and electrochemical microscopy18,19 studies. However, they cannot be utilized as probes for electrochemical measurements at micrometer-size locations and in single cells. In this frontier area of UME research, special attention was directed to the introduction of the flame etching of carbon fibers. Several valuable papers reported this new and simple method for the preparation of ultrasmall carbon fiber cylinder7,8 and disk14,15 UMEs. Typically, the overall tip dimensions reported were in the micrometer range, but some even reached dimensions as small as 400 nm.14 However, contact with flame and the high temperatures involved in the flame etching procedure cause some changes in the characteristics of the fiber material. This may result in, e.g., poor mechanical strength of the fibers, rough surface, and low control over a designated tip size, which sometimes limits the utiity of the UMEs produced via the flame etching route. Similar difficulties were mentioned to require longer electrochemical conditioning of the carbon fiber UMEs.20 In this paper, the fabrication, characterization, and possible application of carbon fiber cone nanoelectrodes (CFCNEs), with overall tip dimensions ranging from 50 to 500 nm in diameter, are presented. For the preparation of these electrodes, the employment of argon ion beam etching is fully reported for the first time. Besides the basic electrochemical behavior and properties of the developed CFCNEs, their potential utility is demonstrated by voltammetric measurements of two neurotransmitters, dopamine and 5-hydroxytryptamine. EXPERIMENTAL SECTION Apparatus. Cyclic (CV), linear sweep (LSV), and differential pulse (DPV) voltammetric measurements were performed employing a Model 366 bipotentiostat (EG&G PAR, Princeton, NJ), a CMBP-1 bipotentiostat (Jiangsu Electroanalytical Instruments, Taixian, China), and a Model 273 potentiostat-galvanostat (PAR) interfaced with a PC and Model 270 electrochemical analysis software (PAR). A Model 485 picometer (Keithley, Cleveland, OH) and a 901-PA analyzer (Ningde Instruments, Ningde, China) were used to amplify the current signals, when necessary. A three-electrode configuration, consisting of a carbon fiber cone nanoelectrode (CFCNE), a platinum wire, and a saturated calomel electrode (SCE) as the working, counter, and reference electrodes, respectively, was employed for all voltammetric measurements. The CFCNE was dipped in a conventional fashion, so that the whole carbon fiber cone tip was exposed to the measurement solution. Unless indicated otherwise, all potentials given in this paper are referred to the SCE. All electrochemical experiments were carried out in a 20 mL glass cell mounted in a laboratory-made copper mesh Faraday cage at room temperature (24 ( 2 °C). Scanning electron microscopy (SEM) was performed employing X650 (Hitachi, Tokyo, Japan) and JSM-T220 (JEOL Ltd., Tokyo, Japan) scanning electron microscopes. Carbon fibers were etched on a WW-3 ion beam thinner (Wuhan University Work(17) Montenegro, M. I. Res. Chem. Kinet. 1994, 2, 1-80. (18) Bard, A. J.; Fan, F. R. F.; Pierce, D. T.; Unwin, P. R.; Wipf, D. O.; Zhou, F. Science 1991, 254, 68-74. (19) Mirkin, M. V.; Fan, F. R. F.; Bard, A. J. J. Electroanal. Chem. 1992, 328, 47-62. (20) Golas, J.; Osteryoung, J. Anal. Chim. Acta 1986, 181, 211-218.
Figure 1. Schematic drawing of a carbon fiber cone nanoelectrode: E, epoxy resin; G, glass capillary tube; W, copper wire; S, copper fiber joint with silver paint; C, carbon fiber; T, sharpened fiber tip; F, glass/fiber interface.
shop, Wuhan, China). For micro-scale operations in CFCNE fabrication, a 17 XA inverted microscope (Shanghai Optical Instruments, Shanghai, China) and an Olympus micromanipulator (Narishige, Japan) were employed. Reagents and Standard Solutions. All chemicals were of analytical reagent grade purity and were used as received. All solutions were prepared with redistilled water. Prior to electrochemical measurements, the solutions were deaerated for 10 min with pure nitrogen. Dopamine (DA) and 5-hydroxytryptamine (5HT) stock solutions were prepared by dissolving the required amounts of DA and 5-HT (both received from Sigma, St. Louis, MO) in 0.2 mol/L phosphate buffer solution (pH ) 7.0). Working standard solutions were prepared by diluting the stock solutions with 0.2 mol/L phosphate buffer solution prior to use. Ferrocene was synthesized and purified in the laboratory. Fabrication of the Carbon Fiber Cone Nanoelectrodes. (A) Carbon Fiber Etching. Carbon fibers (7 µm in diameter and about 15 mm in length, Goodfellow Co., Oxford, UK) were initially cleaned by sonication for 5 min first in acetone then in 1:1 nitric acid and finally in redistilled water. Afterward, they were allowed to dry in air at room temperature. The cleaned and dried fibers were positioned in the center of the target of the ion beam thinner and etched with an argon ion beam. Needle-type carbon fibers with sharpened tips, ranging in diameter from about 50 nm to several hundred nanometers, were obtained by controlling the experimental conditions and settings (acceleration voltage, 10004000 V; beam current, 1.5-5.0 mA; incident angle, 5-15°; and etching time, 8-15 h). More details on the carbon fiber etching procedure are given elsewhere.21 (B) CFCNE Construction. Glass capillary tubes were pulled with an Olympus PE-2 pipet puller to form very thin capillaries with inner diameters of 2-10 µm. A single etched carbon fiber was mounted on the end of a copper wire (0.5 mm in diameter) and fixed by means of silver conducting paint. The carbon fiber was then inserted into a pulled glass capillary under a microscope. The carbon fiber was carefully made to pass through the capillary so that ∼10-100 µm of it was left protruding. After this, the copper wire was fixed at the opposite end of the glass capillary tube with an epoxy glue (Dongfen Ad-U, Guangzhou, China). Finally, the fiber was sealed into the capillary mouth by heating the glass/fiber interface for about 2 s inside a small heating coil under vacuum. A CFCNE constructed in this way is schematically presented in Figure 1. RESULTS AND DISCUSSION Carbon Fiber Cone Nanoelectrode Fabrication. The results of argon ion beam etching of carbon fibers and CFCNE fabrication are illustrated in Figures 2 and 3 with a series of SEM (21) Zhang, X.; Zhou, X.; Zhang, W. Chinese Patent 94104755.5, 1994.
Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
3339
Figure 2. Scanning electron microscopy pictures of parent, ion beam-etched, and flame-etched carbon fibers. (A) A single parent (unetched) carbon fiber, together with five ion beam-etched fibers, at the same magnification scale. (B) A conically ion beam-etched carbon fiber. (C) A flame-etched carbon fiber.
Figure 4. Optical microscopy picture of a carbon fiber cone nanoelectrode. The scale bar represents 10 µm. Figure 3. Scanning electron microscopy pictures of ion beametched carbon fiber cone tips of the already-fabricated CFCNEs: fiber tips with tip diameters of about 300 (A) and 100 nm (B) and glass/ fiber interface (C). No gold sputtering was applied in A and C.
pictures. To enable a direct comparison between the untreated and ion beam-treated fibers, picture A in Figure 2 shows a SEM image of a single parent, i.e., unetched, carbon fiber, together with five etched fibers, at the same magnification. Picture B in Figure 2 shows the image of a (sharpened) part of an ion beametched carbon fiber, with dimensions of about 100 µm in length and several hundred nanometers in tip diameter. Evidently, the ion beam etching process results in cone- or needle-shaped carbon fibers. For reference, in picture C of Figure 2, a flame-etched fiber is displayed, produced in our laboratory by following exactly the established procedure.14 Its surface is obviously strongly affected by flame burning off under poorly controllable conditions. Pictures A and B in Figure 3 display SEM images of ion beametched carbon fiber tips with dimensions of about 300 and 100 nm in diameter, respectively, of the already-fabricated CFCNEs. The SEM images of these tips clearly indicate that the fiber surface 3340 Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
after ion beam etching is remarkably smooth, which is a very important factor, particularly in low-current electrochemical measurements. Picture C in Figure 3 illustrates the glass/fiber interface, after the fabrication of a CFCNE has been finalized by sealing a fiber into a pulled glass capillary at sufficiently high temperature in vacuum. It reveals that the sealing was accomplished successfully, which is further evident from the optical microscopy picture of the sealing region, presented in Figure 4. It should be stressed that this direct heat sealing is accomplished in a very short time period (about 2 s) at a designated location (performed inside a small heating coil) and under vacuum (no oxygen present). Hence, the mechanical and surface characteristics of a fiber cone are negligibly affected by this sealing process. The mechanical strength of the carbon fibers etched via the proposed ion beam procedure was compared to that of the carbon fibers etched via the flame procedure (performed in our laboratory). Twenty-four pairs of CFCNEs with both types of fiber tips of equal tip dimensions, but all below 200 nm in diameter, were tested by introducing the tips into membranes of different plant
or animal single cells, which was performed with the optical microscope and micromanipulator under apparently identical conditions. The test results showed that about 80% of the flameetched tips were broken, while only about 10% of the ion beametched tips were damaged. It may be concluded, therefore, that the mechanical strength of the carbon fibers processed by ion beam etching is much greater than that of the carbon fibers etched by flame. Apparently, this characteristic of the CFCNEs produced by ion beam etching makes them very useful, in particular for the tissue and single-cell electrochemical measurements where mechanical strength, in addition to the small size, should play a key role. After fabrication and experimentation with more than 400 CFCNEs, the success rate of their production was calculated. Starting from the parent carbon fibers, the success rate of ion beam etching for the fibers with tip diameters of up to 500 nm, when optimum settings and experimental conditions are provided, is nearly 99%. The success rate of the complete fabrication procedure for the CFCNEs with tip diameters of up to ∼100 nm is about 50%, and that for the CFCNEs with tip diameters of up to ∼500 nm is about 80%. All the above results confirm that the CFCNE fabrication method developed in this laboratory is appropriate and successful. The proposed argon ion beam etching of carbon fibers is believed to be an excellent tool in fabricating the nanometer-size electrodes with assessable surface characteristics. By carefully varying the parameters (acceleration voltage, beam current density, and beam incident angle), from high settings in the beginning to low settings toward the end of the etching procedure, it is possible to produce conically and uniformly shaped carbon fiber tips with a physically and electrochemically smooth surface. While the ion beam treatment enables thinning of carbon fibers to the designated shape and dimension, it apparently does not exert any influence on the known electrochemical characteristics of the carbon fiber surface. Voltammetric Behavior of the Carbon Fiber Cone Nanoelectrodes. Each CFCNE produced was initially tested by cyclic voltammetry measurements of 1.0 × 10-3 mol/L potassium ferricyanide in 0.2 mol/L phosphate buffer solution of pH ) 7. Such cyclic voltammograms, recorded at different potential scan rates, are displayed in Figure 5. Sigmoidal-shaped voltammograms were observed, indicating the nonlinear diffusion behavior at CFCNEs. Furthermore, from Figure 5, it is obvious that the limiting current obtained at a CFCNE changed only little on varying the potential sweep rate over the range from 10 to 500 mV/s, implying a high rate of diffusion flux at the CFCNE. In addition, the voltammograms in Figure 5, exhibiting relatively low background discharge currents, proved our estimation based on observations of the SEM pictures, that the surfaces of the CFCNEs, owing to the ion beam etching, are smooth. Furthermore, our CV measurements with freshly made CFCNEs showed that neither chemical nor any electrochemical pretreatment or conditioning was needed prior to their first use. One of the important areas of the application of ultramicroelectrodes is for studies and measurements in resistive media. Here, the use of epoxy-sealed microelectrodes might often be restricted to aqueous solutions. This prompted us to examine the performance of our CFCNEs in nonaqueous media, since no epoxy sealing was involved with these nanoelectrodes. Figure 6 shows the LSV responses for ferrocene in an organic solvent,
Figure 5. Cyclic voltammograms of 1.0 × 10-3 mol/L potassium ferricyanide in 0.2 mol/L phosphate buffer solution (pH ) 7.0) obtained with a carbon fiber cone nanoelectrode at different scan rates: (a) 10, (b) 50, (c) 100, (d) 200, (e) 300, and (f) 500 mV/s. Initial and final potential, -0.1 V; vertex potential, 0.5 V; CFCNE tip diameter, 300 nm.
Figure 6. Linear sweep voltammograms of 1.0 × 10-3 mol/L ferrocene in acetonitrile obtained with a carbon fiber cone nanoelectrode at different concentrations of supporting electrolyte (NaClO4): (a) 0, (b) 2.0 × 10-3, and (c) 5.0 × 10-2 mol/L. Scan rate, 10 mV/s; initial potential, -0.1 V; final potential, 0.6 V; CFCNE tip diameter, 100 nm.
acetonitrile, in the absence (a) and in the presence of low (b) and high (c) concentrations of a supporting electrolyte (NaClO4). Apparently, even in the absence of supporting electrolyte, a welldefined voltammogram was obtained, clearly indicating that there was practically no effect of the ohmic resistance of the measurement solution. A similar steady-state current response, with only a small positive shift in the E1/2 value, can be observed when comparing the CFCNE behavior in the absence of electrolyte (curve a) to that at higher (curves b and c) concentrations of the added supporting electrolyte. The small ohmic losses, coupled with relatively high analytical sensitivity, seem promising for the application of the CFCNEs in resistive media, e.g., for detection in liquid chromatography and flow injection analysis, as virtually no addition of a supporting electrolyte would be required. From the analytical standpoint, the ultramicroelectrodes should also exhibit a well-defined concentration dependence. The Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
3341
Figure 7. LSV calibration plots of potassium ferricyanide obtained at carbon fiber cone nanoelectrodes with different tip diameters: (a) 50, (b) 100, and (c) 300 nm. Supporting electrolyte solution, 0.2 mol/L phosphate buffer (pH ) 7.0); other conditions as in Figure 5.
calibration plots created from the voltammetric data for the series of concentrations of potassium ferricyanide, obtained at three CFCNEs, differing in tip dimensions, are presented in Figure 7. A good linear proportionality between the steady-state currents and the ferricyanide concentration was observed for all nanoelectrodes examined. Least-squares treatment of these data yielded slopes of 10.5, 15.0, and 24.4 nA/mmol for the electrodes with tip dimensions of 50, 100, and 300 nm in diameter, respectively. In fact, the ratio of sensitivities does not follow the ratio of the tip dimensions, nor is the current response simply related to the geometric surface area of the corresponding CFCNEs. However, this discussion is not within the focus of this paper. Measurements of Dopamine and 5-Hydroxytryptamine at the CFCNEs. To understand the actual role of dopamine (DA) and 5-hydroxytryptamine (5-HT) as neurotransmitters, methods capable of directly monitoring the concentrations of DA and 5-HT in neurons would be extremely useful. According to the sufficiently small dimensions of the CFCNEs developed in our laboratory, as well as their physical and electrochemical characteristics, they could be considered appropriate for measuring DA and 5-HT in vivo. Hence, the CFCNEs were investigated for their responses and analytical performance for both neurotransmitters. As can be seen from Figure 8, well-defined cyclic voltammograms of DA and 5-HT were recorded utilizing the CFCNEs. The voltammograms correspond to the oxidation of DA and 5-HT, respectively, and show the expected sigmoidal behavior. The behavior of DA presented in Figure 8A is consistent with previously reported data obtained at the carbon fiber cylinder microelectrodes,22 proving that the voltammetric integrity of the carbon fibers is retained after etching by an ion beam thinner. When successive potential scans were applied in 5-HT solution, a small decrease in the steady-state oxidation current was noted, which corresponds to the observations reported elsewhere.23 In further examinations of the analytical applicability of the developed carbon fiber cone nanoelectrodes, the voltammetric measurements of low concentrations of DA and 5-HT were carried out with DPV operation at the CFCNEs with tip dimensions of (22) Ewing, A. G.; Dayton, M. A.; Wightman, R. M. Anal. Chem. 1981, 53, 18421847. (23) Lau, Y. Y.; Wong, D. K. Y.; Luo, G.; Ewing, A. G. Electroanalysis 1992, 4, 865-869.
3342 Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
Figure 8. (A) Cyclic voltammogram of 1.0 × 10-3 mol/L dopamine in 0.2 mol/L phosphate buffer solution (pH ) 7.0) at a carbon fiber cone nanoelectrode with tip diameter of 100 nm. Scan rate, 50 mV/ s; initial and final potential, -0.1 V; vertex potential, 0.5 V. (B) Cyclic voltammogram of 1.0 × 10-3 mol/L 5-hydroxytryptamine; all other conditions as in (A).
Figure 9. Differential pulse voltammograms of dopamine in 0.2 mol/L phosphate buffer solution (pH ) 7.0) at a carbon fiber cone nanoelectrode with tip diameter of 300 nm. (a) Baseline. (b) DA concentration, 5.0 × 10-6 mol/L. (c) DA concentration, 1.0 × 10-5 mol/L. Scan rate, 20 mV/s; pulse repetition time, 0.3 s; pulse time, 50 ms; pulse height, 20 mV.
100 nm in diameter. The differential pulse voltammograms of DA, including the baseline of the blank supporting electrolyte solution, are shown in Figure 9. Using the DPV data, the calibration plot obtained for DA was linear over the range from 5.0 × 10-7 to 1.0 × 10-4 mol/L, having a slope of 78.5 pA/µM, a y-axis intercept of 2 pA, and a linear correlation coefficient of 0.995. The relative standard deviations at the DA concentrations of 5.0 × 10-6 and 1.0 × 10-4 mol/L (n ) 8) were found to be 2.6% and 1.4%, respectively, and the limit of detection, calculated to S/N ) 3, was 1.0 × 10-7 mol/L. Using again the corresponding DPV data, the calibration plot obtained for 5-HT was linear over the range from 2.0 × 10-6 to 1.0 × 10-4 mol/L, exhibiting a slope of 75.4 pA/µM, a y-axis intercept of 3.6 pA, and a linear correlation coefficient of 0.991. The limit of detection found was 5.0 × 10-7 mol/L. Significantly, these calibration ranges correspond well with the estimated concentration levels of these and related substances in mammalian cells.24 CONCLUSIONS A novel method for the fabrication of carbon fiber cone nanometer-size ultramicroelectrodes (CFCNEs), employing the (24) Cooper, B. R.; Jankowski, J. A.; Leszczyszyn, D. J.; Wightman, R. M.; Jorgenson, J. W. Anal. Chem. 1992, 64, 691-694.
ion beam etching of carbon fibers and direct heat sealing of a needle-shaped carbon fiber into a pulled glass capillary, is reported. The new fabrication method offers several significant advantages: (i) the argon ion beam treatment allows carbon fibers to be etched to designated tip dimensions, as small as 50 nm in diameter or even less; (ii) the method is suitable for the production of CFCNEs in large quantities and is adequately rapid; (iii) the main procedure of the CFCNE fabrication, the ion beam etching, can be carried out with no special skill being required; and (iv) the method proposed is convenient for performing CFCNE fabrication with a high success rate, e.g., the success rate of fabrication of CFCNEs with overall tip dimensions of up to 500 nm is about 80%. The carbon fiber cone nanoelectrodes presented in this work for the first time exhibited some interesting improvements. Due to direct heat sealing, i.e., with no epoxy sealing involved, these electrodes can be employed in practically any aqueous and nonaqueous media. The CFCNEs produced by the proposed method are readily applicable with no chemical or electrochemical pretreatment or conditioning required. The developed CFCNEs include some attractive and promising characteristics. Due to low ohmic losses, they are suitable for detection in resistive nonaqueous media, with no need of the addition of a supporting electrolyte, e.g., for use in liquid chromatography and other continuous-flow systems. The experimental evidence of the dopamine and 5-hydroxytryptamine
measurements indicates the CFCNEs can be considered useful for in vivo and ex vivo detection of neurotransmitters. Owing to the nanometer tip dimensions of the CFCNEs, their smooth surface, and their high mechanical strength, these electrodes seem advantageous for electrochemical measurements in tissue ultramicro-scale environments and in single cells. Nevertheless, considering a possible modification of the CFCNEs, e.g., by coating the surface with organic (polymer) or inorganic films, and some further improvements in the fabrication procedure, their potential area of applications is still to be thoroughly investigated. Research activities in some of these directions are in progress in our laboratories. ACKNOWLEDGMENT This project was supported by the National Natural Science Foundation of China and partially supported by the Ministry of Science and Technology of the Republic of Slovenia, which are gratefully acknowledged. We thank Mr. Hua Tong for his assistance in obtaining the SEM pictures and Dr. Wenghui Zou for her assistance in ion beam etching of carbon fibers. Received for review January 31, 1996. Accepted July 5, 1996.X AC9600969 X
Abstract published in Advance ACS Abstracts, August 15, 1996.
Analytical Chemistry, Vol. 68, No. 19, October 1, 1996
3343
