Technical Note pubs.acs.org/ac
Atomic Force Microscopy of Electrochemical Nanoelectrodes Wojciech Nogala, Jeyavel Velmurugan, and Michael V. Mirkin* Department of Chemistry and Biochemistry, Queens College-CUNY, Flushing, New York 11367, United States S Supporting Information *
ABSTRACT: Nanometer-sized electrodes have recently been used to investigate important chemical and biological systems on the nanoscale. Although nanoelectrodes offer a number of advantages over macroscopic electrochemical probes, visualization of their surfaces remains challenging. Thus, the interpretation of the electrochemical response relies on assumptions about the electrode shape and size prior to the experiment and the changes induced by surface reactions (e.g., electrodeposition). In this paper, we present first AFM images of nanoelectrodes, which provide detailed and unambiguous information about the electrode geometry. The effects of polishing and cleaning nanoelectrodes are investigated, and AFM results are compared to those obtained by voltammetry and SEM. In situ AFM is potentially useful for monitoring surface reactions at nanoelectrodes.
T
disk radius. Estimating a from eq 1 is problematic because steady-state voltammograms provide no information about nanoelectrode geometry, which is never perfect (see below) and has to be checked independently. The visualization issue is even more pressing in studies of surface reactions at nanoelectrodes. For instance, to investigate nucleation/growth processes, one has to relate the current response to both the initial electrode size/geometry and the change resulting from electrodeposition of metal.14 Without a means for independent verification, such analysis can be ambiguous. The size and geometry of a nanoelectrode can be evaluated by using it as a probe in the scanning electrochemical microscope (SECM).15 While high SECM feedback (e.g., the tip current increasing by the factor of ∼10 near the conductive surface) can provide strong evidence that a nanoelectrode is essentially flat and well polished,7,11 a lower feedback often observed in current-distance curves is hard to interpret. Possible origins of such a response include either recessed or protruding tip geometry, surface contamination, or poor tip/ substrate alignment. One should also notice that only a sharp nanoelectrode with a very thin insulating sheath can be used as an SECM probe. In this Technical Note we present a direct method for visualizing nanoelectrode surfaces by atomic force microscopy (AFM).
he development of nanometer-sized electrodes made it possible to study processes and phenomena that would not be accessible by larger electrochemical probes.1 Nanoelectrodes were used to study electrochemistry of single molecules2 and single nanoparticles,3 investigate mass transport processes on the nanoscale,4 measure rapid kinetics,5−7 and perform quantitative experiments inside living cells.8 Several unusual phenomena such as electrochemistry through glass,9 surface diffusion of adsorbed redox species at the Pt/glass interface,10 and the effects of a partially formed electrical double layer11 could only be observed in nanoscale systems. Experiments at nanoelectrodes are often hindered by visualization difficulties. The knowledge of the electrode shape and size is essential for quantitative experiments; and significant shape irregularities (e.g., the recession of the conductive surface into the surrounding insulator) may cause very large errors in the determined kinetic parameters.12 Neither optical nor electron microscopy provide adequate means for visualization of nanoelectrodes. The SEM resolution is not sufficiently high to characterize electrodes smaller than ∼50 nm radius; moreover, insufficient z-axis resolution makes it hard to distinguish between flat, recessed, and protruding nanoelectrodes. TEM can provide a side view of a very small (e.g., ≤3 nm radius13) electrode, but it gives no information about the exposed metal surface. Additionally, it is not easy to use either SEM or TEM for in situ monitoring of electrochemical processes. In most publications on nanoelectrochemistry, a nanoelectrode was assumed to be disk-shaped and flush with the surface of surrounding insulator, and its radius (a) was evaluated from the diffusion limiting current id = 4nFDc*a
■
EXPERIMENTAL SECTION Chemicals. Ferrocenemethanol (FcCH2OH, 97%) from Aldrich (Milwaukee, WI) was sublimed before use. Other chemicals were used as received. Aqueous solutions were prepared from deionized water (Milli-Q, Millipore Co.). Electrodes and Electrochemical Cells. Pt and Au nanoelectrodes were fabricated as described previously.7 Briefly,
(1)
where n is the number of transferred electrons, F is the Faraday constant, D and c* are the diffusion coefficient and the bulk concentration of the redox species, respectively, and a is the © 2012 American Chemical Society
Received: March 15, 2012 Accepted: May 17, 2012 Published: June 6, 2012 5192
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
Analytical Chemistry
Technical Note
an annealed 25-μm Pt or Au wire (Goodfellow) was pulled into a glass capillary under vacuum with the help of a Sutter P-2000/ G laser pipet puller. The pulled electrodes were polished either on 50 nm alumina lapping tape or 100-nm diamond lapping tape (Precision Surfaces International) under video microscopic control. Pt wires were sealed into borosilicate glass and Au wires into soda-lime glass capillaries (Drummond; 1.0-mm o.d., 0.2-mm i.d.). Electrochemical AFM images and voltammograms were obtained in a commercial liquid cell (Park Systems), which was mounted on the stage of XE-120 scanning probe microscope, using an EI-400 potentiostat (Ensman Instruments). In a two-electrode setup, a Ag-wire was used as a quasi-reference. Current images were obtained with the nanoelectrode biased to a potential corresponding to diffusion-limited oxidation of FcCH2OH, and the current was plotted as a function of lateral position of the AFM probe. SECM measurements were performed using a previously described home-built SECM instrument.7,8 AFM Imaging. An XE-120 scanning probe microscope (Park Systems) was employed to image the nanoelectrodes in a noncontact, intermittent, or contact mode, either in air or in solution; and PPP-NCHR and PPP-CONTSCR AFM probes (Nanosensors) were used for noncontact and contact imaging, respectively. An ∼4.5 cm long nanoelectrode was mounted vertically with its polished surface facing the AFM probe using a homemade sample holder (Figure 1) attached to an XY piezo positioning stage. The cantilever was positioned above the nanoelectrode with the help of an optical microscope. A different custom-made holder was used for liquid phase imaging. In a noncontact mode, the tip was brought within a close proximity of the sample using the approach function, and then the nanoelectrode was moved laterally in 200 nm steps to
bring the AFM probe to its apex. (In XE-120 the tip travels along z-axis, and the sample is moved in the x−y plane). The travel direction was selected to affect z-axis retraction of the piezo actuator in a close-loop mode. This corresponded to sliding of the slanted tip surface along the edge of the glass insulating sheath of the electrode. When the piezo approached its upper limit, the z-stage motor was retracted by 1 μm to maintain the z-axis piezo actuator within its range (12 μm). This approach allows imaging of sharp objects, including needle-like SECM tips with a nanometer-scale thickness of the insulating sheath at the tip.
■
RESULTS AND DISCUSSION Noncontact topography imaging is very convenient for preliminary characterization of nanoelectrode geometry. An image (Figure 2) was obtained in air with no direct contact
Figure 2. Noncontact topographic image of a polished Pt nanoelectrode in air (A) and a steady-state voltammogram of 1 mM FcCH2OH obtained at the same electrode in 0.2 M KCl solution (B). (A) The scan rate was 0.5 Hz. The red line corresponds to the shown cross-section. (B) The potential sweep rate was v = 50 mV s−1.
between the AFM probe and the nanoelectrode and, thus, no possibility of damage or contamination of the sample surface. The electrode in Figure 2A has a 25−30 nm effective radius and appears to be ∼5 nm recessed into the glass insulator. The steady-state diffusion limited current to such an electrode can be calculated from eq 216
Figure 1. Experimental setup used for AFM imaging of nanoelectrodes in air. (A) A glass-sealed, polished nanoelectrode (1) is positioned under the AFM probe (2). (B) A scheme of the nanoelectrode holder: (1) brass base, (2) plastic screw, (3) nanoelectrode, and (4) XY piezo positioning stage.
il = id /(1.0354 + 1.2621H + 0.01155 ln H ) 5193
(2)
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
Analytical Chemistry
Technical Note
where H is the recess depth normalized by the effective disk radius (l/a) and id is the diffusion limiting current to the nonrecessed electrode of the same radius given by eq 1. Using a = 28 nm, l = 5 nm, D = 7.6 × 10 −6 cm 2 /s for ferrocenemethanol (FcCH2OH),7a and c* = 1 mM, one obtains from eq 2 il = 6.8 pA, which is very close to the diffusion limiting current value in Figure 2B (∼6.7 pA). The triangular shape of the cross-section in Figure 2A is due to the convolution of the tip shape and the sample geometry. It is possible that the tip did not reach the bottom of the cavity, and its actual depth could be somewhat larger. However, a good agreement between the AFM and voltammetric results indicates that noncontact imaging provides a reliable estimate for the effective recess depth. This slightly recessed electrode (H = 5 nm/28 nm = 0.18) is suitable for quantitative nanoelectrochemical experiments. However, some errors could be expected if it was used for measuring electron transfer kinetics because the current distribution near the edge of the conductive surface at the recessed electrode is significantly more uniform than at the inlaid disk. One should notice that no existing electrochemical or microscopic technique could provide equally detailed information about nanoelectrode geometry. An image of a considerably recessed electrode is shown in Figure 3A. The diffusion limiting current (∼9 pA) measured in 1.2 mM FcCH2OH (Figure 3C) is in good agreement with il = 9.4 pA calculated from eq 2 with l ≈ 40 nm, a ≈ 52 nm found from Figure 3B. From the voltammogram in Figure 3C, one would not be able to tell that this electrode is recessed. The effective radius calculated from Figure 3C without taking into account the recessed geometry would have been as small as 20 nm. Moreover, kinetic experiments (and other geometrysensitive experiments) at such an electrode could yield misleading results. Recessed nanoelectrodes have been employed for different types of experiments, including measurements in ultrasmall volumes.11 AFM characterization of recessed probes can greatly improve the reliability of such experiments. In the case of a deeply recessed nanoelectrode (i.e., l > a), a sharper AFM probe (e.g., a carbon nanotube probe) should be used for more accurate measurement of l. An electrode with the conductive core protruding from the glass sheath can also be characterized by AFM. The electrode imaged in Figure 4A has a ≈ 83 nm and height h ≈ 20 nm. The glass roughness in this case is relatively high, and to clearly distinguish between the conductive Pt surface and the surrounding glass one can compare the topographic image (Figure 4A) to the phase shift image of the same electrode (Figure 4C) obtained in the intermittent contact mode. The contrast in the latter is due to different interactions of the AFM probe with Pt and glass surfaces. Contact mode imaging (Figure S1A in Supporting Information) provides better resolution than the noncontact mode, but positioning a more flexible AFM contact probe above the nanoelectrode apex is more difficult. The imaging set point has to be selected carefully to avoid scratching the electrode surface by the sharp AFM tip. By comparing a lateral force image recorded in the contact mode (Figure S1B in the Supporting Information) to the corresponding topographic image (Figure S1A in the Supporting Information), one can validate the shape and position of the electrode surface due to the difference between friction coefficients of metal and glass. Alternatively, a conductive AFM probe may be used for distinguishing between the metal surface and insulating glass (Figure S2 in the Supporting Information). However, even with a very low bias,
Figure 3. Noncontact topographic image of a recessed Pt nanoelectrode in air in 3D (A) and 2D (B) and a steady-state voltammogram of 1.2 mM FcCH2OH (C). For other parameters, see Figure 2.
conducting AFM images of nanoelectrodes are affected by higher noise. An essentially “perfect” flat electrode whose conductive surface is flush with the surface of the surrounding insulator can be hard to visualize by noncontact AFM. A well-polished electrode in Figure 5 exhibits extremely low roughness (∼1 nm over a 1 μm2 surface area) of both Pt surface and glass. The electrode active area is undetectable in the topographic image (Figure 5A), whereas the voltammogram in Figure 5C yields 5194
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
Analytical Chemistry
Technical Note
AFM tip. The effective radius value, a = 37 nm, obtained from Figure 5C is reliable because the electrode is flat and flush with the insulator surface (Figure 5A) and its surface reactivity is uniform (Figure 5B). One should also notice that the topography of a nanoelectrode can be imaged in solution before and immediately after an electrochemical experiment, which is essential for visualization of nanoelectrochemical processes (e.g., electrodeposition of metals at nanoelectrodes18). Figure 6 shows an example of characterization of the same nanoelectrode by AFM and SECM. The radius value extracted from the SECM approach curve (Figure 6A), a = 47.5 nm, was essentially the same as that calculated from the diffusion limiting current (48.2 nm). However, this electrode could not be brought sufficiently close to the substrate surface (the experimental curve in Figure 6A deviated from the theory at shorter separation distances, d < ∼1.2a) for reliable evaluation of its geometry. The difficulties in bringing the tip closer to the substrate could be caused by different problems, including a recessed Pt surface, tip/substrate misalignment, large RG (i.e., the ratio of glass radius to that of the Pt tip), and the contamination of either the Pt or glass surface by relatively large particles. The topographic AFM image of the same electrode (Figure 6B) confirms the radius value determined by SECM and shows that the Pt surface is essentially flush with the flat surrounding glass (the recess depth a) and a relatively large physical size of the 5195
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
Analytical Chemistry
Technical Note
Figure 5. Topography (A) and substrate current (B) AFM images recorded simultaneously during noncontact imaging of a Pt nanoelectrode (a ≈ 37 nm) with cross sections shown below images for fast scan axis direction. Scan rate: 0.5 Hz. Cyclic voltammogram (C) of the sample at 50 mV s−1. AFM images and voltammogram recorded in 1.2 mM FcCH2OH in 0.2 M KCl.
Information). Moreover, a = 50 nm was obtained by fitting an SECM approach curve to the theory (Figure S4D in the Supporting Information). This set of data is consistent with the assumption that the electrode surface was cleaned by annealing. However, Figure 4 suggests a different explanation of the annealing effect. Before annealing, the metal surface was flat and indistinguishable from surrounding glass in the topographic image (not shown), and a = 83 was calculated from the diffusion limiting current (curve 1 in Figure 4B). The annealing of this electrode for 1 h at 120 °C resulted in the increased current (curve 2 in Figure 4B), essentially unchanged radius (∼80 nm) and marked protrusion (∼20 nm) of Pt (Figure 4A) confirmed by the phase-shift image (Figure 4C). The increase in diffusion current can be attributed to the protrusion effect.21 SEM images in Figure S4 in the Supporting Information do not show this effect because of insufficient z-axis resolution. Longer annealing of a nanoelectrode can cause dramatic changes in its geometry. For instance, the annealing of an essentially flat 15-nm-radius Pt nanoelectrode (Figure S5A in the Supporting Information) for 18 h at 120 °C resulted in the formation of a large spheroidal conductive structure, with the radius ∼5 times that of the original electrode (Figure S5B in the Supporting Information). Although nanometer-sized metal particles melt at significantly lower temperatures than the melting point of the bulk metal,22 it is surprising that Pt could be extruded from glass and reshaped at such a low temperature. Nevertheless, the agreement between the size of the annealed electrode and the diffusion current (which increased by the factor of ∼7 after annealing; cf. parts C and D of S5 in the Supporting Information) suggests that the formed structure is metallic. Unlike Pt, the annealing of Au nanoelectrodes causes the metal surface to recede into glass. In Figure S6 in the Supporting Information, a polished Au electrode essentially flat before annealing (Figure S6A in the Supporting Information) became ∼30 nm recessed after 1 h annealing at 150 °C (Figure S6B in the Supporting Information), and ∼50 nm recessed after further annealing at 180 °C for 1 h (Figure S6C in the Supporting Information). The decrease in the diffusion limiting current to this electrode upon annealing (Figure S6D in the Supporting Information, curves 2 and 3) confirms Au surface receding. Overall, annealing does not seem to be a good way of cleaning either Pt or Au nanoelectrodes. In contrast, immersing
Figure 6. (A) Experimental SECM current-distance curve (symbols) obtained with a polished Pt nanoelectrode approaching an insulating glass substrate and theoretical curve (solid line19) calculated for a = 47.5 nm and RG = 19. The current is normalized by id = 29 pA. (B) Noncontact topographic image of the same electrode in air. The red line corresponds to the shown cross-section. The white spots are small particles present on the surface.
5196
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
Analytical Chemistry
Technical Note
(12) Baranski, A. S. J. Electroanal. Chem. 1991, 307, 287. (13) Li, Y.; Bergman, D.; Zhang, B. Anal. Chem. 2009, 81, 5496. (14) Chen, S.; Kucernak, A. J. Phys. Chem. B 2003, 107, 8392. (15) Mirkin, M. V.; Fan, F.-R. F.; Bard, A. J. J. Electroanal. Chem. 1992, 328, 47. (16) Sun, P.; Mirkin, M. V. Anal. Chem. 2007, 79, 5809. (17) Burt, D. P.; Wilson, N. R.; Janus, U.; Macpherson, J. V.; Unwin, P. R. Langmuir 2008, 24, 12867. (18) Wang, Y.; Noël, J.-M.; Velmurugan, J.; Nogala, W.; Mirkin, M. V.; Lu, C.; Guille Collignon, M.; Lemaître, F.; Amatore, C. Proc. Natl. Acad. Sci. U.S.A. 2012, DOI: doi:10.1073/pnas.1201552109. (19) Cornut, R.; Lefrou, C. J. Electroanal. Chem. 2007, 608, 59. (20) Agyekum, I.; Nimley, C.; Yang, C.; Sun, P. J. Phys. Chem. C 2010, 114, 14970. (21) Myland, J. C.; Oldham, K. B. J. Electroanal. Chem. 1990, 288, 1. (22) Couchman, P. R.; Jesser, W. A. Nature 1977, 269, 481.
a nanoelectrode in piranha solution (i.e., volumetric 3:1 concentrated sulfuric acid to 30% hydrogen peroxide solution. Caution! This solution is a very strong oxidizing agent and very dangerous to handle in the laboratory. Protective equipment including gloves, goggles, and face shields should be used at all times) usually improves its response with no apparent change in geometry detected in our AFM images. Another typical problem in nanoelectrochemical experiments is solution leakage through the cracks in the insulating sheath. Such cracks can be detected in noncontact AFM images (Figure S7 in the Supporting Information) In summary, AFM imaging of a nanoelectrode in air and in solution can provide detailed information about its geometry and surface reactivity that would be hard to obtain by any other technique. This information is essential for reliable interpretation of nanoelectrochemical experimental data. The nanoelectrodes characterized by AFM are not damaged; they can be employed in electrochemical experiments and as SECM probes. Our experiments also revealed surprising effects of the lowtemperature annealing, which can result in the recession of the Au electrode surface into glass, protrusion of Pt nanoelectrodes, and formation of nanoscale surface structures. The development of methodologies for in situ AFM monitoring of surface reactions on nanoelectrodes, including nucleation/growth of metal nanoclusters,18 is currently underway in our laboratory.
■
ASSOCIATED CONTENT
S Supporting Information *
AFM and SEM images of and voltammograms obtained at nanoelectrodes. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The support of this work by the National Science Foundation (Grants CHE-0957313 and CHE-1026582) is gratefully acknowledged. We thank Dr. Jean-Marc Noël for the conductive AFM images (Figure S2 in Supporting Information).
■
REFERENCES
(1) Murray, R. W. Chem. Rev. 2008, 108, 2688. (2) Fan, F.- R. F.; Bard, A. J. Science 1995, 267, 364. (3) Li, Y.; Cox, J. T.; Zhang, B. J. Am. Chem. Soc. 2010, 132, 3047. (4) Krapf, D.; Quinn, B. M.; Wu, M.-Y.; Zandbergen, H. W.; Dekker, C.; Lemay, S. G. Nano Lett. 2006, 6, 2531. (5) Penner, R. M.; Heben, M. J.; Longin, T. L.; Lewis, N. S. Science 1990, 250, 1118. (6) Watkins, J. J.; Chen, J.; White, H. S.; Abruna, H. D.; Maisonhaute, E.; Amatore, C. Anal. Chem. 2003, 75, 3962. (7) (a) Sun, P.; Mirkin, M. V. Anal. Chem. 2006, 78, 6526. (b) Velmurugan, J.; Sun, P.; Mirkin, M. V. J. Phys. Chem. C 2009, 113, 459. (8) Sun, P.; Laforge, F. O.; Abeyweera, T. P.; Rotenberg, S. A.; Carpino, J.; Mirkin, M. V. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 443. (9) Velmurugan, J.; Zhan, D.; Mirkin, M. V. Nat. Chem. 2010, 2, 498. (10) Zhan, D.; Velmurugan, J.; Mirkin, M. V. J. Am. Chem. Soc. 2009, 131, 14756. (11) Sun, P.; Mirkin, M. V. J. Am. Chem. Soc. 2008, 130, 8241. 5197
dx.doi.org/10.1021/ac300744t | Anal. Chem. 2012, 84, 5192−5197
