J. Phys. Chem. 1994, 98, 12410-12414
12410
Voltammetry in the Presence of Ultrasound. Sonovoltammetry and Surface Effects Richard G. Compton; John C. Eklund, Stephen D. Page, Giles H. W. Sanders, and Jonathan Booth Physical Chemistry Laboratory, Oxford University, South Parks Road, Oxford, OX1 3QZ United Kingdom Received: May 18, 1994@
The effect of power ultrasound in the modification of voltammetric processes is investigated and the role of surface activation by means of electrode erosiodroughening characterized. The potential benefits of sonovoltammetry are illustrated with two examples. First, the FladC potential for the passivation of nickel in air-saturated aqueous KOH solution is shown to be anodically shifted in the presence of ultrasound. Second, the voltammetric study of Cr(C0)6 in acetonitrile solution is shown to be feasible without the passivation observed under silent conditions where the electrolysis results in surface-active species the adsorption of which block further electron transfer.
Introduction
Transducer
Various workers have reported beneficial results from exposing electrochemical cells to the effects of power ultrasound. Areas of application include electroplating,1,2deposition of polymer film^,^,^ and electro~ynthesis.~~~ Several physical mechanisms exist by which ultrasound might modify electrode processes including (i) the enhancement of mass transport to and from the electrode surface resulting from cavitation in solution, (ii) the continuous activation of the electrode surface, (iii) the formation of ions, radicals, and other high-energy intermediates during transient cavitation, and (iv) the ultrasonic mediation of chemical processes associated with heterogeneous electron transfer steps. The first of these possible mechanisms was addressed in a previous paper7 which established a quantitative theoretical description of electrochemical mass transport in the presence of ultrasound, which was validated by a variety of experiments at both microelectrodes and electrodes of traditional dimensions.' The present paper seeks to identify and interpret the role of surface activation and/or cleaning as a mechanism by which ultrasound can control or modify the mechanism of electrode reactions. For this purpose, we make use of the wellthermostated cell shown in Figure 1, the transport properties of which have been characterized previously? and in which an ultrasonic horn is immersed in the solution of interest at a known distance above the working electrode under investigation. This arrangement is used to investigate the influence of ultrasound in the erosion and roughening of platinum and other electrode surfaces through interrogation by means of in situ ac impedance spectroscopy allied to ex situ atomic force microscopic characterization of sonicated surfaces. These effects are shown first to be advantageous in inhibiting the passivation of nickel surfaces at anodic potentials. Second, they are demonstrated to facilitate the study of reactive organometallic (in this study the one-electron oxidation of Cr(C0)6) and other species where surface passivation precludes meaningful voltammetric measurements in the absence of ultrasound. Experimental Section Three separate homs were employed in this work. These were supplied by Heat Systems (Model W380) and Sonics &
* Author to whom correspondence should be sent. @
Abstract published in Advance ACS Abstracts, October 15, 1994.
SONIC HORN SCE reference electrode
\
I To control unit of sonic horn
Pt 102 resistance
Copper cooling coil connected to thermostatted water bath
pt disc "macro" or "micro"
electrode
Figure 1. Sonoelectrochemicalcell used for the voltammetric and other studies described.
Materials (Models VC385 and VCX400) and had titaniumtipped horn probes (13 mm diameter) extended by 127 mm and operating at 20 kHz. Power levels up to and including 63 W cmW2were employed and calibrated calorimetrically according to the procedure of Mason et aL6 Thermostating of the electrochemical cell was accomplished by means of a copper cooling coil inserted in the solution through which water was circulated from a constant-temperature bath. By limitation of the sonication time to less than 1 min, this arrangement enabled the voltammetric measurements to be conducted at constant temperature (to within 2 "C). Platinum microdisk electrodes were obtained from Bioanalytical Systems (West Lafayette, IN) and had radii of between 2.5 (f0.3) and 60.0 (33.0) pm as calibrated electrochemically. Platinum macroelectrodes of radii between 0.05 (f0.002) and 0.39 (fO.O1) cm were mounted in an insulating Teflon sheath. All electrodes were carefully polished using diamond lapping compounds (Kemet, Kent, UK) of decreasing size down to 0.25 pm. A nickel electrode fabricated from a nickel wire (99.95% Ni) of 1 mm diameter and an active length of 1.3 cm was utilized in the FladC potential studies. The 1.3 cm length was held perpendicular to the axis of the horn at a distance of 30 (35)
0022-365419412098-12410$04.50/0 0 1994 American Chemical Society
Voltammetry in the Presence of Ultrasound
mm. The Ni electrode was initially smoothed down with fine emery paper, and then as with all electrodes, it was polished with diamond lapping compound to 0.25 pm. Platinum flags and A1 studs used in the atomic force microscopy (AFM) experiments were polished to 0.1 pm using diamond lapping compound of the appropriate size. Voltammetric and ac impedance measurements were carried out using a Solartron 1286 electrochemical interface (ECI) under computer control or an Oxford electrode potentiostat. Both of these operate in a conventional three-electrode mode. For ac measurements, the Solartron was used in combination with a 1250 frequency response analyzer (FRA). In all experiments, a carbon rod served as a counter electrode and a saturated calomel reference electrode was located close to the working electrode surface as shown in Figure 1. For ac impedance measurements, the FRA and ECI were controlled by a Viglen Genie Professional 4DX33 personal computer using a NI PC2.2 GBIB interface card and ZPLOT software package (Schlumberger Electronics Ltd.) for data acquisition and control of the impedance measurements. Microscopy experiments were performed on a SFM-BD2 Park Scientific Instruments (PSI, 1171 Borregas Ave., Sunnyvale, CA) atomic force microscope. The scan rate used to record the images taken was 1 Hz. The images reported were obtained using a microlever which was 100 p m long and 22 pm wide with an approximate force constant of 0.21 N m-l and a resonance frequency of about 66 kHz. The microlevers were fabricated from silicon nitride, and the pyrimidal tips had nominal radii of less than 400 A. Aqueous solutions were made up using Elgastat (High Wycombe, Bucks, UK) UHQ grade water, resistivity 18 MQ. The organic solvent used throughout was dried9 acetonitrile (Fisons, dried, distilled), and tetrabutylammonium perchlorate (TBAP; Kodak, puris) served as the background electrolyte. CrC06 (Fluka, '98%) and KOH (BDH, AnalaR) were used as received. Solutions were thoroughly purged of oxygen by bubbling through the solution argon that had been dried with calcium chloride and then, in the case of the organic solvent, presaturated with acetonitrile. Results and Discussion Initial attempts to identify possible effects of ultrasound on electrode surfaces utilized AC impedance spectroscopy. A platinum electrode (0.6 cm diameter) was polished smooth using a succession of finer diamond lapping compounds down to 0.10 pm in size and placed in the sonoelectrochemical cell shown in Figure 1. The latter contained 0.1 M TBAP/acetonitrile solution. The ac impedance spectrum was recorded in the frequency range 1-65 000 Hz first in the absence of ultrasound. For all impedance measurements, the cell was thermostated at 20 "C and the electrode was held at a potential of 0.0 (f0.2) V (vs SCE). The sonic horn, located 35 mm above the electrode surface to ensure repr~ducibility,~ was then switched on the electrode surface continuously exposed to power ultrasound (60 W cm-2) for an extended period of up to 15 min. The impedance spectrum was recorded at regular intervals throughout. Figure 2 shows a Cole-Cole plotlo obtained from an unsonicated electrode and one which had been exposed to ultrasound for a period of 420 s. The shape of these plots is typical of those observed in all experiments and is characterized by a tiny real (2') component and appreciable imaginary components such that the plot takes the form of a near vertical line. The latter would be that expected for a perfectly smooth electrode displaying capacitative behavior only.1° However, it
J. Phys. Chem., Vol. 98, No. 47, 1994 12411 1.03Hz
t
0000-
6000
.
IaIBefore Sono
-
1
1.93Hz
Z'/ohms Figure 2. Cole-Cole plots obtained at a platinum electrode located in 0.1 M TBAP/acetonitrile solution. In the two cases the electrode was (a) unsonicated and (b) exposed to power ultrasound (60 W cm-*) for 420 s.
0 .o
I
1.0
2.0
3.0
4.0
5.0
Log ,o(w/rads-l 1 Figure 3. Analysis of the data shown in Figure 2b according to eq 1. The good straight line obtained confiis that the effect of ultrasound is to roughen the electrode surface, increasing its fractal dimension D. The inferred value of D is 2.24.
may be seen that in reality the line deviates noticeably from the absolute vertical and this is consistent with the constant phase angle admittance expected under real conditions for a fractally rough electrode ~ u r f a c e . ~It~can - ~ be ~ shown that such a surface generates an interfacial impedance of the form
z' = dUfi)= where j = (-l)ll2, 5 is the frequency (rad s-l), and d is a frequency-independent real constant. The exponent a is fractional and for ideally polarizable electrodes lies between l/2 and 1. Thus, perfectly smooth electrodes are characterized by a = 1 and show ideal capacitative behavior. In general, a = [D - l]-l, where D is the effective fractal dimension of the surface (2 < D < 3),11-'14from which it may be seen that the case of D = 3 gives a = l/*, which is a result previously deduced for a porous e l e c t r ~ d e . ~ ~ . ~ ~ The impedance spectra measured as described above were analyzed in terms of eq 1 on the basis of which plots of log { -Z"} against log w should be linear and of slope a.17Figure 3 shows the data from Figure 2 analyzed in this manner. Good straight lines are seen and confirm the conceptual basis adopted for the interpretation. The slopes of plots such as Figure 3 permit the inference of the fractal dimension of the electrode
12412 J. Phys. Chem., Vol. 98, No. 47, 1994
jlBl 2.26
2.24 2.222.20
-
2.l0-m. b
&''"O
100 200 300 400 500 600 700 800 900 lo00
Time /s Figure 4. Evolution of the fractal dimension, D,of a platinum electrode surface in 0.1M TBAP/acetonitrile solution during exposure to ultrasound (power 60 W cm-2).
surface, D, and, in conjunction with the experiments reported above, how it varies with time after exposure to ultrasound of power 60 (35)W cm-* in acetonitrile/O.l M TBAP solution. Figure 4 shows that D starts reasonably close to 2 corresponding to a freshly polished electrode and increases steadily for a period of -10 min after sonication, whereafter it remains approximately steady. Note that the effect of the ultrasound is to roughen the electrode surface, increasing the value of D, which evolves to a steady value after a period of -10 min. We suggest that the roughening results from microjets of solution which are squirted onto the electrode due to the collapse of cavitation bubbles near the latter.**
Compton et al. Sonically induced surface roughening was further investigated using atomic force microscopy (AFM) for the ex situ imaging of platinum electrodes that had been exposed to ultrasound for varying periods of time. Polished (0.1 pm) platinum flags of approximate size 4 mm x 4 mm and rigidly mounted on sem sample studs before incorporation into the sonoelectrochemical cell were used for this exercise. Each flag was exposed to 60 W cm-* ultrasound from a horn located 35 ( f 5 ) mm from its surface in a medium of 0.1 M TBAP/acetonitrile. After exposure, the flags were removed from the solution, thoroughly rinsed with acetonitrile, and examined by AFM. Figure 5a shows an image of a platinum surface unexposed to ultrasound. It can be seen that it is smooth at the micrometer scale, which is as expected in the light of the electrode polishing procedure in which the finest diamond grit used was 0.1 pm. Figure 5b shows the effect of as little as 120 s of ultrasonic irradiation on the electrode surface from a horn positioned 30 mm away as in Figure 1. A pronounced roughening of the surface has occurred, and the platinum has become uneven at the 1 pm scale. Figure 5c shows that prolonging the exposure to 300 s gives an increased roughening. Yet further roughening occurs if the horn-electrode separation is reduced and the exposure time is increased, as can be inferred from Figure 5d. These direct observations are consistent with the inferences made from ac impedance spectroscopy that a platinum electrode surface is appreciably roughened after a few minutes of exposure to ultrasound. It might be expected that the scale and roughness of the sonicated surface would be material dependent. This anticipation is confirmed by Figure 6, which shows the exercise equivalent to that conducted above for platinum applied to the much softer metal aluminum. Figure 6a shows that the polished
Figure 5. AFM images taken ex situ showing (a, left top) a polished platinum electrode unexposed to ultrasound, after (b, left bottom) 120 and (c, right top) 300 s of irradiation (60 W cm-2) from a horn positioned 30 mm above the electrode surface, and (d, right bottom) after 600 s of irradiation (60 W cm-2) for a horn positioned 10 mm above the electrode surface.
J. Phys. Chem., Vol. 98, No. 47, 1994 12413
Voltammetry in the Presence of Ultrasound
I
200NA
tlk +O 80V
E / V (vs SCE 1
-080V
Figure 7. Voltammograms recorded for the oxidation of a nickel electrode in air-saturated aqueous KOH solution (a) in the absence and (b) in the presence of ultrasound. In both cases a voltage scan rate of 100 mV s-l was used.
Figure 6. AFM images taken ex situ showing (a, top) a polished aluminium electrode unexposed to ultrasound and (b, bottom) after 300 s of irradiation (60 W cm-*) from a horn positioned 30 mm above the electrode surface.
surface is smooth at the 0.1 pm scale while Figure 6b shows the roughening induced by 300 s of sonication (60 W cm-2) from a horn positioned 30 mm above the surface. The surface is seen to be much “spikier” than the corresponding platinum surface, and comparitatively more erosion of the metal has occurred. We now turn to a consideration of the effect of ultrasound on surface electrochemical processes and first examine the passivation of nickel electrodes. Air-saturated solutions containing 1.0 M KOH were used, and the nickel electrode was initially held at a potential of -0.6 V (vs SCE) at which no current due to metal oxidation flows. An anodic voltage scan of 100 mV s-l was then applied to the nickel electrode in the absence of ultrasound. The resulting cyclic voltammogram is depicted in Figure 7a, which shows first a rise in oxidation current as the metal oxidizes and then a rapid passivation of the electrode surface at potentials positive of the so-called Had6 p ~ t e n t i a l . ’ ~ This - ~ ~ passivation is thought to be caused by a thin surface film of oxygen atoms chemisorbed to the surface of the nickel.21 It can be seen from Figure 7a that the FladC potential under the conditions studied was +0.344 f 0.030 V (vs SCE) in agreement with previous studies.20 This value was found to be consistent for the scan rate range used (10-200 mV s-’) Figure 7b shows the corresponding experiment conducted with ultrasound (60 W cm-2) directed onto the electrode surface from a horn some 30-35 mm away. It can be seen that the magnitude of the current flowing is essentially unaltered by the ultrasound, c o d i n g the surface-bound nature of the oxidative process and the absence of mass transport control. However, the FladC potential is shifted anodically by
-60 mV in the presence of ultrasound (for all the voltage scan rates utilized). This shift dropped in magnitude on reduction of the ultrasonic power; at 40 mW cm-2 the shift was 20 mV and at 15 mW cm-2 it was 5 mV. When the ultrasound was switched off, the Had6 potential immediately returned to its original value, confirming that thermal effects were not responsible for the potential shift. The shift of the FladC potential to more positive potentials may be attributed to the cleaning or surface erosion effect of ultrasound. The latter causes cavitation in the solution. The collapse of cavitation bubbles causes jets of solution to be launched at the electrode with high kinetic energy, thus displacing species attached to its surface.l8 The greater the ultrasonic power the greater the expected displacement of species from the surface. As a result, a more positive potential is required to grow the passive film on the nickel electrode since the rate of film growth is required to be greater to compete with the electrode erosion process and so build up an insulating film. Finally we illustrate the potential benefits of sonic electrode cleaning and consider the oxidation of Cr(C0)6 in 0.1 M TBAP/ acetonitrile solution. This system has been extensively studied22 and is known to undergo one-electron oxidation at a potential of +1.53VZ2 (vs SCE) at platinum electrodes. However, the oxidation process is observed to passivate the electrode, and successive voltage scans reveal a current-voltage curve of steadily decreasing magnitude.22 Accordingly, investigations were made to see whether the irradiation of the electrode by ultrasound might clean the electrode so as to maintain its activity. Figure 8 shows typical results obtained at a platinum electrode of 3 mm radius. Ultrasonic irradiation of the electrode after the final scan depicted in Figure 8 for a period of 90 s using the experimental arrangement of Figure 1 with d = 30 mm restored the voltammetric response to that of the first voltage cycle. Measurements carried out using the same electrode with continuous ultrasonic irradiation showed substantially less passivation and, by using a microelectrode of radius 4.4 pm, under similar conditions passivation effects could be eliminated fully. Figure 9 shows results obtained in this
Compton et al.
12414 J. Phys. Chem., Vol. 98, No. 47, 1994
material to the electrode surface. Use of the mass transport model described elsewhere' enabled the inference of a value of 1.95 x cm2 s-l for the diffusion coefficient of Cr(C0)6.
scan
Conclusions The irradiation of electrode surfaces by power ultrasound has been shown to induce their erosion and roughening. This may be beneficially exploited to permit the study of (i) processes at electrodes under conditions that would otherwise lead to their passivation and (ii) the voltammetry of species that otherwise results in the generation of surface-active species which may poison the electrode activity. Acknowledgment. We thank SERC for a studentship for J.C.E., Zeneca (Fine Chemical Manufacturing Organisation) for the loan of their W380 ultrasonic horn, David Walton (Coventry University) for valuable discussions, and Marcus Browning for exciting our interest in high-frequency effects. References and Notes (1) (2) Books: (3) (4) 145. (5) 333.
7
1.av E I V b SCE)
1 ov
Figure 8. Cyclic voltammogram obtained at a platinum electrode (radius =)"3 for the oxidation of Cr(C0)6 (2.0 mM) in 0.1 TBAP/ acetonitrile solution using a voltage scan rate of 100 mV s-l. Progressive electrode passivation is seen as a result of successive potential cycles.
E I V ( v s SCE)
0.8
1.8
E/V(vs SCEl
0.8
Figure 9. Steady-state voltammogram measured at a microelectrode of radius 4.4 p m continuously irradiated with ultrasound (44 W cm-*) for the oxidation of Cr(C0)6 (0.32 mM) in 0.1 M TBAP/acetonitrile solution.
manner. Note that the form of the voltammogram is changed from that of Figure 8 and shows a clearly defined transportlimited plateau indicating a steady supply of electroactive
Walker, R. Chem. Br. 1990, 26, 251. Brown, B.; Goodman, J. E. High Intensity Ultrasonics; Liffe New York, 1965. Akbulut, U.; Toppare, L.; Yuntas, K. Polymer 1986, 27, 803. Oswana, S.;Ito, M.; Tanaka, K.; Kuwano, J. Synth. Met. 1987, 18, Mason, T. J.; Lorimer, J. P.; Walton, D. J. Ultrasonics 1990, 28,
(6) Chyla, A.; Lorimer, J. P.; Smith, G.; Walton, D. J. J . Chem. Soc., Chem. Commun. 1989, 603. (7) Compton, R. G.; Eklund, J. C.; Page, S. D.; Walton, D. J., submitted for publication. (8) Mason, T. J.; Lorimer, J. P.; Bates, D. M. Ultrasonics 1992, 30, 140. (9) Compton, R. G.; Coles, B. A,; Pilkington, M. B. G.; Bethell, D. J . Chem. SOC., Faraday Trans. 1 1990, 86, 663. (10) Brett, C. M. A,; Oliveira Brett, A. M. Electrochemistry. Principles, Methods and Applications; Oxford University Press: Oxford, UK, 1993; p 229. (11) Nykios, L.; Pajkossy, T. Electrochim. Acta 1985, 30, 1533. (12) Nykios, L.; Pajkossy, T. Electrochim. Acta 1986, 33, 1347. (13) Mulder, W. H.; Sluyters, J. H. Electrochim. Acta 1988, 33, 303. (14) Keddam, M.; Takenouti, H. Electrochim. Acta 1988, 33, 445. (15) de Levie, R. Electrochim. Acta 1964, 9, 1231. (16) de Levie, R. Electrochim. Acta 1965, 10, 113. (17) Compton, R. G.; Waller, A. M.; Block, H.; Chapples, J. J . Appl. Electrochem. 1990, 20, 23. (18) Mason, T. J.; Lorimer, J. P. Practical Sonochemistry; Ellis Hanvood: New York, 1988; p 28. (19) Bard, A. J.; Lund, H. Encyclopedia of the Electrochemistry of the Elements; Dekker: New York, 1979; Vol. 3, p 309. (20) Weininger, J. L.; Breiter, M. W. J . Electrochem. SOC. 1966, 113, 1133. (21) Reiger, P. H. Elecfrochemistry;Prentice Hall: New York, 1987; p 429. (22) Pickett, C. J.; Pletcher, D. J . Chem. Soc., Dalton Trans. 1975, 879.