Comparison of Voltammetric Data Obtained for the trans - American

Anal. Chem. 2003, 75, 6938-6948

Comparison of Voltammetric Data Obtained for the trans-[Mn(CN)(CO)2{P(OPh)3}(Ph2PCH2PPh2)]0/+ Process in BMIM‚PF6 Ionic Liquid under Microchemical and Conventional Conditions Jie Zhang and Alan M. Bond*

School of Chemistry, Monash University, P.O. Box 23, Clayton, Victoria 3800, Australia

Conventional cyclic voltammetric studies on the oxidation of millimolar concentrations (mg masses) of trans-[Mn(CN)(CO)2{P(OPh)3}(Ph2PCH2PPh2)] (trans-Mn) dissolved in milliliter volumes of bulk ionic liquid, 1-butyl3-methylimidazolium hexafluorophosphate (BMIM‚PF6), give rise to a reversible [trans-Mn]0/+ process. In this study, it is shown that equally well-defined reversible voltammetry can be more economically obtained under microchemical ionic liquid conditions by employing a chemically modified electrode (µg quantities of trans-Mn adhered to a glassy carbon electrode covered with microliter volumes of water-immiscible BMIM‚PF6) in contact with aqueous (0.1 M KPF6) electrolyte. The ability to obtain electrochemical data that are directly relatable to bulk ionic liquid media under these microchemical conditions is principally associated with the dissolution of electrogenerated solid [trans-Mn]+ in the layer of waterimmiscible BMIM‚PF6 present at the electrode/ionic liquid/aqueous electrolyte interface. If the BMIM‚PF6 layer is sufficiently thick, mass transport of the dissolved species is governed by semi-infinite linear diffusion. Under these conditions, the voltammetric waveshape and position, but not the current magnitude are the same as those found when conventional bulk ionic liquid conditions are employed. In contrast, use of very thin layers produces voltammograms that exhibit the characteristics expected for a reversible process in which the mass transport process is predominantly governed by finite rather than semi-infinite diffusion. A theoretical model has been developed that describes the transformation from thick- to thin-layer type behavior as the thickness of the ionic liquid layer is decreased. Practical uses of conventional high-temperature molten salts1 are limited because they are generally moisture sensitive. In contrast, room-temperature ionic liquids, which are equivalent to molten salts, often have attractive properties, such as negligible vapor pressure, low toxicity, high chemical and thermal stability, * Corresponding author. E-mail: [email protected]. (1) (a) Wilkes, J. S. Green Chem. 2002, 4, 73 and references therein. (b) Lipsztajn, M.; Osteryoung, R. A. Inorg. Chem. 1985, 24, 716. (c) Sahami, S.; Osteryoung, R. A. Anal. Chem. 1983, 55, 1970.

6938 Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

and the ability to dissolve a wide range of compounds.2 Consequently, ionic liquids have been suggested as “green” replacements for the volatile conventional organic solvents employed in organic synthesis,2 solar cell applications,3 and solvent extraction processes.4 Large potential windows and high conductivities, which allow studies to be undertaken without added supporting electrolyte,2 are additional advantages for electrochemists. As a result of these attractive features, electrochemical studies in ionic liquids have become relatively common (see refs 5 and 6 for example). In recent times, much attention is being devoted to the introduction of microchemical (green chemistry) approaches that minimize the quantities of solvents and chemicals employed in chemical investigations.7 Conventional electrochemical studies with ionic liquids employ 1-10-mL volumes and millimolar concentrations (mg masses) of dissolved redox-active material. In this paper, initially, we employed these conventional conditions in order to establish the basic electrochemical characteristics associated with the voltammetric oxidation of trans-[Mn(CN)(CO)2{P(OPh)3}(Ph2PCH2PPh2)] (trans-Mn) in bulk quantities of 1-butyl-3-methylimidazolium hexafluorophosphate (BMIM‚PF6) (2) (a) Dupont, J.; de Souza, R. F.; Suarez, P. A. Z. Chem. Rev. 2002, 102, 3667. (b) Sheldon, R. Chem. Commun. 2001, 2399. (c) Welton, T. Chem. Rev. 1999, 99, 2071. (3) (a) Stathatos, E.; Lianos, R.; Zakeeruddin, S. M.; Liska, P.; Gra¨tzel, M. Chem. Mater. 2003, 15, 1825. (b) Kubo, W.; Kitamura, T.; Hanabusa, K.; Wada, Y.; Yanagida, S. Chem. Commun. 2002, 374. (c) Bonhoˆte, P.; Dias, A. P.; Papageorgiou, N.; Kalyanasundaram, K.; Gra¨tzel, M. Inorg. Chem. 1996, 35, 1168. (4) (a) Visser, A. E.; Swatloski, R. P.; Reichert, W. M.; Mayton, R.; Sheff, S.; Wierzbicki, A.; Davis, J. H.; Rogers, R. D. Environ. Sci. Technol. 2002, 36, 2523. (b) Visser, A. E.; Holbrey, J. D.; Rogers, R. D. Chem. Commun. 2001, 2484. (c) Chun. S.; Dzyuba, S. V.; Bartsch, R. A. Anal. Chem. 2001, 73, 3737. (5) (a) Kakiuchi, T.; Tsujioka, N. Electrochem. Commun. 2003, 5, 253. (b) Quinn, B. M.; Ding, Z. F.; Moulton, R.; Bard, A. J. Langmuir 2002, 18, 1734. (c) Boxall, D. L.; O’Dea, J. J.; Osteryoung, R. A. J. Electrochem. Soc. 2002, 149, E468. (d) Boxall, D. L.; Osteryoung, R. A. J. Electrochem. Soc. 2002, 149, E185. (e) Hagiwara, R.; Hirashige, T.; Tsuda, T.; Ito, Y. J. Electrochem. Soc. 2002, 149, D1-D6. (f) Wadhawan, J. D.; Schro¨der, U.; Neudeck, A.; Wilkins, S. J.; Compton, R. G.; Marken, F.; Consorti, C. S.; de Souza, R. F.; Dupont, J. J. Electroanal. Chem. 2000, 493, 75. (6) (a) Hultgren, V. M.; Mariotti, A. W. A.; Bond, A. M.; Wedd, A. G. Anal. Chem. 2002, 74, 3151. (b) Zhang J.; Bond, A. M. Anal. Chem. 2003, 75, 2694. (7) For recent review, see for example: (a) Haswell, S. J.; Watts, P. Green Chem. 2003, 5, 240. (b) Anastas, P. T.; Kirchhoff, M. M. Acc. Chem. Res. 2002, 35, 686. 10.1021/ac034921e CCC: $25.00

© 2003 American Chemical Society Published on Web 11/11/2003

ionic liquid. To ascertain whether equivalent ionic liquid electrochemical data could be obtained under microchemical conditions, glassy carbon electrodes modified with microgram masses of solid microparticles of trans-Mn were coated with a layer of hydrophobic BMIM‚PF6 (1 µL or less volume) prior to being placed in contact with conventional volumes of 0.1 M KPF6 aqueous electrolyte solution. This microchemical technique employs a configuration that is closely related to the one used in studies of electron-transfer reactions across immiscible organic solvent/aqueous electrolyte interfaces as pioneered by Anson and co-workers8 and subsequently employed by other groups.9 However, we directly use microgram quantities of adhered solid rather than redox-active compound dissolved in the organic solvent (requires larger masses), and another difference is that the ionic liquid does not require addition of electrolyte to achieve adequate conductivity. Voltammetric measurements under these microchemical conditions have been undertaken in this study with various quantities of trans-Mn solid and thicknesses of BMIM‚PF6 layers, to probe the mass transport phenomena that take place within the pure phases and across the electrode/microparticle/ionic liquid/ aqueous interface. A simplified theoretical model, developed on the basis of these findings, explains the relationship of voltammetric data obtained under microchemical conditions to that found when trans-Mn is dissolved in bulk BMIM‚PF6. EXPERIMENTAL SECTION Chemicals. Trans-Mn (structure 1) was synthesized, purified,

and characterized as described in the literature.10 Cobalticinium hexafluorophosphate ([Co(Cp)2][PF6], 98%, Stream Chemicals) and KPF6 (98%, Sigma-Aldrich) were used as received from the manufacturer. BMIM‚PF6 (structure 2) (>97%, Sigma-Aldrich) was

dried over basic alumina for at least 12 h prior to use, when employed as a bulk solvent in conventional voltammetric experiments. (8) See, for example: (a) Shi, C.; Anson, F. C. J. Phys. Chem. B 2001, 105, 8963. (b) Chung, T. D.; Anson, F. C. J. Electroanal. Chem. 2001, 508, 115. (c) Shi, C.; Anson, F. C. J. Phys. Chem. B 2001, 105, 1047. (d) Shi, C.; Anson, F. C. Anal. Chem. 1998, 70, 3114. (e) Hubbard, A. T.; Anson, F. C. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1970; Vol. 4, p 129. (9) See, for example: (a) Barker, A. L.; Unwin, P. R. J. Phys. Chem. B 2000, 104, 2330. (b) Shafer, H. O.; Derback, T. L.; Koval, C. A. J. Phys. Chem. B 2000, 104, 1025. (10) (a) Hogan, C. F.; Bond, A. M.; Neufeld, N. K.; Connelly, N. G.; Llamas-Rey, E. J. Phys. Chem. A 2003, 107, 1274. (b) Connelly, N. G.; Hassard, K. A.; Dunne, B. J.; Orpen, A. G.; Raven, S. J.; Carriedo, G. A.; Riera, V. J. Chem. Soc., Dalton Trans. 1988, 1623. (c) Bombin, F.; Carriedo, G. A.; Miguel, J. A.; Riera, V. J. Chem. Soc., Dalton Trans. 1981, 2049.

Instrumentation and Procedures. Voltammetric and uncompensated resistance measurements were undertaken with a BAS 100B (Bioanalytical Systems) electrochemical workstation. A conventional three-electrode electrochemical cell was employed. For conventional experiments with bulk quantities of ionic liquid as solvent, 1-mm-diameter glassy carbon (GC) and Pt wire were used as the working and counter electrodes, respectively, and a Ag wire dipped in a tube containing BMIM‚PF6 and separated from the test solution by a glass frit was employed as the quasireference electrode. The potential of the quasi-reference electrode was then calibrated against that of the IUPAC-recommended [Co(Cp)2]+/0 process using a 1 mM [Co(Cp)2][PF6] solution (Co(Cp)2 ) cobaltocene) as an internal reference.11 For voltammetric measurements under microchemical conditions, Ag/AgCl (3 M NaCl) placed in the aqueous electrolyte phase was used as the reference electrode. The Pt wire counter electrode also was placed in the aqueous electrolyte phase in these experiments. All voltammetric data were collected at 20 ( 1 °C, and solutions were degassed with nitrogen for between 5 and 10 min prior to undertaking electrochemical measurements. The procedure for mechanical attachment of microcrystalline particles onto electrodes surfaces has been described in detail elsewhere.12,13 In brief, between 1 and 5 mg of trans-Mn compound was placed on weighing paper. The GC electrode was then pressed onto the paper substrate and rubbed over the solid. In this manner, microgram amounts of the solid were transferred to the electrode surface to form microparticle islands exhibiting a layered structure and having 10-100-µm-diameter dimensions.10a Since the electrode surface is maintained parallel to the paper substrate during the course of transfer of solid, the thickness of every island is likely to be similar. Although the fabrication of the microparticlemodified electrodes is not well controlled, the peak potentials and waveshape characteristics of the voltammograms were not found to be highly sensitive to either the amount of solid (if sufficient) or the size of the microparticles. The mass (high or low terminology is used below in the Results and Discussion section to define two classes of behavior) of trans-Mn adhered on the glassy carbon electrode is controlled by pressing the electrode onto (high or low) quantities of trans-Mn initially present on the paper substrate. The consistency of mass level transferred (high or low) to the glassy carbon electrode surface can be qualitatively confirmed by visual or microscopic inspection of the highly colored (deep yellow) microcrystalline material. The suitably modified electrode was then placed either directly in contact with ionic liquid solution or in microchemical-type experiments coated with a layer of ionic liquid (1 µL for what is designated as a “thick” layer and less than or equal to 0.1 µL for what is designated as a “thin” layer in the Results and Discussion section) via use of a microsyringe. Finally, this chemically modified electrode was placed in contact with 0.1 M KPF6 aqueous electrolyte solutions. The microchemical approach produces a cell configuration that is analogous to that used in the thin-layer cell technique involving investigations of redox reactions occurring (11) Stojanovic, R. S.; Bond, A. M. Anal. Chem. 1993, 65, 56. (12) Bond, A. M.; Cooper, J. B.; Marken, F.; Way, D. M. J. Electroanal. Chem. 1995, 396, 407. (13) Bond, A. M. Broadening electrochemical horizons: principles and illustration of voltammetric and related techniques; Oxford University Press: Oxford, 2002; Chapter 5.

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

6939

Figure 1. Schematic illustration of the two forms of chemically modified GC electrode/solution interface used in this study. Scheme 1: An array of microparticles adhered to an electrode in contact with bulk ionic liquid. Scheme 2: Electrode/microparticle array/ionic liquid layer/aqueous electrolyte microchemical configuration.

across organic solvent/aqueous electrolyte interfaces.8 When required, 1 mM [Co(Cp)2]+ was present in the ionic liquid for reference potential calibration purposes. The solubility of BMIM‚PF6 in water (10 mM level or ∼2% in mass)14 is comparable to that of water-immiscible organic solvents, such as 1,2-dichloromethane and benzonitrile,15 that are commonly employed in electrochemical studies of reactions that occur across water/organic solvent interfaces.16,17 In the situation prevailing in our studies, time-dependent dissolution of the ionic liquid layer was prevented by presaturation of the aqueous electrolyte solution with the BMIM‚PF6. The ionic liquid used also was presaturated with water in order to maintain constant viscosity.18 A schematic representation of the chemically modified working electrode in contact with bulk ionic liquid or an ionic liquid layer/aqueous electrolyte interface is shown in Figure 1. In Figure 1 (scheme 1), the reference and counter electrodes are in contact with bulk ionic liquid, while in scheme 2, these electrodes are present in the aqueous electrolyte phase. Prior to the attachment of solid, the GC working electrode was polished with a 0.3-µm Al2O3 (Buehler) slurry, washed successively with water and acetone, and finally dried with tissue paper. (14) (a) Swatloski, R. P.; Visser, A. E.; Reichert, W. M.; Broker, G. A.; Farina, L. M.; Holbrey, J. D.; Rogers, R. D. Green Chem. 2002, 81. (b) Wong, D. S. H.; Chen, J. P.; Chang, J. M.; Chou, C. H. Fluid Phase Equilib. 2002, 194, 1089. (c) Fadeev, A. G.; Meagher, M. M. Chem. Commun. 2001, 295. (15) The Merck Index, 12th ed.; Budavari, S., O’Neill, M. J., Eds.; Merck: Whitehouse Station, NJ, 1996. (16) Girault, H. H.; Schiffrin D. J. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1989; Vol. 15, p 1. (17) Liquid Interfaces in Chemical, Biological, and Pharmaceutical Applications; Volkov, A. G., Ed.; Marcel Dekker: New York, 2001. (18) Schro¨der, U.; Wadhawan, J. D.; Compton, R. G.; Marken, F.; Suarez, P. A. Z.; Consorti, C. S.; de Souza, R. F.; Dupont, J. New J. Chem. 2000, 24, 1009.

6940

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

Figure 2. Comparison of experimental (-) and simulated (O) cyclic voltammograms obtained at a 1-mm-diameter GC disk electrode (scan rate 0.1 V s-1) for oxidation of 7.4 mM trans-Mn dissolved in BMIM‚PF6 ionic liquid. The simulated voltammogram was calculated for a reversible one-electron-transfer process using D ) 9.1 × 10-9 cm2 s-1 for both trans-Mn and [trans-Mn]+, electrode area 0.008 57 cm2, Ru ) 4500 Ω, and T ) 293 K.

RESULTS AND DISCUSSION Voltammetry of trans-Mn Dissolved in Bulk BMIM‚PF6. The [trans-Mn]0/+ process, when trans-Mn is dissolved in CH3CN or CH2Cl2, is a well-defined one-electron reversible process at a glassy carbon electrode (eq 1).10a

trans-Mn(organic solvent) a [trans-Mn]+(organic solvent) + e- (1) To be able compare results obtained by microchemical methods at an electrode/ionic liquid/aqueous electrolyte interface with those in the dissolved state in an ionic liquid, details concerning the voltammetric oxidation of trans-Mn dissolved directly in bulk BMIM‚PF6 were initially established. Although the solubility of trans-Mn in BMIM‚PF6 is reasonably high (10 mM level), the rate of dissolution is very slow. Consequently, to obtain concentrations in the millimolar range, solid trans-Mn particles in contact with the ionic liquid had to be left in an ultrasonic bath for ∼2 h at ∼40 °C. A cyclic voltammogram obtained at a glassy carbon electrode for the [trans-Mn]0/+ process using a trans-Mn concentration of 7.4 mM and a scan rate (v) of 0.1 V s-1 is shown in Figure 2. As is the case when trans-Mn is dissolved in organic solvents, a very well-defined one-electron-transfer process is observed. The voltammetric process in BMIM‚PF6 is described by eq 2.

trans-Mn(ionic liquid) a [trans-Mn]+(ionic liquid) + e- (2) The linear dependence of peak current on the square root of v (0.01-5 V s-1) suggests that eq 2 describes a reversible diffusioncontrolled process in BMIM‚PF6 on the designated voltammetric time scale. Excellent agreement between experimental and simulated voltammograms is obtained for a reversible oneelectron-transfer process with diffusion coefficients of 9.1 × 10-9 cm2 s-1 (for both trans-Mn and [trans-Mn]+), a reversible potential

Figure 3. Comparison of an experimental (-) (scan rate 0.1 V s-1) cyclic voltammogram obtained when an array of solid trans-Mn microparticles is adhered to a 1-mm-diameter GC disk electrode (highmass case) that is placed in contact with BMIM‚PF6 and simulated data (O) that represents a cyclic voltammogram for the case where trans-Mn is dissolved in bulk ionic liquid. The simulated voltammogram has been calculated using the parameters given in the caption for Figure 2 but is normalized to the peak current of the experimentally obtained voltammogram in order to demonstrate the shape and peak potential equivalence of voltammograms obtained from adhered and dissolved material.

of 1.664 V versus [Co(Cp)2]+/0, and an uncompensated resistance, Ru, value of 4500 Ω (the measured value of Ru obtained via use of the technique available with the BAS instrument was 4000 ( 500 Ω). The simulated result was obtained by using the Digisim simulation software package (Bioanalytical Systems, Inc. West Lafayette, IN).19 The fact that the measured peak-to-peak separation (∆Ep) is 63 mV in BMIM‚PF6 and slightly higher than the theoretical value of 56 mV for a reversible process at 20 °C20,21 is predominately therefore attributable to the influence of uncompensated resistance, Ru. The unusually small diffusion coefficient value of 9.1 × 10-9 cm2 s-1 obtained in BMIM‚PF6 is attributable to the very high viscosity of this ionic liquid in its pure form. Consequently, a significant contribution to this value from electron hopping is feasible, particularly if this process is diffusion controlled. However, even with this very low diffusion coefficient value, any contribution from this source was calculated22,23 to be negligible ( 0, interface 2: c (Ox+ (ionic liquid)) F (E - E 0′ ) exp solid) RT c (Red (solid))

[

c (Ox+ (ionic liquid)) F (E - E 0′ ) exp il ) RT c (Red (ionic liquid))

[

]

]

t > 0, interface 2: ∂c D ) 0 for the three species involved in eqs 5 and 6 ∂x

∑

t > 0, interface 1 (x ) d1): Dred,solid

∂c(Red (solid)) )0 ∂x

t > 0, interface 3 (x ) d2): Dred,il

∂c(Red (ionic liquid)) )0 ∂x

Dox,il

∂c(Ox+ (ionic liquid)) )0 ∂x

In these equations, F, R, and T have their usual meanings; c* is the initial concentration of Red(solid); E is the working electrode 0′ and E 0′ potential versus the reference scale; E solid il are the formal potentials of processes described by eqs 5 and 6, respectively; Dred,solid, Dred,il, and Dox,il are the diffusion coefficients of Red (solid), Red (ionic liquid), and Ox+ (ionic liquid), respectively; and d1 and d2 are the thicknesses of the solid and ionic liquid phases, respectively, which are assumed to be constant. The simulation of the voltammetric theory using the Crank-Nicolson implicit finite difference method31 can be efficiently undertaken as described in the previous studies.6b,24 Since we are interested in understanding the conditions related to mass and hence thickness of the assumed binary phase that determine the shape and position of voltammograms, the simulated results employ the following arbitrary parameters: c* ) 2 M and diffusion coefficients of 1 × 10-10 and 1 × 10-8 cm2 s-1 for the solid- and solution-state species, -1 respectively; E 0′ il ) 0.40 V; v ) 0.1 V s ; and radius of electrode 0.05 cm. Other simulation parameters used are defined in the 0′ figure captions. The value of 0.30 V used for E solid , being less positive than 0.40 V, ensures that the simulated voltammogram obtained for the adhered solid when the product dissolves is identical to that of fully dissolved species6b when the mass transport is governed by semi-infinite diffusion. This equivalence has been observed in the voltammetry of a wide range of redoxactive microparticles adhered onto an electrode surface and then placed in contacted with bulk ionic liquid.6,24 (31) Britz, D. Digital Simulation in Electrochemistry, 2nd ed.; Springer-Verlag: New York, 1988.

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

6945

Figure 7. Influence of phase thicknesses d1 (solid microparticle phase) and d2 (ionic liquid phase) on simulated cyclic voltammograms (scan rate 0.1 V s-1) and according to the reaction scheme given in Figure 6. (a) d1 ) 3 × 10-4 cm and d2 ) 3 × 10-3 cm, (b) d1 ) 3 × 10-4 cm and d2 ) 1 × 10-4 cm, (c) d1 ) 1 × 10-5 cm and d2 ) 3 × 10-3 cm, (d) d1 ) 1 × 10-5 cm and d2 ) 1 × 10-4 cm, and (e) d1 ) 1 × 10-5 cm and d2 ) 1 × 10-5 cm. Other parameters used in the simulation are defined in the text.

From the experimental data presented in Figure 5, it is obvious that the mass (thickness) of both trans-Mn islands and the thickness of the BMIM‚PF6 layer are crucial in determining the voltammetric characteristics. It is well known that the presence of a phase boundary does not affect a semi-infinite diffusion process when the phase thickness is greater than 6(Dttotal)1/2 (ttotal is time taken for the voltammetric measurement).20,31 Thus, for the diffusion coefficient values given above and a ttotal value of 16 s, a thickness of 3 × 10-5 cm for the solid phase or 3 × 10-4 cm for the ionic liquid layer would achieve semi-infinite diffusion conditions. Consequently, considerably smaller phase thicknesses than these values can then be employed in the simulation, to examine the influence of the phase boundary on the mass transport. The theoretical dependence of both the solid and ionic liquid layer phase thicknesses is shown in the simulations presented in Figure 7. This trend is in agreement with experimental data that indicated the characteristics of voltammograms change from those associated with a semi-infinite diffusion6946 Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

mediated reversible process for high mass of redox probe-thick ionic liquid layer configuration to those associated with classical thin-layer voltammetry when the thickness of both phases are reduced. In experimental studies, the “high-mass” and “low-mass” combinations referred to in discussion of data are equated to “high” and “low” solid-phase thickness, respectively, in the theoretical model. Again, as observed experimentally, simulated voltammograms exhibit characteristics between these two extreme cases when the thickness of one phase is reduced. As a means of demonstrating the influence of the thickness of the solid and ionic liquid phases on the diffusion processes, the concentration profiles of the species involved in the redox reaction at E ) 0.40 V (value of E 0′ il ) are displayed in Figure 8 for the oxidation sweep of a voltammogram. To clearly present the data, all concentrations in Figure 8 are divided by c* to obtain the normalized concentration, Cnorm. The real distance, x, from interface 2 to any position inside the solid phase (or ionic liquid layer) also is divided by the thickness d1 (or d2) to obtain the

Figure 8. Concentration profiles of species involved in the redox reactions in the initial oxidative sweep direction of cyclic voltammograms at a potential E ) 0.40 V. Parameters used in the simulation are the same as those for Figure 7.

normalized distance, Xnorm (This is not the same as the normalized distance term commonly associated with numerical simulations.). Thus, the maximum value of Xnorm in each phase is 1, although d1 and d2 are very different. When semi-infinite diffusion is applicable (thick layer case in Figure 8a), the concentrations of the redox species remain at their bulk value for most of the normalized distance and a reversible voltammogram having the expected characteristics of this kind of diffusion-mediated process is observed (Figure 5a and Figure 7a). When the thickness of one of the phases is reduced (Figure 8b and c), the mass transport in one phase is blocked by the phase boundary, which is evidenced by the clear deviation of the concentration from the initial value red at the phase boundary (Xnorm ) 1); thus, |i ox p /i p |deviates from 1 (Figure 5b and c, and Figure 7b and c). When the thicknesses of both phases are reduced (thin-layer case in Figure 8d), the diffusion processes in two phases are blocked. Thus, a voltammogram with characteristics lying between a diffusion-controlled

and the thin-layer cell cases is observed (Figure 5d and Figure 7d). When the thicknesses of the two phases are further reduced, the diffusion process eventually becomes unimportant. Under these conditions, almost no concentration gradient is present in the two phases (Figure 8e), so almost ideal thin-layer type cell behavior is obtained. Despite the excellent qualitative agreement of experimental voltammograms and those simulated under conditions where a binary-phase model is employed to describe the conversion of solid trans-Mn to solid [trans-Mn]+, this should not be taken as evidence that the activities of solid trans-Mn and [trans-Mn]+ are proportional to their molar concentration ratio. Other mechanisms involving unit activity of individual phases or nucleation growth processes could in all probably be used to represent the solidsolid conversion reaction that is assumed to have occurred prior to the dissolution process and still lead to the voltammetric characteristics observed experimentally. Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

6947

CONCLUSIONS A microchemical technique that is based on the use of microgram masses of solid microparticles adhered to an electrode that is coated with microliter volumes of ionic liquid prior to the chemically modified electrode being placed in contact with aqueous electrolyte solution can provide an efficient method of establishing data equivalent to that obtained via voltammetric methods requiring the use of bulk ionic liquids. The principles of this microchemical technique have been established using the [trans-Mn]+/0 redox probe and BMIM‚PF6 as the ionic liquid phase and where trans-Mn, [trans-Mn]+, and the ionic liquid are all insoluble/immiscible in water.

6948

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

ACKNOWLEDGMENT The authors thank the Australian Research Council and the Monash University Research Fund for financial support of this project. Professor Neil Connelly (School of Chemistry, University of Bristol) is acknowledged for his generous provision of trans[Mn(CN)(CO)2{P(OPh)3}(Ph2PCH2PPh2)].

Received for review August 7, 2003. Accepted October 2, 2003. AC034921E