Characterization of Batch-Microfabricated Scanning Electrochemical

Anal. Chem. 2005, 77, 424-434

Characterization of Batch-Microfabricated Scanning Electrochemical-Atomic Force Microscopy Probes Phillip S. Dobson and John M. R. Weaver

Department of Electronic and Electrical Engineering, Glasgow University, Glasgow, G12 8LT Mark N. Holder, Patrick R. Unwin, and Julie V. Macpherson*

Department of Chemistry, University of Warwick, Coventry CV4 7AL

A procedure for the batch microfabrication of scanning electrochemical-atomic force microscopy (SECM-AFM) probes is described. The process yields sharp AFM tips, incorporating a triangular-shaped electrode (base width 1 µm, height 0.65 µm) at the apex. Microfabrication was typically carried out on 1/4 3-in. wafers, yielding 60 probes in each run. The measured spring constant of the probes was in the range 1-1.5 N m-1. To date, processing has been carried out twice successfully, with an estimated success rate for the fabrication process in excess of 80%, based on field emission-scanning electron microscopy imaging of all probes and current-voltage measurements on a random selection of ∼30 probes. Steady-state voltammetric measurements for the reduction of Ru(NH3)63+ in aqueous solution indicate that the electrode response is well-defined, reproducible, and quantitative, based on a comparison of the experimental diffusion-limited current with finite element simulations of the corresponding mass transport (diffusion) problem. Topographical imaging of a sputtered Au film with the SECM-AFM probes demonstrates lateral resolution comparable to that of conventional Si3N4 AFM probes. Combined electrochemicaltopographical imaging studies have been carried out on two model substrates: a 10-µm-diameter disk ultramicroelectrode (UME) and an array of 1-µm-diameter UMEs, spaced 12.5 µm apart (center to center). In both cases, an SECM-AFM probe was first employed to image the topography of the substrates. The tip was then moved back a defined distance from the surface and use to detect Ru(NH3)62+ produced at the substrate, biased at a potential to reduce Ru(NH3)63+, present in bulk solution, at a diffusion-controlled rate (substrate generation-tip collection mode). These studies establish the success of the batch process for the mass microfabrication of SECMAFM tips. The atomic force microscope (AFM) has revolutionized surface imaging, enabling the visualization of surface topography at * To whom correspondence should be addressed. E-mail: j.macpherson@ warwick.ac.uk.

424 Analytical Chemistry, Vol. 77, No. 2, January 15, 2005

ultrahigh resolution (typically approaching the nanometer level), for a wide range of surfaces in a variety of different media, including air, vacuum, and liquid.1 The AFM probe consists of a sharp tip attached to a force-sensing cantilever. For imaging, the tip scans across the sample, and the cantilever deflects in response to force interactions between the tip and substrate. The force is generally used as a feedback mechanism to maintain a fixed tipsample separation. In this way, the AFM is able to topographically image surfaces irrespective of their conductivity, in contrast to scanning tunneling microscopy, which is limited to conducting or semiconducting samples. In its conventional form, AFM lacks chemical specificity. Molecular recognition becomes possible by tailoring the chemical functionality of the probe, the substrate, or both in a technique termed chemical force microscopy (CFM).2 However, CFM is limited in that a particular tip can usually only be employed to identify a particular surface chemistry or distinguish between two distinct types of chemical character. It is increasingly recognized that scanned probe microscopy techniques based on electrochemical principlessspecifically scanning electrochemical microscopy3shave strengths in imaging that AFM lacks.4 Namely, electrochemistry is an attractive tool for the chemical identification of electroactive ions and molecules close to surfaces and interfaces. For example, through the use of amperometry, the potential of a probe electrode can be tuned to detect a particular species of interest and measure the resulting concentration quantitatively, by virtue of the current flow.5,6 Additionally, amperometry provides a means to rapidly alter the chemical composition of a solution adjacent to a surface.7,8,9 Recent progress has thus seen the integration of submicrometer-sized (1) (a) Binnig, G.; Quate, C. F.; Gerber, Ch. Phys. Rev. Letts. 1986, 56, 930. (b) Rugar, D.; Hansma, P. K. Phys. Today 1990, 43, 23. (c) Wiesendanger, R. Scanning Probe Microscopy and Spectroscopy; Cambridge University Press: Cambridge, U.K., 1994. (2) (a) Noy, A.; Vezenov, D. V.; Lieber, C. M. Annu. Rev. Mater. Sci. 1997, 27, 381. (b) Frisbie, C. D.; Rosnyai, L. F.; Noy A., Wrighton, M. S.; Lieber, C. M. Science 1994, 265, 2071. (c) Lee, G. U.; Chrisey, L. A.; Colton, R. J. Science 1994, 266, 771. (3) Bard, A. J., Mirkin, M. V., Eds. Scanning Electrochemical Microscopy; Marcel Dekker: New York, 2001. (4) Hansma, H. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 14678. (5) Basame, S. B.; White, H. S. J. Phys. Chem. B 1998, 102, 9812. (6) Scott, E. R.; White, H. S.; Phipps, J. B. Anal. Chem. 1993, 65, 1537. 10.1021/ac048930e CCC: $30.25

© 2005 American Chemical Society Published on Web 12/08/2004

electrodes in AFM probes, a technique that has been termed combined scanning electrochemical microscopy (SECM)-AFM.10 SECM-AFM has many recognized advantages over SECM. For example, in conventional imaging form, SECM does not employ a feedback mechanism to maintain a constant tip-sample separation. Thus, the response of the probe ultramicroelectrode (UME), used as the imaging tip, depends on both the substrate (electro)activity and topography; separating these two contributions to an SECM image can be difficult. Moreover, the topographical and electrochemical resolution of SECM is governed by the size of the probe, which is often in the micrometer range or sometimes greater. In contrast, the dual-sensing capabilities of an SECMAFM probe means that it is possible to independently determine topography (via the AFM component) and surface activity (via the electrochemical component). Thus, SECM-AFM offers the possibility to both (i) correlate surface activity with surface topography and (ii) change the local solution conditions electrochemically and examine the effect on surface structure. Such measurements are possible at much higher spatial resolution than attainable with SECM alone. We note that the combination of SECM with shear force measurements11 also offers the possibility of dual topography-activity measurements, but, to date, the topographic imaging capabilities do not match those of SECMAFM. Several different approaches have been adopted to fabricate SECM-AFM probes. The simplest approach is to coat conventional Si or Si3N4 probes with Pt or Au to make the tip and cantilever electroactive. Such probes have been used to induce dissolution reactions, via an electrochemical perturbation of the solution conditions, and to monitor the resulting change in surface topography,12-14 as well as to probe diffusion through individual nanoscale pores (100-nm diameter) in a membrane.15 To improve the spatial resolution of the electrochemical component of SECMAFM probes, initial work focused on a hand-fabrication procedure whereby a 50-µm-diameter Pt microwire was flattened (to form the cantilever), etched (to form the tip), and electrically insulated at all but the apex of the tip.16 This resulted in the formation of cone-shaped tip electrodes with a characteristic electrode dimension in the range 10 nm to 1 µm. Au SECM-AFM probes of conical or spherical tip geometry have also been reported, using modifications to this procedure.17 Given that the tip is electrically conducting it is also possible to use these probes for conducting-AFM measurements (C-AFM), in air or under solution, with the added advantage that there is no loss in tip conductivity from tip wear.18,19 (7) (a) Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1994, 98, 1704. (b) Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1995, 99, 3338. (8) Yatziv, Y.; Turyan, I.; Mandler, D. J. Am. Chem. Soc. 2002, 124, 5618. (9) Wipf, D. O. Colloid Surf. A 1994, 93, 251. (10) Gardner, C. E.; Macpherson, J. V. Anal. Chem. 2002, 74, 576A. (11) (a) Hengstenberg, A.; Kranz, C.; Schuhmann, W.; Chem. Eur. J. 2000, 6, 1547. (b) Ballesteros Katemann, B.; Schulte, A.; Schuhmann, W. Electroanalysis 2004, 16, 60. (12) Macpherson, J. V.; Unwin, P. R.; Hillier, A. C.; Bard, A. J. J. Am. Chem. Soc. 1996, 118, 6445. (13) Jones, C. E.; Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. B 2000, 104, 2351. (14) Jones, C. E.; Unwin, P. R.; Macpherson, J. V. Chem. Phys. Chem. 2003, 4, 139. (15) Macpherson, J. V.; Jones, C. E.; Barker, A. L.; Unwin, P. R. Anal. Chem. 2002, 74, 1841. (16) Macpherson, J. V.; Unwin, P. R. Anal. Chem. 2000, 72, 276-285. (17) Abbou, J.; Demaille, C.; Druet, M.; Moiroux, J. Anal. Chem. 2002, 74, 6355.

Electrochemical imaging with hand-fabricated probes can be carried out in contact mode, at fixed height,20 or at constant distance from the substrate surface, i.e., following the underlying surface contours (via lift-mode imaging).20 Although the nature of the fabrication process results in some variability between probes, in terms of electrode geometry, size, and physical characteristics, these tips have proved useful in several applications including the following: active-site identification and quantification;20 in situ modification and visualization of crystal surfaces;16 and correlating the surface structure of a membrane with the underlying transport activity.16 For more widespread applications of SECM-AFM, tip preparation methods are required that yield probes with reproducible and defined characteristics. Microfabrication technologies represent an attractive way to address these issues, with the possibility of scaling up the preparation procedure. Kranz and co-workers have made progress using focused ion beam (FIB) methods to create a defined electrode structure in an AFM tip.21 In this approach, a conventional Si3N4 AFM probe was first sputter-coated with a 100300-nm-thick layer of Au, followed by ∼700-800-nm-thick layer of insulating Si3N4, SiO2 or parylene. Successive millings of the probe, using FIB, were employed to redefine and sharpen the geometry of the AFM tip, resulting in the formation of either a submicrometer ring or square-framed electrode, around the AFM tip. With this design, the electrode images at a constant distance from the substrate. A similar geometry has been adopted for the fabrication of pH-sensitive iridium oxide electrodes situated around the base of STM tips.22 Imaging with the Kranz probes can be achieved using both contact and tapping modes under fluid.21,23,24 However, each tip is fabricated individually using the FIB procedure, and to date, current images obtained with such probes have not been analyzed quantitatively. In this paper, we describe the use of direct write electron beam lithography (EBL) to define both the shape of the AFM probe25 and geometry of the integrated electrode, at the wafer scale. Using EBL, a triangular-shaped gold electrode, ∼1 µm base width and 0.65-µm height, is positioned at the apex of the lithographically defined AFM tip. For the purpose of this paper, we have focused on this one electrode design, although many others are possible using the EBL approach. Diffusional mass transport to the triangular electrode in bulk solution has been simulated using the finite element modeling package FEMLAB, and the resulting limiting currents have been found to be in excellent agreement with those observed experimentally for reduction of Ru(NH3)63+ in aqueous solution (with excess supporting electrolyte). The effectiveness of the probe in quantitatively correlating surface structure with electroactivity is demonstrated through generation(18) Macpherson, J. V.; Gueneau de Mussy, J. P.; Delplancke, J. L. Electrochem. Solid-State Lett. 2001, 4, E33. (19) Boxley, C.; White, H. S.; Gardner, C. E.; Macpherson, J. V. J. Phys. Chem. B 2003, 107, 379. (20) Macpherson, J. V.; Unwin, P. R. Anal. Chem. 2001, 73, 550. (21) Kranz, C.; Friedbacher, G.; Mizaikoff, B.; Lugstein, A.; Bertagnolli, E. Anal. Chem. 2001, 73, 2491. (22) Amman, E.; Beuret, C.; Indermuhle, P. F.; Kotz, R.; de Rooij, N. F.; Siegenthaler, H. Electrochim. Acta 2001, 47, 327. (23) Kueng, A.; Kranz, C.; Lugstein, A.; Bertagnolli, E.; Mizaikoff, B. Angew. Chem., Int. Ed. 2003, 42, 3238. (24) Kueng, A.; Kranz, C.; Mizaikoff, B.; Lugstein, A.; Bertagnolli, E. Appl. Phys. Lett. 2003, 82, 1592. (25) Zhou, H.; Mills, G.; Chong, B. K.; Midha, A.; Donaldson, L.; Weaver, J. M. R. J. Vac. Sci. Technol. A 1999, 17, 2233.

Analytical Chemistry, Vol. 77, No. 2, January 15, 2005

425

Figure 1. Schematic illustrations of (a) the dimensions of the SECM-AFM cantilever and tip (in micrometers), (b) the geometry of the triangularshaped electrode in the vicinity of the tip (not to scale), and (c) low- and (d) high-resolution FE-SEM images of the tip and cantilever geometry of a typical SECM-AFM probe.

collection imaging experiments with 10- and 1-µm-diameter disk electrodes as the substrate. EXPERIMENTAL SECTION SECM-AFM Probes. The microfabrication procedure adopted for the production of SECM-AFM probes is outlined in detail elsewhere.26,27 Microfabrication was typically carried out on 1/4 3-in. wafers, yielding 60 probes. The cantilever and tip were formed from a 500-nm-thick plasma enhanced chemical vapor deposition (PECVD) layer of Si3N4, yielding the tip and cantilever geometry shown in Figure 1a. EBL was used to define the Au contact pad, mirror, and contact lines (100 nm thick), which ran from the contact pad, situated on the main body of the probe, up to the tip of the cantilever. In the vicinity of the tip, the metal lines had a width of 1 µm. A further layer of PECVD Si3N4, 75 nm thick, was employed to insulate the Au lines. Exposure of the Au tip electrodesa triangular-shaped electrode, 1 µm in base width and (26) Dobson, P. S.; Holder, M. N.; Unwin, P. R.; Macpherson, J. V.; Weaver, J. M. R. In preparation. (27) Mills, G.; Zhou, H.; Midha, A.; Donaldson, L.; Weaver, J. M. R. Appl. Phys. Lett. 1998, 72, 2900.

426


0.65 µm in height, as shown in detail in Figure 1sand the contact pad was achieved using a resist mask and reactive ion etch.26 Scanning electron micrographs of a typical SECM-AFM probe are shown in Figure 1c and d, which clearly highlight the tip and cantilever geometry and dimensions. Materials. All solutions were made up from Milli-Q reagent water (Millipore Corp.) and comprised Ru(NH3)63+ (Strem Chemicals, Newbury Port, Ma, USA) at a concentration of 5 mM in 0.1 M KNO3 (Fisher Scientific). Two types of substrate were examined for combined SECM-AFM imaging. The first substrate was a diskshaped UME, formed by sealing a 10-µm-diameter platinum wire (Goodfellow, Cambridge, U.K.) into a glass capillary.28 The UME was shortened to a length of ∼1-2 mm and then set into a cylinder of epoxy resin,20 which acted as a support for the substrate when placed into the AFM fluid cell. This substrate was polished with 0.05-µm alumina on a moistened polishing cloth (Buehler, Coventry, U.K.). The second substrate was an UME grid array, comprising 14 400 1-µm-diameter disk electrodes connected in series. The electrodes were in a square region of (28) Wightman, R. M.; Wipf, D. O. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1989; Vol. 15, p 145.

1.5-mm length and separated by a center-center distance of 12.5 µm, to prevent the overlap of diffusion fields from neighboring electrodes when the array was operated amperometrically.29 To prepare the substrate, a silicon oxide wafer was initially coated with a thin layer of Pt. This was then covered with a 60-nm-thick layer of Si3N4, selectively patterned and dry etched to reveal a grid of 1-µm-diameter electrodes. Instrumentation. A Nanoscope E AFM with a fluid cell (Digital Instruments, Santa Barbara, CA) was used for all experiments. The AFM was equipped with a J scanner that had a maximum scan range of 120 µm × 120 µm. The AFM was sited inside a home-built Faraday cage, on top of a custom-built granite bench incorporating vibration isolators. Electrical connection to SECM-AFM probes was made by means of a wire attached to the spring clasp on the fluid cell. To aid the connection, the clasp was coated in silver DAG (Acheson Electrodag, Agar Scientific, Stanstead, Essex, U.K.) before finally insulating the probe body with a 1:1 superglue-nail varnish mixture. Electrochemical measurements were made using a CHI750A bipotentiostat (CH Instruments, Austin, TX). For electrochemical imaging, either the SECM-AFM tip current or the substrate current was recorded to the auxiliary input of the AFM using a DI breakout box. Procedures. Cyclic voltammograms (CVs) were run on both the substrate and the SECM-AFM tip before the experiment began to determine the potentials at which the two electrodes should be held. For the substrate generation-tip collection experiments, the substrate was held at -0.4 V (vs Ag/AgCl), sufficient to reduce Ru(NH3)63+ at a diffusion-controlled rate, while the tip was held at +0.1 V, sufficient to oxidize Ru(NH3)62+ at a diffusion-controlled rate. All generation-collection experiments were carried out in a four-electrode mode with an Ag/AgCl wire as the reference electrode and a platinum wire as the counter electrode. In all experiments reported herein, a contact mode topography image was recorded of the substrate surface. The tip was then retracted from the surface a known distance, using the stepper motor of the AFM, and the electrochemical image was collected (fixed height imaging20). Finite Element Modeling. The models reported were created with the commercial finite element modeling package, FEMLAB (COMSOL Ltd., Oxford, U.K.), using the Chemical Engineering Module. Simulations were carried out on a desktop PC (Dell Dimension 2350, Pentium 4 2.5 GHz, 512 Mb RAM). It typically took ∼5-10 min to simulate each current response and diffusion profile of interest, for the three-dimensional diffusion problems described herein.

Figure 2. Steady-state voltammograms recorded at (a) 50 and (b) 10 mV s-1 for the reduction of 5 mM Ru(NH3)63+ in 0.1 M KNO3, at the same triangular-shaped SECM-AFM tip before (a) and after (b) 65 min of electrochemical imaging.

RESULTS AND DISCUSSION Bulk Voltammetry at SECM-AFM Probes. Before considering the imaging applications, it was first necessary to establish that bulk voltammetric measurements with the SECM-AFM tips were well-defined, quantitative, and reproduciblesboth from probe to probe and over a sufficiently long time for a given probe. Figure 2a shows a typical CV (scan rate 50 mV s-1) for the reduction of 5 mM Ru(NH3)63+ (in 0.1 M KNO3) at an SECM-AFM tip located in the AFM fluid cell. The probe was withdrawn a long distance from the substrate surface so that the bulk diffusion response could be measured. The voltammetry for this simple reversible

redox process is well-defined with a reasonably clear limiting current of ∼0.55 nA. For this measurement, the solution was not degassed; hence the slight slope on the limiting plateau is likely to be due to the onset of reduction of dissolved oxygen, present at a concentration of 0.25 mM.30 The limiting current response at a fixed potential was stable to within (2%, even when the electrode was operated for periods in excess of 1 h, sufficient to obtain many images in an SECMAFM experiment (vide infra). Figure 2b shows the response of the same probe (scan rate 10 mV s-1) after 65 min of electrochemical imaging, in which the probe had been held at a potential of 0.1 V corresponding to the limiting current for substrate generation-tip collection of Ru(NH3)62+. It can be seen that the voltammetric characteristics are similar to those shown in Figure 2a. The CV response of 30 probes tested thus far (15 from each of two batches of 60 probes) have been found to be similar to those depicted in Figure 2, displaying the classic steady-state signature for diffusion to a small electrode. One of the motivations for batch fabrication was to produce probes that were not only reproducible but for which diffusion

(29) Scharifker, B. F. J. Electroanal. Chem. 1988, 120, 61.

(30) Pletcher, D.; Sotiropoulos, S. J. Electroanal. Chem. 1993, 356, 109.


427

Figure 3. FEMLAB simulation showing the concentration profile at the electroactive part of the SECM-AFM probe (triangular electrode 1-µm base, 0.65-µm height) for diffusion-controlled electrolysis of 5 mM Ru(NH3)63+, in excess background electrolyte.

could be readily modeled. As highlighted in the Experimental Section, we used FEMLAB for this purpose because of its efficiency and ready adaptability to complex three-dimensional geometries. All simulations were carried out in real space with experimentally relevant parameters. The steady-state diffusion equation describing the transport of Ru(NH3)63+ to the SECMAFM tip in three dimensions is given by

0)D

(

)

∂2c ∂2c ∂2c + + ∂x2 ∂y2 ∂z2

(1)

where c is the concentration of Ru(NH3)63+, which has a diffusion coefficient, D, and x, y, and z are the Cartesian coordinates of the system. For all reported experiments, the bulk concentration, c* ) 5 mM and D ) 8.8 × 10-6 cm2 s-1.31 For the simulation of the bulk diffusion problem, the electrode surface concentration of Ru(NH3)63+ was set to zero and there was a no-flux boundary condition on all other parts of the tip surface. The concentration of Ru(NH3)63+ was set to the bulk value at the extremes of the simulation domain, ensuring that these were sufficiently far removed from the tip to allow the bulk concentration to recover naturally. Initial simulations considered the SECM-AFM tip, cantilever, and probe body, but from the results obtained, it was clear that consideration of only the cantilever and tip was necessary for simulation of the particular problems presented herein, resulting in a more efficient approach. Figure 3 shows the concentration profile at the electroactive part of the SECM-AFM probe, for diffusion-controlled consumption of Ru(NH3)63+, highlighting the highly localized nature of the diffusion process. For the geometry-defined (triangle height and base length of 0.65 and 1.0 µm, respectively), and electroactive mediator concentrations employed, the tip current calculated was (31) Birkin, P. R.; Silva-Martinez, S. Anal. Chem. 1997, 69, 2055.

428


0.57 nA, in good agreement with the experimental data shown in Figure 2. The close similarity of the simulated and experimental current indicates that the cantilever insulator coating, in the vicinity of the metal connecting wires, is pinhole free, as demonstrated experimentally by extensive tests. Specifically, Ag electrodeposition (2 mM AgNO3 in 0.1 M KNO3 solution) and subsequent SEM analysis have shown that Ag growth occurs only at the apex of the tips, over an area consistent with limiting current measurements. The use of thin metal contact wires that run from the probe body to the tip apex (two metal wires 1 µm in width, which run the length of the 120-µm-wide and 151.5-µm-long cantilever) minimize the possibility of pinholes forming in the region of underlying metal. Occasionally, larger bulk tip currents were observed, as shown in the voltammogram in Figure 4a for the reduction of 5 mM Ru(NH3)63+ (0.1 M KNO3) at 10 mV s-1, where the limiting current is ∼1.2 nA. During the microfabrication process, it is possible that there may be a small variation in the area of gold exposed at the tip apex due to the slightly nonuniform nature of electron beam resist deposition on a structured surface. This process may result in slightly larger electrodes than the predicted triangular electrode; i.e., there is a small additional contribution from a rectangular portion of the electrode (see Figure 4b inset). This situation was found to occur at probes that were situated toward the edges of the wafer. However, through simulations it is a relatively simple task to determine the exposed electrode area and identify the length of the exposed band, by simply measuring the limiting current. For band electrodes with the dimensions specific to these probes, width 1 µm, the parameter φ ) Dt/r2 (where t is time and r the smallest dimension of the electrode), which describes the diffusional characteristics of the electrode, is .10.32,33 This dictates that the diffusional response of the electrode is strongly (32) Wehmeyer, K. R.; Deakin, M. R.; Wightman, R. M. Anal. Chem. 1985, 57, 1913.

Figure 4. (a) Steady-state voltammogram recorded at 10 mV s-1 for the reduction of 5 mM Ru(NH3)63+ in 0.1 M KNO3 solution at an SECM-AFM probe with an active electrode area comprising a triangular-shaped Au electrode (1-µm base, 0.65-µm height) and exposed Au band. (b) Plot of the simulated limiting current (reduction of 5 mM Ru(NH3)63+ in 0.1 M KNO3) as a function of exposed band length, together with an inset cartoon defining the electrode geometry considered.

dominated by radial diffusion and steady state will prevail at slow scan rates, in line with the FEMLAB simulations. This analysis does not take into account the presence of the adjoining triangularshaped electrode, which will further act to increase the radial diffusion properties of the overall electrode geometry, leading to steady-state behavior at the scan rates employed in these studies. Figure 4b shows a plot of the simulated limiting current as a function of exposed band length, together with an inset cartoon defining the electrode geometry considered. This plot can be used to calibrate the electrode area quantitatively. Topographical Imaging. The spring constant, k, of the batchfabricated SECM-AFM probes was measured using the thermal noise method.34 Through measurements on tips from different batches, k was found to be in the range 1.0-1.5 N m-1. These tips were suitable for imaging in both contact mode and tapping mode (including lift mode to collect the current signal) under solution,35 although in this paper, we only present contact mode topographical images. Figure 5 shows a typical 1 µm × 1 µm contact mode AFM image (height and deflection), recorded using (33) Porat, Z.; Crooker, J. C.; Zhany, Y.; Le Mest, Y.; Murray, R. W. Anal. Chem. 1997, 69, 5073. (34) Hutter, J. L.; Bechhoefer, J. Rev. Sci. Instrum. 1993, 64, 1868. (35) Holder, M. N.; Wilson, N. R.; Macpherson, J. V. Unpublished data.

a typical SECM-AFM probe, of an Au surface (100 nm thick, with 20-nm Cr underlay) sputtered onto glass. The surface shows islands, which correspond to the Au grains, with lateral dimensions in the range ∼30-50 nm and heights of ∼2-8 nm, as expected for an Au surface prepared in this manner. These features correspond well with those seen using a conventional Si3N4 tip, highlighting the topographical imaging qualities of this type of probe. The conductivity properties of the tip apex were investigated by applying a potential between the tip and substrate and measuring the current (through a current limiting resistor) as the tip scanned over the Au surface (C-AFM).18,20 In the majority of the cases, little or no conductivity was observed. This is not unexpected given the probe geometry in which metal wires are only present on one side of the tip; the rest of the tip consists of Si3N4 (see Figure 1b). Moreover, it is possible that the electrode could recede very slightly from the tip apex during the final reactive ion etch. For C-AFM applications, modifications to the microfabrication procedure could be made to produce probes wthat are conducting at the tip apex. Quantitative Combined Electrochemical-Topographical Imaging. FEMLAB simulations were employed as a guide to the most effective electrochemical imaging strategies with these SECM-AFM tips. The feedback mode has been used extensively in SECM applications to quantify processes such as lateral charge transport and conductivity in thin films,36 the rate of heterogeneous electron transport at metal electrodes,37 and fast coupled solution reactions,38 among many possible studies.3 Initial simulations considered the form of the approach curves to both a conducting (pure positive redox feedback) and an inert (negative redox feedback) surface. Typical approach curves are shown in Figure 6 for the triangular-shaped tip geometry already defined, showing the steady-state diffusion-limited currentsas a function of tip-sample separationsfor each of these surfaces. For the positive feedback case, the concentration of Ru(NH3)63+ at the active part of the tip surface was set to zero, while the bulk concentration was attained on the substrate surface. For negative feedback, a zero-flux boundary condition held for the substrate, with all other parameters as already defined for the bulk solution simulation. The data are presented as the steady-state limiting tip current, it, normalized by that in bulk solution, i(∞), as a function of tip-substrate separation, d. It can be seen from the data in Figure 6, that the bulk current holds, for all but the closest distances, and that even with the probe just 200 nm from the surface the current ratio is only 1.19 (positive feedback) and 0.86 (negative feedback). This type of behavior has also been found for conical electrodes16,20,39 and highlights that feedback experiments are not likely to be the most sensitive application of this type of probe geometry. For electrochemical imaging on ultra-small-length scales, in general, the most effective strategy is likely to be the tip collection mode,3,20 since under the high diffusion conditions to small (36) (a) Mandler, D.; Unwin, P. R. J. Phys. Chem. B 2003, 107, 407. (b) Zhang, J.; Barker, A. L.; Mandler, D.; Unwin, P. R. J. Am. Chem. Soc. 2003, 125, 9312. (37) Wipf, D. O.; Bard, A. J. J. Electrochem. Soc. 1991, 138, 469. (38) Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1991, 95, 7814. (39) Mirkin, M. V.; Fan, F. R. F.; Bard, A. J. J. Electroanal. Chem. 1992, 328, 47.


429

Figure 5. Typical 1 µm × 1 µm contact mode height (a) and deflection (b) AFM images of an Au surface sputtered onto glass, recorded in air using a typical SECM-AFM probe. The Au grains are clearly visible in both images.

Figure 6. Simulated feedback approach curves, for the reduction of 5 mM Ru(NH3)63+ at the triangular-shaped tip geometry defined in Figure 1b. Curves are shown for the steady-state diffusion-limited current, as a function of tip-sample separation, for approach to an insulating (b) and conducting (9) surface. Each point represents the simulated current response at a defined tip-substrate separation. The limiting current, it, has been normalized with respect to the current that flows at the tip electrode, far from a surface, i(∞).

structures, it becomes increasingly difficult to observe surface kinetic effects above the background current for hindered diffusion,40 which is always present in feedback mode imaging. In contrast, in the collection mode, there is no contribution from hindered diffusion to the tip current and the probe simply picks up particular species generated at the substrate surface (with zero background current in the ideal case). Consequently, this mode has found major application in visualizing active sites on surfaces5,6,41 and imaging biological surfaces and processes, such as the activity of immobilized enzymes.40,42 Although not termed SECM, the investigation of cellular processes with UMEs gener(40) Pierce, D. T.; Unwin, P. R.; Bard, A. J. Anal. Chem. 1992, 64, 1795.


ally utilizes collection strategies.43 These are areas where we ultimately envisage future applications of combined SECM-AFM. Having suggested that tip collection strategies are likely to represent the most sensitive approach for SECM-AFM imaging with probes of the defined geometry, the diffusion of electroactive species from a disk-shaped Pt UME substrate (10-µm diameter) was investigated. Figure 7 shows typical AFM topography images (height and deflection) recorded with the substrate in an aqueous solution containing 5 mM Ru(NH3)63+ and 0.1 M KNO3. The disk electrode is clearly observed, as are fine scratches in the surface resulting from the 0.05-µm-diameter alumina polishing procedure. These images again highlight the high topographical resolution attainable with these probes. For subsequent electrochemical imaging, the tip was backed away ∼500 nm from the surface using the stepper motor and used to detect Ru(NH3)62+ generated by reduction of Ru(NH3)63+ at the substrate electrode. The substrate was held at a potential of -0.4 V, while the tip was held at 0.1 V, so that the electrodes were operating under diffusion control, with respect to the species detected. With the probe 500 nm above the surface, the substrate current barely changed compared to the bulk value (with no probe), indicating that the probe did not significantly perturb diffusion to the substrate. This was also observed to be the case through simulation, as shown in Figure 8, which demonstrates that the usual hemispherical diffusion profile of electroactive species to the 10-µm-diameter circular substrate is largely maintained, even in the presence of the SECM-AFM probe. The simulations also demonstrated (inset to Figure 8) that a collector (41) Hengstenberg, A.; Blo ¨chl, A.; Dietzel, I. D.; Schuhmann, W. Angew. Chem., Intl. Ed. 2001, 40, 905. (42) (a) Wittstock, G.; Schuhmann, W. Anal. Chem. 1997, 71, 55059. (b) Oyamatsu, D.; Hirano, Y.; Kanaya, N.; Mase, Y.; Nishizawa, M.; Matsue, T. Bioelectrochemistry 2003, 60, 115. (43) (a) Amatore, C.; Arbault, S.; Bruce, D.; de Oliveira, P.; Erard, M.; Vuillaume, M. Faraday Discuss. 2000, 116, 319. (b) Garris, P. A.; Kilpatrick, M.; Bunin, M. A.; Michael, D.; Walker, Q. D.; Wightman, R. M. Nature 1999, 398, 67.

Figure 7. Typical AFM height (a) and deflection (b) images of a 10-µm-diameter Pt disk UME sealed in glass and polished flat, recorded with an SECM-AFM probe. The substrate immersed in an aqueous solution containing 5 mM Ru(NH3)63+ and 0.1 M KNO3.

Figure 8. FEMLAB simulation demonstrating that the usual hemispherical diffusion profile of electroactive species (5 mM Ru(NH3)63+ in 0.1 M KNO3) to a circular 10-µm-diameter Pt disk substrate, biased at a potential to reduce Ru(NH3)63+ at a diffusion-limited rate, is largely maintained in the presence of the SECM-AFM probe, which serves as a collector electrode. The simulations also demonstrate (higher resolution inset to the figure) that a collector electrode of this size and geometry serves as a relatively noninvasive probe of the steady-state concentration profile generated by the substrate.

electrode of this size and geometry is essentially a noninvasive probe of the steady-state concentration profile generated at the 10-µm-diameter disk UME, especially at larger tip-substrate separations.

Figure 9a shows a typical two-dimensional current collection image recorded in the substrate generation-tip collection mode, at a tip-substrate separation of 500 nm. This image was obtained in the same area as the topography images displayed in Figure 7. Analytical Chemistry, Vol. 77, No. 2, January 15, 2005

431

Figure 9. (a) Typical current collection image, recorded at an SECM-AFM probe (in the substrate generation-tip collection mode) for the diffusion-limited oxidation of Ru(NH3)62+ at the tip, generated at a 10-µm-diameter Pt disk UME via the reduction of Ru(NH3)63+. A tip-substrate separation of 500 nm was employed, and the bathing solution contained 5 mM Ru(NH3)63+ in 0.1 M KNO3. This image was obtained in the same area as the topography images displayed in Figure 7. (b) Simulated SECM-AFM collection current response, for the substrate generation-tip collection experiment, outlined in (a), for the tip situated at defined (b) distances above the center of a 10-µm-diameter Pt disk substrate. The tip separation distance, d, has been normalized with respect to the radius of the substrate electrode, as ) 5 µm.

The current image took 512 s (256 lines at 0.5 Hz) to acquire, and during this time, no deterioration in the amperometric signal was observed. A clearly defined current map is seen, for the detection of Ru(NH3)62+ diffusing away from the substrate. With the tip located above the center of the substrate, a maximum current of ∼0.64 nA was detected. This correlated well with the value of 0.61 nA expected at this tip-substrate separation, as shown by the data in Figure 9b, which is the simulated current response for the substrate generation-tip collection problem, with the tip located above the center of the substrate at different separations. Note that, at very small tip-substrate separations, the theoretical tip current approaches a value slightly higher than that for bulk diffusion to the tip. This is expected,44 because at close distances there will be some redox feedback between the tip and (44) Amatore, C.; Bento, M. F.; Montenegro, M. I. Anal. Chem. 1995, 67, 2800.


the surface, with this reversible mediator couple, which is accounted for in the FEMLAB simulation. High-Resolution Combined Electrochemical-Topographical Imaging. To demonstrate the capabilities of these probes for high-resolution topographical and electrochemical imaging, the surface topography and diffusion characteristics of an array of 1 µm diameter UMEs electrogenerating Ru(NH3)62+ were imaged, as shown in Figure 10 (20.4 µm × 20.4 µm scan image). The solution initially contained 5 mM Ru(NH3)63+, and the array electrodes were biased at a potential to generate Ru(NH3)62+ at a diffusion-controlled rate. The SECM-AFM probe is clearly able to resolve the topography of the array electrodes, revealing electrodes of width 1.00 ( 0.05 µm (Figure 10a). The corresponding two-dimensional unfiltered current collection image, recorded in the substrate generation-tip collection mode, at a tip-substrate separation of 1 µm is shown in Figure

Figure 10. (a) Topography and (b) unfiltered fixed-height current maps for the diffusion-controlled tip detection of Ru(NH3)62+, generated from the diffusion-limited reduction of Ru(NH3)63+ at an array of 1-µm-diameter substrate electrodes, in a solution containing 5 mM Ru(NH3)63+ and 0.1 mol dm-3 KNO3. Both images were recorded in the same 20.4 µm × 20.4 µm area. For topographical imaging, the tip was held in contact with the surface (unbiased), while electrochemical data were acquired with the tip imaging at a fixed height of 1 µm from the surface of the substrate. For electrochemical imaging, the black spots correspond to peak currents in the range 130-140 pA above the background current. (c) Limiting tip current-x scan distance responses as the tip was scanned over the center of the 1-µm-diameter UMEs, situated top middle (i) and far left (ii) of the height image in (a).

10b. The diffusion of electrogenerated Ru(NH3)62+, away from the array of 1-µm-diameter electrodes, is clearly detectable using the SECM-AFM tip and is revealed as “hot spots” in the tip current collection image. The maximum currents detected in these regions, with the tip ∼1 µm from the surface, are in the range 130-140 pA above the background current. Figure 10c shows plots of the limiting tip current versus x scan distance as the tip scanned over the 1 µm diameter UMEs, situated top middle (i) and far left (ii) in the topography image in Figure 10a. The current responses are clearly peak-shaped, as expected for diffusion away from a disk source.5 Interestingly, one of the UMEs (bottom of Figure 10a), although topographically identifiable, is not generating Ru(NH3)62+, as shown by the absence of a tip collection current in this surface location. Panels a and b of Figure 10 thus illustrate the ability of the SECM-AFM probe to correlate the structure of a surface with its (electro)chemical activity at high spatial resolution. CONCLUSIONS Batch microfabrication has been shown to produce SECMAFM probes capable of quantitative topographical and electro-

chemical imaging of solid surfaces in contact with electrolyte solutions. The probe design includes a triangular electrode at the tip end, which is most suitable for electrochemical imaging in the collection mode. The tips are well-defined, reproducible, and longlived, giving stable topographical and electrochemical images, as evidenced by generation-collection concentration profiling of microelectrode substrates. A feature of the work herein is that mass transport (diffusion) to the tip is readily treated by simulation, which allows current data to be interpreted quantitatively. We envisage many potential applications of this type of combined SECM-AFM probe, including investigations of the reactivity of solid surfaces (dissolution, crystal growth, and corrosion) and the measurement of fluxes associated with cellular processes (e.g., respiration). The microfabrication process described herein could be readily adapted to other probe geometries, including, for example, disk-shaped electrodes and multiple electrode devices (e.g., ring-disks and dual-disks). There is also the possibility of enhancing the range of detectable species by introducing potentiometric detection formats. Finally, there is scope for enhancing the time resolution and noise characteristics of electrochemical component of this type of probe by incorporatAnalytical Chemistry, Vol. 77, No. 2, January 15, 2005

433

ing a preamplifier directly on the probe body as part of the microfabrication procedure.

Fellowship. Discussions with Dr. Andrew Glidle (Department of Electronic Engineering, Glasgow University) are gratefully acknowledged.

ACKNOWLEDGMENT

Received for review July 22, 2004. Accepted October 15, 2004.

We thank EPSRC (GR/R34738/01) for funding this project. J.V.M. is grateful to the Royal Society for a University Research

434


AC048930E

Characterization of Batch-Microfabricated Scanning Electrochemical

Recommend Documents