Article pubs.acs.org/ac
Mapping Electroactivity at Individual Catalytic Nanostructures Using High-Resolution Scanning Electrochemical−Scanning Ion Conductance Microcopy Michael A. O’Connell* and Andrew J. Wain National Physical Laboratory, Hampton Road, Teddington TW11 0LW, United Kingdom S Supporting Information *
ABSTRACT: Combined scanning electrochemical−scanning ion conductance microcopy (SECM−SICM) has been used to map the electroactivity of surfaces decorated with individual features at the 100−150 nm scale. Dual channel capillary probes consisting of an open SICM barrel, and a solid carbon SECM electrode enabled correlation of surface activity with accurate topographical information. Measurements were validated by approach curve analysis and imaging of model systems in feedback and substrate generation−tip collection modes and then applied to the examination of two nanostructured test substrates. First, electronically isolated gold nanodisk arrays were imaged using a simple electrochemical redox mediator, in which a clear positive feedback signal was observed at the SECM electrode, and the topographical channel compared well with AFM imaging. Second, platinum nanosphere ensembles were mapped using platinum-modified carbon probes to detect oxygen consumption in a redox competition mode, demonstrating the means to study electrocatalytic processes at individual nanoparticles. This work demonstrates the value of high-resolution SECM−SICM for low-current amperometric imaging of nanosystems, and is a step toward quantitative measurement of electrokinetics at the single particle level.
U
(thus maintaining a small and constant tip−sample separation) and retaining the breadth of information that can be gleaned from classical SECM. To meet these criteria, it is necessary to produce probes that combine a miniaturized, addressable electrode with a means of highly accurate positional feedback. A range of approaches have been reported, notably SECM−AFM13,15−24 and shear forceSECM,25−34 as well as others, including intermittent-contact (IC)-SECM35−37 and alternating current (AC)-SECM methods.38−40 These approaches vary considerably in resolution, topographical accuracy, ease of implementation and, often of vital importance, ease and consistency of tip fabrication. Although not widely adopted, many of these techniques are beginning to be offered commercially or via minimal modification of commercial equipment. Of the various techniques employing force−based positional feedback, SECM−AFM offers the high resolution inherent to AFM but requires highly complex and costly probe manufacture. ICSECM offers a simpler force feedback approach employing standard SECM probes, but with much higher contact forces (potentially damaging either the sample or tip). In shear forceSECM, the force response is highly sensitive to many factors and is very short-range, and although demonstrated as a robust
nderstanding the effect of nanoscale heterogeneities at electroactive interfaces is paramount to the development and optimization of a range of energy conversion technologies such as fuel cells, electrolyzers, batteries and related catalytic systems.1,2 Electrochemical imaging techniques provide a means to gather spatially resolved information about interfacial phenomena relevant to these applications, in some cases enabling its correlation with surface structure and composition. Such methods offer a route to developing a better understanding of complex electrochemical systems beyond that which can be accessed from macroscale, particle ensemble measurements. Furthermore, localized activity measurements are widely applicable to nonelectrode dynamic surfaces, such as those relevant to heterogeneous catalysis and biology.3−5 Conventional scanning electrochemical microscopy (SECM), the parent technique from which most electrochemical imaging tools are derived, involves the use of a microelectrode probe to perturb or sample the local electrolyte solution immediately adjacent to the surface of interest. Given the wealth of chemical insight that can be gleaned from micrometer-scale SECM and the variety of systems within its grasp,3,6−8 much work has been performed to increase the resolution and accuracy of the technique.5,9 Indeed, there are a number of diverse examples of measurement of electrochemical signals from features well below the micrometer scale.10−14 The ultimate goal of these endeavors is to monitor electrochemical processes at the nanoscale while both faithfully tracking the surface topography © XXXX American Chemical Society
Received: August 6, 2014 Accepted: November 15, 2014
A
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Figure 1. (a) Schematic of combined SECM−SICM experiment, (b) CV for a typical SECM electrode in 1 mM FcMeOH/0.1 M KNO3 and (c) SEM image of a probe apex showing open (dark) and carbon-filled (light) barrel with surrounding glass.
method for mapping at the micrometer scale, imaging at high spatial resolution is rare. Positional control in such techniques has also benefited from the development of advanced scanning patterns.31,37 Some recent methods have employed capillary probes with positional feedback based on ion conductance,41−45 which have the advantage of simple probe fabrication and do not require any mechanical interaction force. One such approach, scanning ion conductance microscopy (SICM),44,46−50 utilizes an electrode placed inside the capillary biased against one in external solution, between which a small ionic current flows. The positional feedback response is based on the drop in migration current due to physical hindrance of ion flow as a function of tip−surface separation. A simple single capillary can be used for pure topographical imaging, an approach growing in its own right, particularly for biological imaging applications, due to its efficacy for low- (or non-) contact imaging in aqueous media. One advantage of this technique is that SICM probes of the desired size, from nanometers upward, can be fabricated via a trivial and inexpensive laboratory procedure using a benchtop laser puller. For combined SECM−SICM (shown in Figure 1a), introduced by the groups of Korchev and Matsue41 and Hersam,42 inclusion of a second “faradic” electrode enables simultaneous SECM measurement during topographical scanning. Integration of an SECM electrode into the capillary probe architecture can be performed by either coating of a standard single capillary with an electrode material, followed by insulation and exposure of the completed probe,41,42 or by using dual barrel (theta) capillary and filling one barrel with a solid electrode material.43 The former method allows for a choice of coating material (Au, Pt, C) but generally requires expensive and laborious focused ion beam (FIB) milling for each probe, whereas the latter is more economical
but has generally been used only for the production of carbon probes, limiting the range of electrochemical systems that can be studied. Recent work has, however, opened the possibility of electrochemically modifying such carbon electrodes with various materials, thus circumventing this restraint51−53 and potentially yielding a source of inexpensive, easily fabricated, and highly tailored dual function probes. So far, diverse examples of applications of SECM−SICM include high-resolution imaging performed on a range of biological samples using simple electrochemical mediators,41,43 as well as further work on model electrode band41−43 and porous structures.43,54 A number of examples have also been reported using potentiometric detection;52,55−57 for example, for the measurement of pH,52,56 which has been found to be aided by a better signal-to-noise ratio than for small probe amperometric detection.55 In many of the above examples of SECM−SICM, imaging was performed on samples comprising features with micrometer scale dimensions, and higher resolution images have been largely limited to simple passive diffusional systems (e.g., effusion of electroactive species from pores) rather than at electrochemically active surfaces. In the present paper, we extend the application of high-resolution SECM−SICM to mapping the reactive chemistry of electroactive features at the 100−150 nm scale and demonstrate imaging of electrocatalytic activity at the single nanoparticle level.
■
EXPERIMENTAL SECTION Materials. All chemicals were purchased from SigmaAldrich at the highest available purity and were used as received. Solutions were prepared using 18.2 MΩ cm water (Millipore Corporation). Quartz theta capillaries (o.d. 1.2 mm, i.d. 0.9 mm) used for dual barreled probes were purchased from B
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
(Goodfellow) via electrodepostion from 2 mM K2PtCl6 in 0.1 M H2SO4, based on the method introduced by Tian et al.61 To control the nanosphere size, an initial nucleation step of typically −0.25 V vs SCE for 20 ms was followed by a square wave deposition profile between 0.2 and 0.6 V at 100 Hz for 15−30 min. SECM−SICM Experiments. Images were collected in “hopping mode” at a pixel density of typically 200 × 200 pixels per image. Hopping heights varied from 100−500 nm at typical approach rates of ∼2−10 μm/s to an imaging set point of 1.5−3% drop of ion current. This corresponds to an approximate working distance of ∼2 tip radii (vide infra). Smaller set points than this were rarely found to give accurate topographical images. Typical scan times for high-resolution imaging were between 30 and 60 min. Images were plotted, filtered, and analyzed using Scanning Probe Image Processor software (SPIP, Image Metrology, Denmark, see the Supporting Information (SI) for details). For band and disk samples, scans were performed in 1 mM FcMeOH with 10 mM KNO3 as the supporting electrolyte, and for platinum nanosphere samples, scans were performed in 0.1 M NaOH.
Sutter Instrument Company (US), and silver wires used for reference electrode fabrication, from Goodfellow (UK). Instrumentation. Combined SECM−SICM scanning was performed on a modified Ionscope SICM instrument (Cambridge, UK) with electrochemical control via either a patch clamp amplifier (Axopatch 200B, Molecular Devices), controlled using a custom-written LabVIEW interface, or a Heka PG340 bipotentiostat. Capillary fabrication was performed using a laser pipet puller (model P-2000, Sutter Instrument Company, US), and electrochemical modification was performed using an Autolab potentiostat (PGStat 302N, Metrohm, Netherlands). Chloridized silver wires were used as Ag/AgCl quasi-reference electrodes during scanning, and a saturated calomel electrode (Hg/Hg2Cl2, SCE) was used during electrochemical modification. The potentials of the Ag/AgCl quasi-reference electrodes were approximately equal to that of SCE in all solutions used. Scanning electron microscopy (SEM) images were recorded using a Zeiss Supra SEM. AFM images were recorded using a Veeco (now Bruker) Multimode with a Nanoscope IIIa controller operated in Tapping mode. Probe Fabrication. Dual capillary probes were fabricated according to a previously reported procedure.43 Quartz theta capillaries were pulled to generate aperture sizes from a few tens of nanometers to ∼1 μm diameter (dependent on the sample) in distributions with ranges of ∼100 nm in general and narrow for very small tips. A stable environment was found to be essential for producing consistent capillary geometries, and success rates of >80% were typically achieved for tips of >100 nm diameter. A single barrel of the theta pipet was then filled with conducting pyrolytic carbon by pyrolysis of butane using a blowtorch. Burn times of ∼30s were found to be sufficient to ensure a nonporous carbon electrode. Electrode diameters were coarsely estimated from cyclic voltammetry in 1 mM FcMeOH, assuming, for simplicity, disk-shaped geometry, free diffusion from the SICM barrel of the (uncharged) mediator and an RG (ratio of insulating surround radius to active electrode radius) value of 1.1 using the equation:8
■
RESULTS AND DISCUSSION Dual function probes were fabricated as described above at a range of sizes to suit individual samples. Probes with apertures ranging from a few tens of nanometers up to ∼1 μm were easily attainable. Figure 1b shows a sample cyclic voltammogram (CV) from a probe at the smaller end of this size range; voltammetry was generally of good quality with sigmoidal shape and low capacitance, as expected for such probes. The diffusionlimited current of ∼10 pA indicates an approximate carbon electrode radius of 30 nm (vide supra), although it is important to note that this is only an estimation, given the complex nature of the probe geometry and the possible blocking effects of surface contaminants. Probes were also imaged using SEM to check for overspill from the carbon electrode deposition or major imperfections in geometry. An example of a slightly larger probe is depicted in Figure 1c, in which the carbon-filled barrel on the right-hand side of the image appears to exhibit no significant recess or protrusions. Further SEM images of probes are provided in SI Figure S3. We note that these probes were highly susceptible to collecting dust particles from the air, as evidenced by the small particles surrounding the tip in this image. The electrochemical response of carbon probes was further evaluated by performing approach curves on different substrates (Figure 2). As expected, negative feedback was observed upon approach to an insulating glass surface (Figure 2a), and positive feedback was evident at a conducting Pt surface (Figure 2b). The SICM current collected during these approach curves is clearly insensitive to the nature of the surface and therefore is suitable as a mechanism for positional feedback. It is also clear that the SECM response is sensitive over a longer range than the SICM response, leading to a much larger change in the normalized current. Comparison of the SECM approach curves to simulated curves calculated on the basis of standard expressions for pure positive and negative feedback (using a tip radius of 330 nm and an RG value of 1.1)8 gives good agreement with the experimental data. Small deviations between experiment and theory in the negative feedback approach curve may result from the irregular nondisk geometry of the probe. The electrode radius used for the fitting procedure corresponds to that estimated from cyclic
Ilim = 5.14nrFDc
where Ilim is the diffusion-limited current, 5.14 is the factor accounting for additional back diffusion for an electrode of RG 1.1, n is the number of electrons transferred, r is the electrode radius, F is Faraday’s constant, D is the diffusion coefficient of the mediator (6.7 × 10−6 cm s−1 for FcMeOH),58 and c is the concentration of mediator. Modification of carbon electrodes with platinum nanoparticles was performed via cathodic deposition in a solution containing 3 mM K2PtCl6 and 0.1 M HClO4. A typical procedure involved initially holding the electrode at a small anodic potential (generally 0.5 V vs SCE), followed by stepping to −0.3 V vs SCE for 5 s. After electrochemical probe sizing or modification, care was taken to avoid drying of electrolyte upon its removal from solution, since this would often cause irreversible blocking of one or both barrels. Therefore, for imaging applications, probes were generally characterized/ modified immediately prior to use. Samples. Platinum interdigitated array (IDA) electrodes (2 μm) on glass were purchased from ALS (Japan), and 150 nm gold disk samples on silicon with a 300 nm insulating SiO2 layer were fabricated using hole−mask colloidal lithography.59,60 Monodisperse platinum nanospheres (diameters ranging from 50 to 200 nm) were grown on glassy carbon substrates C
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
To characterize the imaging response of the probes, initial current-topography mapping experiments were undertaken at the micrometer scale using an addressable Pt interdigitated electrode consisting of continuous 2 μm wide, 50 nm high bands of Pt on glass. Figure 3a depicts a topographical image of the band structure, and Figure 3b shows the concurrent electrochemical feedback response recorded in 1 mM FcMeOH at a SICM set point of 2.0%. The bands are well-defined in both images, with the raised areas of the Pt bands generating positive feedback in the SECM channel. Figure 3c shows the same area in substrate generation−tip collection (SG−TC) mode in which the substrate is biased to oxidize FcMeOH while the tip is poised to detect the electrogenerated FcMeOH+. In this case, an enhanced cathodic current is recorded as expected over the Pt bands. The cross sections from the three images show strong correlation between the channels: the topography is reasonably well-followed, albeit without demonstrating the plateau of the Pt bands, which is in keeping with the expected topography recorded by SICM imaging (in this case, a relatively large ∼300 nm tip was used to suit the sample dimensions).62 The feedback response follows the shape of the bands particularly well, whereas the SG−TC image appears to be more asymmetric, likely indicating artifacts due to the tip geometry. This is an unusual observation, since the SG−TC mode is normally expected to be less distance-dependent than feedback mode. We speculate that in SG−TC mode, the SICM barrel shields one side of the SECM electrode, leading to asymmetric blocking of the FcMeOH+ collection and, hence, a directional response. In this case, the FcMeOH+ is generated only by the surface, and the signal may include FcMeOH+ generated by adjacent bands, further increasing this effect. In contrast, in feedback mode, the FcMeOH would be expected to freely diffuse from the SICM barrel, meaning this barrel would have little effect on the feedback response. It should be noted that operating in hopping mode rather than a typical raster scan would not be expected to remove this asymmetric effect in SG−TC mode since it arises from the tip geometry rather than the scan direction. When using these probes, it was found to be important to monitor changes in both SICM and SECM current over time. SICM currents were generally highly stable, and fluctuations and drift (beyond an initial few minutes following immersion) generally indicated instabilities in the SICM system. For example, reference electrode potential drift, excessive dissolution at the surface, or tip blocking could all prevent stable
Figure 2. Approach curves in 1 mM FcMeOH/0.01 M KNO3 for (a) glass (negative feedback with tip held at 0.4 V vs Ag/AgCl), (b) unbiased platinum band (positive feedback with tip held at 0.4 V vs Ag/AgCl). Measured SECM current is shown in black, simulated current for an RG of 1.1 is shown as red curve with triangles, and SICM current is shown in blue.
voltammetry, assuming a disk-shaped geometry. The validity of this assumption will clearly vary from probe to probe. Nevertheless, the tip−sample separation indicated here is clearly well within the range for highly sensitive SECM measurements. No attempt was made to fit the SICM approach curves, largely because of a lack of established theory for asymmetric probes; however, on the basis of the parameters used for these fitted SECM approach curves, we estimate that the ionic current set points used for imaging (i.e., 1.5−3.0% current decrease) correspond to a working distance of between 1.7 and 2.5 SECM electrode radii.
Figure 3. Images of platinum band electrode in 1 mM FcMeOH/0.01 M KNO3 showing (a) topography (SICM channel), and faradic current in (b) feedback mode and (c) substrate generation−tip collection. For parts a and b, the tip is held at 0.4 V vs Ag/AgCl, whereas for part c, the substrate is held at 0.4 V, and the tip, at −0.1 V vs Ag/AgCl. Cross sections are taken for the bottom line. D
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Figure 4. Images of gold nanodisk array in 1 mM FcMeOH/0.01 M KNO3: (a) SICM topography, (b) concurrent SECM current (feedback mode with tip biased at 0.4 V vs Ag/AgCl), and (c) an AFM image of a similar area of the sample. Cross sections are indicated by dotted lines.
catalysis in the case of nanoparticle-decorated surfaces.63 Hence, localized electrochemical analysis has the potential to address the need for assessing electrocatalyst nanoparticle activity within an ensemble. Oxygen reduction is a far more complex reaction compared with the single electron oxidation of FcMeOH and takes place via an inner sphere 2 or 4 electron mechanism to generate hydrogen peroxide or water. The overpotential for ORR on carbon is sufficiently high that such surfaces can be considered largely inert, rendering the carbon capillary probes used thus far problematic for studying this reaction. Therefore, to use these probes as a local oxygen sensor, it was necessary to amplify the ORR signal by functionalizing the carbon electrode via electrodeposition of platinum.51,53 The deposition procedure was refined on relatively large probes (radius 250−500 nm), in which a cathodic pulse at −0.3 V vs SCE for 5 s in 3 mM K2PtCl6 dissolved in 0.1 M HClO4 was found to give consistent platinization without extensive overgrowth from the electrode. Figure 5 depicts CVs performed before and after platinization of a typical probe. Voltammetry performed using FcMeOH as an outer sphere redox mediator (Figure 5a) indicated no extensive tip enlargement as a result of the platinization process, whereas a significant increase in ORR current was measured in oxygen-saturated 0.1 M NaOH (Figure 5b). Here, both the earlier ORR onset potential and the much larger currents indicate the desired catalytic effect of the platinum functionalization. Current amplification factors up to an order of magnitude were observed at −0.6 V vs SCE. Similar amplification factors were found when scaled down to smaller probes, yielding the necessary sensitivity for imaging when using small probes and lower dissolved concentrations. SEM imaging of the surface of a much larger Pt coated probe is provided in SI Figure S2, although the exact nature of the deposited Pt is difficult to determine. Platinum nanosphere electrocatalyst samples were prepared on glassy carbon substrates by a square wave deposition procedure. An SEM image of a small area of a typical sample is shown in Figure 6a. The platinum nanospheres here are shown to be largely monodisperse, and sizes ranging from 50−200 nm were easily obtainable. Cyclic voltammograms in 0.1 M NaOH were performed to establish the ORR behavior of these Pt nanospheres (Figure 6b). Although there is some ORR activity on the bare glassy carbon, the overpotential for ORR on even a low-density platinum nanosphere-modified sample is clearly lower (onset occurring at approximately −0.05 V vs Ag/AgCl compared with approximately −0.35 V vs Ag/AgCl on bare
scanning. The SECM signal was found to often degrade initially by a few picoamperes but then settle at a lower steady-state current, at which point it was generally reasonably stable over the course of a scan (∼30 min). However, with prolonged use, minor damage to the tip often occurred, leading to discrete increases in limiting current. Such damage was observed more commonly at high imaging set points (above 3.0%, typically) or at rapid approach speeds when set point overshoot became more significant. For imaging features with lateral dimensions at the micrometer scale, probes of a few hundred nanometers were ideal, given the need to image a relatively large area at sufficient pixel density within an acceptable time limit. However, smaller tips were required for high-resolution imaging of nanostructured surfaces. To demonstrate this, we employed a test sample comprising isolated 150 nm diameter, 50 nm high gold disks patterned on silicon. Figure 4 shows a combined SECM− SICM image of this sample, recorded in feedback mode in 1 mM FcMeOH using a probe with an electrode radius of approximately 10 nm based on voltammetry when initially immersed. The gold disks are clearly defined in both the topography (Figure 4a) and SECM (Figure 4b) channels. Comparison with AFM imaging (Figure 4c) indicates that the topography is tracked faithfully, although a small degree of “rounding” of the top of the gold disk is again evident in the SICM response. The SECM feedback response of the gold disks in Figure 4b is very small but clear, with ∼100 fA variation across the entire sample. The response from such small conductive but electronically isolated features may be minute, but it is nonetheless distinct from the full negative feedback response on the surrounding insulating silicon surface, indicating the potential for measuring very small signals. In this case, we note the tip used was much smaller than the features themselves, and hence, positive feedback is to be expected. Finally, we explored the diversification of this technique to enable its application to electocatalysis. The oxygen reduction reaction (ORR) is a much-studied electrocatalytic system of great importance to developing fuel cell technologies and is often performed at nanostructured platinum surfaces supported on carbon. Fuel cell electrocatalysts are typically characterized via macroscopic rotating disk electrode (RDE) voltammetry, and although informative, such ensemble measurements can fall short of reliably determining intrinsic individual nanoparticle electrokinetics. Indeed, recent work has highlighted the potential pitfalls of interpreting RDE voltammetry for electroE
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
carbon), indicating the electrocatalytic properties of the Pt nanospheres. To measure the ORR activity at these nanospheres using SECM−SICM, we employed a simplified redox competition mode SECM,64 in which oxygen is reduced at both the sample and the tip and the activity of the nanospheres is indicated by the local oxygen consumption. The sample was biased in the region of oxygen reduction on the platinum nanospheres only (−0.3 V vs Ag/AgCl), thus minimizing background contributions from the glassy carbon. Figure 7 depicts an SECM−SICM
Figure 7. Images of oxygen reduction at platinum nanospheres using platinized probes in 0.1 M NaOH: (a) SICM topography, (b) SECM image of competitive oxygen reduction signal during which the substrate is held at −0.3 V vs Ag/AgCl and the tip at −0.6 V vs Ag/ AgCl. Cross sections are indicated by the dotted line.
Figure 5. Platinum modification of carbon probes: (a) CV in 1 mM FcMeOH/0.1 M KNO3 before (black) and after (red) platinum deposition and (b) oxygen reduction CV in oxygen-saturated 0.1 M NaOH for the same probe before (black) and after (red) platinum deposition.
image obtained using a ∼100 nm electrode radius probe (we note that consistent platinization of probes smaller than 50 nm radii was found to be challenging). Nanoparticles are well resolved in the SICM topographical image (Figure 7a) and shown to be highly uniform in size, although the measured lateral diameters appear broadened by tip convolution. Nevertheless, the particle spacing in this sample is sufficient to clearly resolve individual Pt nanospheres. The SECM image (Figure 7b) exhibits the depleted oxygen profiles around the particles with excellent definition, indicating the consumption of oxygen by the electrocatalytic nanospheres. Given the proximity of the particles to each other and the tip size, it is difficult to quantitatively compare the oxygen consumption at each particle; however, it is apparent the oxygen consumption at each is similar, which is to be expected for nanospheres of uniform size and structure. Some background inhomogeneity in the oxygen reduction current on the glassy carbon surface is apparent in some areas, which may arise from very small platinum particles on the surface that are beyond the SICM topographical resolution. Although such particles were not discernible in the SEM imaging, this does not definitively exclude the presence of additional Pt at the surface. For example, additional ORR sites may arise from further Pt metal deposited from trace amounts of remaining platinate salt. Nevertheless, the change in current recorded at these sites is significantly smaller than for the topographically resolved nanospheres. Figure 8 shows the effect of varying the surface potential on the ORR current measured at the Pt nanospheres: Figure 8a shows a topographical image of two 80−100 nm particles, with a height cross section of these shown in part b. Part c shows the current response across this section with the surface held at 0 V, −0.3, and −0.6 V against Ag/AgCl. At 0 V, where no ORR
Figure 6. Characterization of Pt nanosphere substrates: (a) SEM micrograph of a typical sample and (b) ORR CVs in 0.1 M NaOH for bare glassy carbon (black) and Pt nanosphere-modified glassy carbon (red).
F
dx.doi.org/10.1021/ac502946q | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
surface features as small as 100−150 nm, advancing highresolution amperometric imaging using this technique beyond simple diffusional systems. The probes used are easily tailored to the required size (in the range of
