J. CHEM. SOC. FARADAY TRANS., 1994, 90(4), 605-608 Dual-cylinder Microelectrodes Part 2.t-Steady-state Generator and Collector Electrode Currents B. J. Seddon,*$ Chang Fa Wang, Wenfeng Peng§ and Xueji Zhang Department of Chemistry, Wuhan University, Wuhan 430072,People’s Republic of China The feedback and collection diffusion current properties of platinum dual-cylinder microelectrodes have been investigated using bipotentiostatic chronoamperometry. The study focuses on the influence of the interelectrode spacing on the steady-state response at independently controlled generator and collector microcylinders. A series of such devices were fabricated onto polyethylene terephthalate (PET) strips using a microscope manipu- lation method. Micro-manipulation techniques allowed dual-cylinder devices to be constructed using 25 pm platinum wire with gap spacings from 3.5 to 188 pm and cylinder lengths 250-500 pm. Recycling efficiencies, expressed as an iJiQ ratio, observed with these feedback dual-microelectrodes could be as high as 83% for closely spaced microcylinders. Electrochemical measurements further indicate that a convergence of the gener- ator and collector currents would follow as the interelectrode distance diminishes below the pm scale where the strength of the recycling flux field at the device is high enough to minimise bulk solution exchange.The response of the new amperometric device is discussed with reference to the relevant theoretical stationary flux model proposed for dual-cylinder microelectrodes. Amperometric dual microelectrodes function according to interelectrode diffusion processes which bring about a rapid and amplified steady-state current response at two closely positioned microelectrodes.Microelectrode devices such as these can be operated in both feedback and collection modes by independent control and measurement of individ- ual electrode potentials and currents. In collection experi- ments, where a collector is potentiostatted at a specified constant potential, the electro-generated species diffuse out from the generator electrode, traverse a microscopic gap and then undergo further redox reactions at a neighbouring col- lector electrode. The measured collection currents may then be utilised to bring improvements to analytical selectivity by monitoring the generated form of a redox analyte, hence overcoming background effects suffered by the generator elec- trode and indeed complicating homogeneous reactions.As a feedback device, however, it is necessary to sustain a high cyclic flux of some electroactive redox couple in the electro- lyte region between the microscopic anode and cathode. In both cases, feedback and collection micro-devices provide a constant current output in techniques such as pulse ampero- metry. Amperometric signals are measured on the nA scale and possess exceptionally low collector background currents which allows sensitive measurements on sub-micromolar analyte concentrations. Further, the onset of a steady-state current, i.e.response time, for these devices is of the order of a few seconds or less, with amperometric amplification over that of single microelectrode devices. Diffusion effects at dual microelectrodes provide us with new possibilities for the development of electroanalytical and sensor techniques over the conventional amperometric microelectrodes whose response is governed primarily by bulk solution mass trans- Dual microelectrodes like the interdigitated arrays (IDA), parallel dual cylinders (PDC) and feedback disc have been cited in a number of analytical problems over recent years, from end-point detectors in amperometric titrimetry,’ t Part 1: Ref. 14. Present address: Laboratoire d’Electrochimie, Institut de Chirnie Physique 3, Ecole Polytechnique Federale de Lausanne, CH-1015 Lausanne, Switzerland.0 Present address: Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sci-ences, People’s Republic of China. to homogeneous immunoassay,6 collection sensing of cate- cholamines and stripping ~oltammetry.~ The experimental variables of importance in feedback and collection electrode and sensor studies are associated with the absolute and relative values of steady-state generator and col- lector currents. Collection current and efficiency terms are most frequently applied. Consequently, much of the charac- terisation work on multiple-band microelectrodes has relied on terminology from earlier diffusion-onvection experi-ment~.”~’~In this case the collection efficiency is the ratio of collector to generator currents. This term suitably marks the ability of a dual-electrode device to detect electro-generated species.The quantity recovered at the collector over trans- port losses is dependent on chemical and hydrodynamic con- ditions and electrode dimensions. The definition is quite satisfactory for rotating ringdisc experiments and dual-electrode flow cells since the generator electrode process remains independent of the collector electrode reaction owing to the macroscopic size of the electrodes and gap, as well as the directionality of solution flow. The description is also suitable for microelectrode systems where no feedback effects are concurrent.Alternatively, in instances where inter-electrode feedback diffusion is present, the generator current becomes modulated by a reverse transport process, i.e. the feedback flux. The observed collector current is then not in response to an independent generator process and hence under such conditions it is inappropriate to use the term col- lection efficiency in the former sense. For microelectrode feed- back devices generally, the ratio of collector current to generator current, as given in eqn. (I), is a measure of the diffusion recycling flux between generator and collector elec- trodes over any reaction or bulk exchange fluxes. Bearing this in mind, we have assigned the collector current: generator current ratio in these measurements at dual-cylinder micro- electrodes as a recycling efficiency which is sensitive to the physico-chemical conditions concerning the device and elec- trolyte solution.N = iJig Dual-cylinder microelectrodes were recently developed as an alternative geometry to laminate dual-band microelec- trodes,’0-’2 and as a more accessible, inexpensive and robust electrode system relative to the microlithographic IDAs.’ A dual-cylinder device is composed of a parallel arrangement of two cylindrical electrodes with microscopic dimensions in the cylinder diameter and the electrode spacing (for analytical applications <25 pm). By contrast, the electrode length is an order of magnitude or more larger; generally >250 pm but typically < 1 mm.The microelectrode devices examined in the present study are described as geometrically symmetrical in terms of cylinder diameter and spatial arrangement as shown in Fig. 1.14 Dual-cylinder microelectrodes possess certain features important to the design of feedback and col- lection experiments. The microscope fabrication scheme offers an extremely versatile method for the incorporation of multiple cylinders with a wide range of electrode and fibre materials, such as carbon fibre, metal micro-wires, chemically modified wires and coated polymer fibres.I3 Moreover, the well defined microcylinder geometry allows for a detailed characterisation of electrochemical systems by semi-analytical and numerical approaches to mass transport and coupled reaction flux problems.._ L d Fe(CN):-Fe(CN),& Fig. 1 (a) Illustration of the dual-cylinder microelectrode showing the device dimensions: cylinder length L, radius T and the gap spacing w, where w is the shortest separation distance between a pair of parallel cylindrical electrodes and is given by 2(d -r). (b) Electro-chemical recycling scheme for the hexacyanoferrate(lII/II) couple at cylindrical generator and collector microelectrodes. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 A theoretical description of the diffusion current for a reversible redox couple at a dual-cylinder microelectrode was recently put forward in Part 1, where the effects of solution conductivity, convection and diffusion into bulk solution were neg1e~ted.l~ In this treatment it was assumed that a steady-state flux field is established between the cylindrical electrodes which culminated in a constant current of equal magnitude at the generator and collector, that is, no redox analyte losses into the bulk electrolyte.This current, which is dependent on the physical properties of the analyte system and the device dimensions, is re-stated in eqn. (2) below : i = nnFDCbL/cosh-'[(w/2r) + 13 (2) where i is the steady-state current delivered at the elec- trodes, with Cband D,the bulk solution concentration and diffusion coefficient of the electroactive species. The number of electrons involved in the electrode reaction is n, with L,r and w being the device dimensions as described in Fig.1. In this work an experimental study of the steady-state currents at the generator and collector of dual-cylinder devices is pre- sented for the case of Fe(CN);-/Fe(CN):-recycling in aqueous KCl electrolyte. The observations are discussed in the context of the theory for diffusion currents at these novel feed back microelectrodes. Experimental Chemicals Chemical reagents used were of analytical standard. Pot- assium hexacyanoferrate(rr1) and potassium chloride were purchased from Shengyang and Chantou Chemical Factories, China. All solutions were prepared with doubly distilled deionised water. The analytical solutions were purged with dry nitrogen for 10 min before electrochemical measurements were made. Electrode fabrication materials included, 'Melinex' polyethylene terephthalate (PET) kindly donated by ICI Films Ltd, platinum wire purchased from Goodfel- lows, and epoxy-resin from RS Components.Dual-cylinder Microelectrodes Dual-cylinder microelectrodes were fabricated by a novel microscope manipulation technique involving a slit-mounting and epoxy-sealant procedure. The dual electrodes are sup- ported on a PET strip for convenience of handling and oper- ation. Details concerning the construction and electrochemical characterisation of dual-cylinder devices have been given elsewhere. 59 Device dimensions were determined by scanning electron microscopy or with an optical micro- scope fitted with micrometer verniers. It was found that the length of the micro-wire, in the size range 250-500 pm, could be controlled to better than 25 pm using the microscope fab- rication method.Furthermore, the technique allows low micrometre control over the gap dimension with estimations of the average gap size along the device length within &1 pm.' The dual-cylinder microelectrodes used in this work consist of a pair of platinum cylindrical electrodes of 25 pm diameter with a microscopic interelectrode spacing of vari- able size (<188 pm).Only devices showing regular geometric features in addition to displaying identical cyclic voltam- metric and chronoamperometric behaviour for the individual cylinder electrodes (i.e. variations in the quasi-steady-state currents ~3%)were used in collection and feedback experi- ments.Table 1 lists the gap dimensions for a series of dual-cylinder microelectrodes under consideration here. Amperometric Measurements Chronoamperometry experiments were carried out using a bipotentiostat, RDE4, Pine Instruments, USA. The generator J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 607 Table 1 Steady-state currents at the generator and collector elec- +-Itrodes for a series of dual-cyclinder devices with variable w 1.0 0.8 -/e3.5 520.0 430.5 45 1.6 15.1 4.7 !! 5.0 469.2 328.2 380.7 23.2 13.8 12.0 339.8 174.3 250.6 35.6 30.4 22.5 250.4 129.2 188.0 33.2 31.3 c35.0 233.7 90.7 156.0 49.8 41.9 , 79.5 194.4 41.1 111.1 75.0 63.0 ii181.5 212.4 32.7 85.1 149.6 61.6 N188.0 185.8 10.4 82.8 124.4 87.4 Experimental data include background subtraction and are normal-ised in cylinder length (L = 1.0 mm) and concentration (1.0 mmol dm-3).Theoretical currents have been calculated from eqn. (2) using experimental values for device dimensions, with D s-', Cb= 1.0 mmol dmP3 and n = 1 for the hexacyanoferrate(III/n) = 7.0 x 10-lo m2 0.2;+ anions in aqueous KCl solution. and collector current-time curves were recorded on an x-y-t recorder 3086 from Hokushin Electric, Japan. Bipotentiosta- tic chronoamperometry of K,Fe(CN), (1.0 mmol dm- 3, was performed in aqueous KCl (0.5 mol dm-3) solution by apply- ing a potential step of +0.4 to 0.0 V us. SCE at the generator electrode to produce Fe(CN):-. The collector electrode was controlled at a constant potential of +0.4 V in order to oxidise hexacyanoferrate(rr) back to hexacyanoferrate(rr1).Three i-t curves were recorded in succession for each experi- ment. The average steady-state current values sample at 10 s are presented in Table 1. The electrochemical cell housed a reference electrode (SCE) and a counter electrode (Pt wire 1 mm in diameter and 10 mm in length) with additional cap fittings to ease handling of the dual-cylinder strip devices. Furthermore, the cell allowed electrolyte deoxygenation, which is often essential for low-concentration experiments and had a maximum solution capacity of 20 ml. The tem- perature in the chronoamperometric experiments reported here was controlled to 26 f1 "C.Results and Discussion Chronoamperometric measurements for the platinum dual- cylinder microelectrodes using the hexacyanoferrate(rrr/Ir) couple are presented in Table 1, where the generator and col- lector steady-state currents are given as a function of the gap dimension. The diffusion currents observed at the generator electrode, i, , are noticeably larger than the corresponding collector electrode currents, i,, for all values of the gap dimension under review. This is an obvious consequence of the generator being supplied with hexacyanoferrate(Ir1) from the electrolyte solution while the collector response is depen- dent on the cross-diffusion of the generator species [hexa- cyanoferrate(rr)]. Furthermore, the experimental values for the generator and collector currents are enhanced and con- verge on diminution of this interelectrode distance, w.The i-w behaviour in the steady-state response is further empha- sised in the recycling efficiency curve shown in Fig. 2 where the recycling efficiency, N, is plotted against the spacing parameter, w, for a series of dual-cylinder microelectrodes. The recycling curve, in addition to illustrating the sensitivity of the iJi, ratio to the gap, indicates that the stationary field recycling model presented earlier,14 where N is unity for any value of w, is not valid over the device dimensions used in this study. The predictions outlined in the stationary-field model, where comparison was made with a microcylinder response, suggested that eqn.(2) might only be valid for a limited case. Nevertheless, the experimental data, which - 1 a ~ I 1 I I I 0.0 40.0 80.0 120.0 160.0 200.0 w/Pm Fig. 2 Recycling curve for a series of dual-cylinder microelectrodes. The plot shows the response of the recycling efficiency, i&, as a function of the micrometer gap spacing, w. demonstrate N values for dual cylinders ranging from 0.06 to 0.83, do show that as the interelectrode distance is reduced (w +0) the device's steady-state response tends to the theo- retical limit, where i, = i,. The i-w data listed in Table 1 also provide a comparison of experimental diffusion currents with their theoretical steady- state values. The result illustrates how well the analytical model predicts the behaviour of this feedback de~ice.~,'~ Strong deviations from eqn.(2) are found in the generator and collector currents for devices possessing large values of the gap spacing, i.e. >5 pm. The current at the generator is also seen to overestimate consistently the values given by the recycling model. However, such deviations become less pro- nounced as the interelectrode gap decreases to a few pm. In contrast, values for the collector current remain below that of the theoretical current since the amperometric response of this electrode is driven by the generator flux with its associ- ated bulk exchanges. For dual-cylinder devices with broad w spacings the collector currents assume low values, typically < 10 nA per mm of cylinder as w approaches 200 pm. For the smallest gap dimension fabricated using the micro-manipulation technique, ca.3.5 pm, the experimental data for i, and i, deviates from the expected values given in eqn. (2) by 15.1% and -4.7%, respectively. Larger deviations follow as the feedback and collection fluxes are minimised over bulk solution fluxes as the collector is placed further from the gen- erator electrode. It is informative to extrapolate this experi- mental data with respect to eqn. (2) in order to understand the performance limitations (e.g. analytical sensitivity, i/C) of dual-cylinder systems. Using i us. l/cosh-'(w) plots it is pos- sible to make estimations of the interelectrode distance neces- sary to approach theory.In this case the generator and collector currents are found to approximate the theoretical currents to better than 5% as the gap dimension diminishes below 2 pm. Furthermore, convergence of experimental cur- rents on the theoretical value is predicted for the present 25 pm cylinder devices as the gap is reduced to CQ. 0.5 pm. Realistically, however, it is unlikely that dual-cylinder devices with such a small dimension could ever be constructed by manual micromanipulation technology. Fig. 3 considers the experimental diffusion currents at dual- and single-cylinder microelectrodes. l4 The steady-state 608 1000.0 800.0 p 600.0 '=-2 t 3 400.0 200.0 .-\r, wc .. I 1, : lo i I I I I I I I I I 0.0 40.0 80.0 120.0 160.0 200.0 Fig.3 The behaviour of the steady-state currents at generator and collector electrodes with varying cylinder gap. The horizontal solid line gives the quasi-stady-state current at a single-cylinder microelec-trode sampled after 10 s pulse. response of the generator electrode is in fact a combination of a cylindrical flux originating from bulk solution and the col-lector feedback process. For small gap sizes (<5 pm), the generator current behaviour is strongly influenced by feed-back effects from the collector electrode which is essentially limiting the spatial extension of the generator diffusion field. Consequently this affords chronoamperometric curves dis-playing a fast onset for steady-state currents in addition to amplification of the diffusion current over that of a single-cylinder device.'0.'3.'4 The generator curve also conveys the effects of the collector electrode positioned at greater distance from the generator.Here, the current at the generator tends to a minimum value which corresponds to single-cylinder behaviour, i.e. i, -,is, as w extends beyond a critical feedback space, wg. At large interelectrode distances therefore the gen-erator response remains relatively unperturbed by electro-chemical processes at the collector electrode. By comparison with the current at a single-cylinder electrode, this distance, wg, is of the order of 80 pm and marks the diffusional dis-tance over which the feedback effect operates for dual-cylinder microelectrodes of this dimension.The collector curve featured in Fig. 3 displays the typical analyte collection properties of dual-cylinder microelectrodes. The current at the collector tends to a background value, in this instance, i, = 0.85 nA, as the interelectrode spacing exceeds ca. 110 pm. Hence, the collector curve defines the maximum collection or sensing distance, w, , beyond which generated species can perturb the collector electrode current from a background value. In these experiments with hexacyanoferrate(rI1) in aqueous KCl solution, the Fe(CN):-species can be detected J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 at the collector over distances > 180 pm from the generator. Further to this, the collector current approaches the theoreti-cal limit with decreasing values of the gap distance.Conclusion Dual-cylinder microelectrodes have been fabricated by micro-scope manipulation procedures to tolerances acceptable for electrochemical characterisation. The amperometric response of this novel feedback device was assessed in terms of the steady-state currents at individual generator and collector microelectrodes. The theoretical treatment previously re-ported set out to explain the diffusion current at dual cylin-ders in terms of a total recycling model. However it was shown here that the model only approximates device per-formance where the interelectrode gap dimension is small relative to the microcylinder diameter. This would be a direct consequence of a high recycling flux between the cylinder electrodes overcoming the effects of bulk solution concentra-tion gradients.The authors acknowledge the co-operation and financial con-tribution of the National Science Foundation of China to the dual-microelectrode projects. B.J.S. wishes to express his gratitude to the students, technical staff and colleagues at Wuhan University for making the visit to China in 1990 rewarding and memorable. References 1 M. S. Harrington and L. B. Anderson, Anal. Chem., 1990, 62, 546. 2 B. J. Seddon, H. H. Girault and M. J. Eddowes, J. Electroanal. Chem., 1989,266,227. 3 K. Aoki, M. Morita, 0. Niwa and H. Tabei, J. Electroanal. Chem., 1988,256,269. 4 H. Tabei, T. Horiuchi, 0. Niwa and M. Morita, J. Electroanal. Chem., 1992,326,339. 5 W. Peng, Z. Zhou and P. Li, J. Electroanal. Chem., 1993,347, 1. 6 W. Peng and B. J. Seddon, Proceedings of the 4th. National Con-ference of Electroanalytical Chemistry, Shanghai, 1990. 7 T. Horiuchi, 0. Niwa, M. Morita and H. Tabei, Anal. Chem., 1992,64,3206. 8 V. T. Shea and A. J. Bard, Anal. Chem., 1987,59,2101. 9 0.Niwa, M. Morita and H. Tabei, Anal. Chem., 1990,62,447. 10 B. J. Seddon, Ph.D. Thesis, Edinburgh University, 1989. 11 W. Peng, Z. Zhou, X. Zhang, B. J. Seddon and Z. Zhao, Fenxi Huaxue, 1992,20,844. 12 Z. Zhou, W. Peng, X. Zhang and B. J. Seddon, Proceedings of the International Beijing Conference and Exhibition on Instrumen-tal Analysis, Science Press, Beijing, 1991, p. F63. 13 C. Wang, B. J. Seddon, W. Peng and X. Zhang, Electrochim. Acta, to be submitted. 14 B. J. Seddon, H. H. Girault, M. J. Eddowes, W. Peng and Z. Zhao, J. Chem. SOC.,Faraday Trans., 1991,87,2603. Paper 3/06341K; Received 25th October, 1993