Anal. Chem. 2000, 72, 3241-3244
Chronopotentiometric Analysis of Highly Resistive Media Joseph Wang* and Baomin Tian
Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, New Mexico 88003
Reciprocal derivative chronopotentiometry is shown to be well suited for performing electrochemistry in highly resistive solutions in connection to conventional-sized electrodes. The fundamentally different operational principles of reciprocal derivative chronopotentiometry (compared to controlled-potential voltammetry) reduces its susceptibility to solution resistance effects. The influences of the electrolyte concentration and constant current upon the reciprocal derivative chronopotentiometric peak area and potential are examined in different solvents. Welldefined, analytically useful peaks are observed in highly resistive media (with low electrolyte/analyte ratios). Narrower peaks, shifted to higher potential values, are observed upon increasing the solution resistance. Yet, the analytical signal (square root of the peak area) is nearly independent of the ohmic drop. The experimental results are compared with the theoretical predictions. Considerable ohmic drop distortions are observed in analogous voltammetric experiments. The defined chronopotentiometric peaks permit convenient quantitation of micromolar analyte concentrations in highly resistive media. Such observations open the door to new electrochemical applications of conventional electrodes that were previously restricted to ultramicroelectrodes. The use of organic solvents is of great interest in electrochemical research and technology.1 Unfortunately, in all controlledpotential methods, the uncompensated solution resistance introduces an error because the effective potential is less than the applied potential.2 Voltammetry in highly resistive media is thus complicated by the ohmic iR drop.1,2 Such ohmic distortion can be minimized using sufficiently small electrodes for which the iR drop is negligible even in highly resistive solutions.1-5 Electronic iR compensation is also included in modern voltammetric analyzers but cannot fully address severe ohmic distortions. In this paper, we report on the use of reciprocal derivative chronopotentiometry (in its reciprocal derivative form) for providing analytically useful data at conventional-sized electrodes in nonaqueous media containing little added electrolyte. Chronopotentiometry is an electrochemical technique in which a constant (1) Coetzee, J. F. Pure Appl Chem. 1986, 58, 1091. (2) Bond, A. M. Electrochim. Acta 1987, 32, 863. (3) Wipf, D. O.; Wightman, R. M. Anal. Chem. 1990, 62, 98. (4) Geng, L.; Ewing, A. G.; Jernigan, J. C.; Murray, R. C. Anal. Chem. 1986, 58, 852. (5) Jaworski, A.; Donten, M.; Stojek, Z. J. Electroanal. Chem. 1995, 305, 106. 10.1021/ac000230w CCC: $19.00 Published on Web 06/17/2000
© 2000 American Chemical Society
current is imposed between the working and auxiliary electrodes, while the resulting potential response is monitored.6,7 This relatively old technique has not received major attention in modern electroanalysis, compared to controlled-potential voltammetric methods. The power of chronopotentiometry has been greatly enhanced in recent years with the introduction of microprocessorcontrolled instruments that register ultrafast events and convert its wave-shaped response to a convenient peak-shaped (differential) display.8 Such development has paved the way to new applications of chronopotentiometry, ranging from stripping measurements of trace metals 9 to the transduction and biosensing of DNA hybridization.10 In the following sections, we will illustrate that the fundamentally different operational principles of chronopotentiometry make it less susceptible to ohmic drop effects. The smaller susceptibility of stripping potentiometry to the solvent resistance has been exploited for extending its scope toward electropositive elements, such as alkali metals.11 The present investigation demonstrates defined and analytically useful chronopotentiometric/conventional electrode data for ferrocene in different organic solvents containing various electrolyte levels. Increasing the solution resistance shifts the reciprocal derivative chronopotentiometric dt/dE vs E curves to higher potentials, results in sharper peaks, and has a negligible effect upon the peak area. The well-defined peaks that are observed in highly resistive media thus permit convenient quantitation of micromolar levels of the target analytes. Such observations promise to open resistive media to electrochemical studies at conventional electrodes. EXPERIMENTAL SECTION Apparatus. Reciprocal derivative chronopotentiometric experiments were carried out using the TraceLab PSU20 system (Radiometer, Copenhagen, Denmark) in connection to an IBM 386 personal computer. Cyclic voltammetry was performed with CV-50W voltametric analyzer (BAS). All experiments were performed at room temperature using a 10-mL cell (model VC-2, BAS). The three-electrode system consisted of a glassy carbon working electrode (3-mm diameter, BAS), a platinum wire counter electrode, and Ag/AgCl wire quasi-reference electrode. (6) Delahay, P.; Mamantov, G. Anal. Chem. 1955, 27, 478. (7) Sturrock, P.; Hughey, J.; Vaudreuil, B.; O’Brien, G.; Gibson, R. J. Electrochem. Soc. 1975, 122, 1195. (8) Jagner, D. TrAC, Trends Anal. Chem. 1983, 2, 53. (9) Ostapczuk, P. Anal. Chim. Acta 1993, 273, 35. (10) Wang, J.; Cai, X.; Rivas, G.; Shiraishi, H. Anal. Chim. Acta 1996, 326, 141. (11) Coetzee, J. F.; Hussan, A.; Petrck, T. R.Anal. Chem. 1983, 55, 120.
Analytical Chemistry, Vol. 72, No. 14, July 15, 2000 3241
Figure 1. Cyclic voltammograms (A) and reciprocal derivative chronopotentiograms (B) for 2 × 10-5 M ferrocene in acetonitrile containing (a) 3, (b) 10, and (c)100 µM THAP. Scan rate (A), 100 mV/s; constant current (B), 0.4 µA.
Figure 2. Reciprocal derivative chronopotentiograms for (a) 0, (b) 10, (c) 20, (d) 30, and (e) 40 µM catechol in water containing (A) 30 µM and (B) 30 mM sodium chloride. Constant current, 0.6 µA.
Reagents. All chemicals were of analytical reagent grade and used as received. All solutions were prepared with double-distilled water. Ferrocene and catechol were received from Sigma, acetonitrile and methylene chloride were received from EM Science and Fisher, respectively, and tetra-n-hexylammonium perchlorate (THAP) was purchased from Alfa Aesar. Procedure. Before use, the glassy carbon disk electrode was carefully polished with 0.05-µm alumina slurry and rinsed with double-distilled water. Reciprocal derivative chronopotentiograms (dt/dE vs E) were recorded by applying the constant current onto the electrode system using a quiescent solution. The chronopotentiograms were recorded following data smoothing, but without baseline fitting or correction. The solution was stirred for 20 s after each run and was kept quiescent for additional 30 s prior to the next one. RESULTS AND DISUSSION Figure 1 compares cyclic voltammograms (A) and reciprocal derivative chronopotentiograms (B), recorded at a conventionalsize glassy carbon electrode, for 2 × 10-5 M ferrocene in acetonitrile solutions containing different levels of the THAP electrolyte. As expected for the voltammetric behavior of macrosized electrodes, a severe peak distortion is observed upon lowering the THAP concentration to 1 × 10-5 (A,b) and 3 × 10-6M (A,a). The peak separation (∆Ep) increases dramatically from 177 (at 1 × 10-4 M) to 622 mV (for 3 × 10-6 M THAP) over this electrolyte concentration range. In contrast, defined peaks are observed for the corresponding chronopotentiometric measurements (B). Sharper peaks are actually observed upon lowering the electrolyte concentration; the peak width at half-height (b1/2) decreases from 105 mV at 1 × 10-4 M to 30 mV at 3 × 10-6 M THAP. The smaller peak width in low electrolyte levels is accompanied by a larger peak height (see discussion below). Note, however, the marked shift of the peak potential [from +0.41 (c) to +1.30 V (a)] with the decreased electrolyte concentration, which indicates that such chronopotentiometric measurements are influenced by solution resistance effects. Analogous measurements of ferrocene in methylene chloride containing different electrolyte concentrations yielded similar trends in the voltam3242 Analytical Chemistry, Vol. 72, No. 14, July 15, 2000
Figure 3. Calibration data in resistive medium. (A) Reciprocal derivative chronopotentiograms for (a) 0, (b) 4, (c) 8, (d) 12, and (e) 16 µM ferrocene. (B) Calibration plots for ferrocene over the 2-50 µM range: (a) A vs C; (B) A1/2 vs C. Constant current, 0.5 µA. Medium, acetonitrile containing 20 µM THAP.
metric (A) and chronopotentiometric (B) data (not shown). In this medium, the chronopotentiometric peak b1/2 decreased from 160 to 70 mV upon lowering the level of THAP from 3 × 10-4 to 1 × 10-5 M, while the peak potential shifted from +0.58 to +0.95 V, respectively. Poorly defined, highly distorted, voltammetric peaks were observed for the dilute electrolyte solutions. The use of aqueous media, containing little or no deliberately added electrolyte, can also benefit from the unique behavior of the reciprocal derivative chronopotentiometric protocol. Figure 2 displays chronopotentiograms, recorded at the macrosized glassy carbon electrode, for aqueous solutions containing 3 × 10-5 (A) and 3 × 10-2 M (B) sodium chloride and increasing levels of catechol (1 × 10-5 - 4 × 10-5 M, b-d). The well-defined peaks observed with both electrolyte levels indicate convenient quantitation of these micromolar catechol concentrations. Similar to the nonaqueous chronopotentiometric operation, sharper and higher peaks are observed using lower sodium chloride concentrations. In view of the higher conductivity of the aqueous solution, no
Figure 4. Effect of the applied current upon the chronopotentiometric peak area in different solutions. (A) 40 µM ferrocene in acetonitrile containing (a) 3, (b) 10, (c) 30, and (d) 100 µM THAP. (B) 40 µM ferrocene in methylene chloride containing (a) 10, (b) 30, (c) 100, and (d) 300 µM of THAP. (C) 20 µM catechol in water containing (a) 3, (b) 10, and (c) 50 µM sodium chloride.
Figure 5. Effect of the applied current upon the chronopotentiometric peak potential in different solutions. (A) 40 µM ferrocene in acetonitrile containing (a) 3, (b) 10, (c) 30, and (d) 100 µM THAP. (B) 40 µM ferrocene in methylene chloride containing (a) 10, (b) 30, (c) 100, and (d) 300 µM THAP. (C) 20 µM catechol in water containing (a) 3, (b) 10, and (c) 50 µM sodium chloride.
apparent change in the catechol peak potential is observed between 3 × 10-5 and 3 × 10-2 M sodium chloride. Overall, the data of Figure 2 indicate convenient quantitation of micromolar analyte levels in low ionic strength aqueous media. Such chronopotentiometric quantitation relies on the attainment of a well-defined concentration dependence. Figure 3A displays reciprocal derivative chronopotentiograms for increasing levels of ferrocene [(4-16) × 10-6 M, b-e] in an acetonitrile solution containing 2 × 10-5 M THAP electrolyte. Despite the resistive medium, the conventional glassy carbon electrode yields well-defined peaks for these micromolar concentrations. As common for reciprocal derivative chronopotentiometric measurements,8 the peak area (obtained by integrating the counts over potential range of the peak), A, corresponds to the transition time signal, τ. The relationship between the analyte concentration, C, and the transition time (and hence the peak area), is given by the Sand equation:
τ1/2 ) A1/2 ) nFaπ1/2D1/2C/2i
(1)
where i is the applied current, a is the electrode area, n is the number of electrons transferred, F is the Faraday constant, and D is the diffusion coefficient. Accordingly, the square root of the peak area, and not the peak area, is proportional to the solute
concentration (Figure 3 B, a vs b). Note that the Sand equation, which describes ideal conditions, is obeyed in such resistive media (that affect other peak parameters). The plot of A1/2 vs the ferrocene concentration is highly linear over the entire (2-50) × 10-6 M range tested (b), with a slope of 2.09 ms1/2/µM and correlation coefficient of 0.9996. A detection limit of ∼2 × 10-6 M ferrocene can be estimated from the signal-to-noise characteristics (S/N ) 3) of Figure 3A,b. The low background, facilitating such micromolar measurements (e.g., Figures 2 and 3a) is attributed to the powerful digital noise filtration and tangent baseline fitting of the instrument-supported software. Figure 4 displays the influence of the applied current upon the peak area for different media and electrolyte concentrations. The ideal behavior of eq 1, is obeyed, and the square root of the peak area increases linearly with the reciprocal of the current for all the tested solutions (i.e., constant iτ1/2 and iA1/2 values). The different slopes of the A1/2 vs i-1 plots reflect differences in the n, D, and C values of the ferrocene and catechol analytes in these media. Note also that A1/2, at a given current, is nearly independent of the electrolyte level (i.e., resistance). Accordingly, the sharper peaks observed upon lowering the electrolyte concentration (e.g., Figure 1) result in larger peak heights, i.e., same peak area. Figure 5 examines the influence of the constant current upon the reciprocal derivative chronopotentiometric peak potential of Analytical Chemistry, Vol. 72, No. 14, July 15, 2000
3243
ferrocene and catechol in acetonitrile (A), methylene chloride (B), and water (C) solutions containing different electrolyte concentrations (a-d). In all cases, the peak potential (at a given current) shifts to higher values upon lowering the level of the supporting electrolyte. Such shift is in agreement to the observations of Figure 1. The peak potential also shifts linearly with the applied current for all solvents and electrolyte concentrations tested. The increased peak potential with the solution resistance and applied current reflects the ohmic drop contribution to the chronopotentiometric and reciprocal derivative chronopotentiometric potentialtime relationships (eqs 2 and 3, respectively),
E ) Eτ/4 - (RT/nF) ln[(τ1/2 - t1/2)/t1/2] + iRu
(2)
dt/dE ) 2nFτP/RT(1 + P)3+ iRu
(3)
where P ) exp(nF/RT)(E - Eτ/4). 12 Accordingly, larger potential shifts with the applied current (i.e., slopes of the Ep vs i plots) are observed for lower electrolyte concentrations. The shape of the response (for an ideal system), defined by the 2nFτP/RT(1 + P)3 term, is expected to be independent of the iRu drop. In practice, the peaks become sharper as the solution resistance increases. Such resistance effects do not compromise the analytical (12) Bi, S.; Yu, J. J. Electroanal. Chem. 1996, 405, 51.
3244
Analytical Chemistry, Vol. 72, No. 14, July 15, 2000
utility, in view of the well-defined chronopotentiograms observed in highly dilute electrolyte solutions and the nearly independence of the signal (A1/2) of the solution resistance. The exact reasons for the observed changes in peak shape are not fully clear. One reason may be the low support ratio (electrolyte/analyte). While the model analyte (ferrocene) is neutral, its reaction product is charged and will be transported away from the surface when the support ratio is lower than 10. Capacitance effects, which influence the portion of the applied current “going” to the faradaic process, should also not be ruled out. In summary, the experiments described above demonstrated that analytically useful data could be obtained in highly resistive solutions in connection to reciprocal derivative chronopotentiometry at electrodes of conventional size. The chronopotentiometric peak area were shown to be independent of the ohmic drop. Such behavior is attributed to the fundamentally different operational principles of chronopotentiometry. These findings promise to open resistive media and new environments to electrochemical studies at traditional electrode systems. ACKNOWLEDGMENT This work was supported by the U.S. DOE (Grant De-FG0796ER2306). Received for review February 28, 2000. Accepted May 8, 2000. AC000230W
