Scanning Electrochemical Cell Microscopy (SECCM) in Aprotic Solvents: Practical Considerations and Applications

: Many applications in modern electrochemistry, notably electrosynthesis and energy storage/conversion take advantage of the “tunable” physicochemical properties ( e.g. , proton availability and/or electrochemical stability) of non-aqueous ( e.g. , aprotic) electrolyte media. This work develops general guidelines pertaining to the use of scanning electrochemical cell microscopy (SECCM) in aprotic solvent electrolyte media to address contemporary structure−electrochemical activity problems. Using the simple outer-sphere Fc 0/+ process (Fc = ferrocene) as a model system, high boiling point (low vapor pressure) solvents give rise to highly robust and reproducible electrochemistry, whereas volatile (low boiling point) solvents need to be mixed with suitable low melting point supporting electrolytes ( e.g. , ionic liquids) or high boiling point solvents to avoid complications associated with salt precipita-tion/crystallization on the scanning (minutes to hours) timescale. When applied to perform microfabrication — specifically the electrosynthesis of the conductive polymer, polypyrrole — the optimized SECCM set up produces highly reproducible arrays of synthe-sized (electrodeposited) material on a commensurate scale to the employed pipet probe. Applying SECCM to map electrocatalytic activity — specifically the electro-oxidation of iodide at polycrystalline platinum — reveals unique ( i.e. , structure-dependent) patterns of surface activity, with grains of specific crystallographic orientation, grain boundaries and areas of high local surface misorientation identified as potential electrocatalytic “hot spots”. The work herein further cements SECCM as a premier technique for structure−function−activity studies in (electro)materials science and will open up exciting new possibilities through the use of aprotic solvents for rational analysis/design in electrosynthesis,

Surface structure/composition has long been known to have a profound effect on heterogeneous reaction kinetics and/or mechanisms. For instance, in electrochemistry, electrodes of practical importance are often nanostructured to introduce/expose specific sites and/or maximize the active surface area. It follows that understanding the structure−property relationships of functional (electro)materials is of utmost importance, especially with the resurgence of electrochemistry in organic (electro)synthesis and renewable energy storage/conversion technologies. 1 Thus, to complement the ever-expanding library of material nanoengineering strategies, 2 there is a great need for techniques that can probe (electro)chemical activity at the scale of surface heterogeneities, e.g., from single defects (e.g., step edges) and/or nanostructures (e.g., nanoparticles, NPs) to the individual grains and grain boundaries of a polycrystal. 3 Scanning electrochemical cell microscopy (SECCM) has emerged as a premier technique for correlative structure−function studies in (electro)materials science. [3][4][5] In SECCM, electrochemical measurements are performed in a statistically-large (typically hundreds to thousands) series of small areas (each nm 2 to μm 2 ) of a surface, defined by a meniscus cell created between a pipet probe filled with electrolyte solution and substrate (working electrode, WE) surface. 6,7 SECCM probes electrochemistry directly at an electrode surface, at a scale commensurate to commonly used structural/compositional characterization techniques (e.g., high-resolution microscopy and/or spectroscopy), allowing structure−activity to be assigned unambiguously. Indeed, this innovative correlative electrochemical multi-microscopy approach has previously been applied to resolve the activity of: steps and terraces in layered (two-dimensional) materials; [8][9][10][11] individual grains and grain boundaries of polycrystalline metals [12][13][14][15] and; single NP or NP agglomerates within an ensemble. [16][17][18] Complementing these applications, SECCM has also successfully been deployed as a tool for highthroughput fabrication (i.e., electrodeposition) and screening of nanostructured materials at electrode surfaces, notably conductive polymers (CPs) 19,20 and metal NPs. 21,22 All of the applications of SECCM highlighted above have mainly been carried out in aqueous electrolyte media, where the scanning protocols (e.g., meniscus cell stability, suitable supporting electrolytes and reference electrodes) are well-established. 4,5 Many applications in modern electrochemistry, including the electrosynthesis and energy storage/conversion (e.g., Li-ion batteries and supercapacitors) take advantage of the tunable physicochemical properties of non-aqueous electrolyte media. For example, through careful selection of the solvent and supporting electrolyte, the proton availability, donor/acceptor (i.e., Lewis base/acid) properties, electrochemical stability window etc. can be tuned, 23 allowing electrochemical reaction pathways to be controlled to an extent. Although SECCM has previously been applied in aprotic electrolyte media, including in ionic liquids (ILs) 24,25 and organic solvents (and gels), 26,27 the few studies available do not explicitly consider how the solvent/supporting electrolyte affects the stability of the meniscus cell, and the range of solvents studied is quite limited.
Herein, the feasibility of performing SECCM in non-aqueous media, specifically those based on aprotic solvents, is explored systematically. Taking organic carbonates as a model class of aprotic solvent, general guidelines are established that dictate the choice of solvent/supporting electrolyte to ensure a stable and reproducible electrochemical meniscus cell on the scanning (minutes to hours) timescale. Using the established guidelines, SECCM is applied to perform: (1) local electrodeposition of CP (polypyrrole) arrays and; (2) nanoscale structure−activity mapping of an electrocatalytic process (I − oxidation). All-in-all, the established guidelines and highlighted applications greatly expand the solvents that can be used in SECCM and further establish it as a robust tool for structure−function studies in (electro)materials science and electroanalytical chemistry.

EXPERIMENTAL SECTION
Chemical reagents and electrode materials. Details in the Supporting Information (SI), Section S1. Electrochemical measurements. Electrochemical experiments were performed on a home-built SECCM workstation, as previously reported. 4,5,7,28 In this configuration, shown schematically in Figure 1a, a single-barreled pipet probe (fabrication detailed below) was filled with electrolyte solution and mounted on a z-piezoelectric positioner (P-753.3, Physik Instrumente, Germany). The WE (nanocrystalline Au or polycrystalline Pt, Figure 1a) was mounted atop an xy-piezoelectric positioner (P-621.2, Physik Instrumente). The SECCM probe was initially positioned above the WE using coarse xy-micropositioners (M-461-XYZ-M, Newport, U.S.A.), and subsequently lowered into the near-surface position using a stepper motor (8303 Picomotor Actuator, Newport) in tandem with an optical camera (PL-B776U camera, 4× lens, Pixelink, U.S.A.).
The SECCM probe was approached to the WE surface, using a surface current (isurf) threshold of ca. 1.5 pA to detect when the meniscus-surface contact had been made and to stop further translation. 6,10,29 Note that the probe itself never made contact with the WE surface. Electrochemical (i.e., amperometry and voltammetry) measurements were performed in the confined area defined by the meniscus cell created between the SECCM probe tip and WE surface. Mapping was carried out using a standard hopping mode protocol, as per previous reports. 29,30 In brief, the SECCM probe was approached to the WE surface at a series of predefined locations in a grid ( Figure 1a) and, upon each landing, an independent electrochemical measurement was made, building up a spatially-resolved amperometric (isurf−t) or voltammetric (isurf−E) 'image' of the substrate. The final position of the z-piezoelectric positioner at approach was also used to build up a topographical map of the substrate (WE), synchronously.
During operation, the SECCM set up was placed on a vibration isolation platform (Minus K, U.S.A.) located within an aluminum faraday cage equipped with heat sinks and acoustic foam. The QRCE potential was controlled, with respect to ground and the current flowing at the WE (i.e., isurf), held at a common ground, was measured using a home-built electrometer. isurf was measured every 4 µs, and averaged in 256 blocks to give a data acquisition rate of 4 × (256 + 1) = 1028 µs (one extra iteration is used to transfer the data to the host computer). Data acquisition and instrumental control was carried out using an FPGA card (PCIe-7852R) controlled by a LabVIEW 2016 (National Instruments, U.S.A.) interface running the Warwick Electrochemical Scanning Probe Microscopy (WEC-SPM, www.warwick.ac.uk/electrochemistry) software.
The single-barreled SECCM probes were fabricated from borosilicate filamented capillaries (GC120F-10, Harvard Apparatus, U.S.A.) using a CO2-laser puller (P-2000, Sutter Instruments, U.S.A.). The pipet probes were highly reproducible, with tip diameters (dt) of approximately 200 nm (Figure 1b (Figure 1b-iii), characterized using electron microscopy. 31 Pulling parameters are included in the SI, Section S2. After the SECCM probes were filled with the analyte solution using a MicroFil syringe (World Precision Instruments Inc., U.S.A.), the QRCE (Ag/AgCl or Ag/AgI) was then inserted into the analyte solution, and mounted on the zpiezoelectric positioner, ready for use. Surface characterization. Details in the SI, Section S1. Data processing and analysis. Details in the SI, Section S1. Note that all electrochemical/topographical images are presented without any data interpolation.

RESULTS AND DISCUSSION
Selection of solvent and/or supporting electrolyte. The suitability of different solvent/supporting electrolyte combinations for deploying SECCM in aprotic solvent media was the first practical aspect to be explored. In order to do so, the ferrocene/ferrocenium (Fc 0/+ ) process was considered: Fc 0/+ is known to be a mechanistically simple, outer-sphere process in a range of solvents, characterized by rapid electrontransfer kinetics. 32,33 Indeed, previous SECCM studies (with nanometric pipet probes) have demonstrated that water-soluble Fc derivatives undergo ideal, reversible electron-transfer processes at metal and carbon electrodes. 8,34 Thus, by investigating the Fc 0/+ process at a topographically flat, uniformly active metallic electrode (nanocrystalline Au, herein), any deviation from ideality was taken as an indicator that the solvent/supporting electrolyte combination was not suitable for SECCM applications.
Linear-sweep voltammograms (LSVs) obtained by averaging 121 independent measurements in the SECCM configuration (dt ≈ 750 nm, Figure 1b-ii) are shown in Figure 2. Organic carbonates were chosen as "representative" aprotic solvents herein due to their widespread use in Li-ion battery electrolytes. 35 As shown Figure 2a, in propylene carbonate, PC (0.1 M tetra-nbutylammonium hexafluorophosphate, [NBu4][PF6]) the Fc 0/+ process is highly reproducible, giving rise to an ideal (Nernstian, vide infra) sigmoidal LSV with half-wave potential (E1/2) and steady-state limiting current (ilim) values of 0.413 ± 0.002 V vs Ag/AgCl quasi-reference counter electrode (QRCE) and ca. 15.1 ± 0.2 pA, respectively. Note that a relatively high voltammetric scan rate (υ) of 0.5 V s −1 was chosen to allow many measurements to be carried out on a short timescale, in line with previous LSV-SECCM mapping studies. 10,29 Note that the system cannot achieve a true steady-state on this rapid experimental timescale (evidenced by the slightly peak-shaped LSV in Figure 2a, indicative of a transient contribution to masstransport). Although a true steady-state could be approached through the use of a slower voltammetric scan rate or smaller pipet probes (vide infra), previous SECCM studies have shown that non-steady-state effects can be readily accounted for by performing complementary finite element method (FEM) simulations. 6,28 Also note that given the diffusion coefficient of Fc (DFc) in PC is ca. 3 × 10 −6 cm 2 s −1 (Ref 36), ilim is ca. 10% of that expected at the same sized microdisk electrode (ca. 200 pA), in agreement with previous SECCM reports. 6,28,34  [PF6]), the Fc 0/+ process is significantly less reproducible in the SECCM configuration, giving rise to distorted sigmoidal LSVs with E1/2 and ilim values of ca. 0.41 ± 0.02 V vs Ag/AgClQRCE and ca. 36 ± 3 pA. The shape of the average LSV is distorted relative to an ideal case (e.g., Figure  2a), particularly in the mass-transport controlled region (i.e., E > 0.5), which is particularly evident when comparing individual i−E curves, as shown in the SI, Section S3 ( Figure S1). The disparity between the Fc 0/+ response in DMC and PC is attributed to differences in the electrochemical meniscus cell stability brought on by the precipitation (crystallization) of [NBu4][PF6] at the SECCM probe tip during scanning. While this is not usually an issue for SECCM in aqueous media, where supporting electrolyte concentrations of up to 1 M have been reported, 16 DMC is relatively volatile (boiling point, BP = 91°C, vapor pressure (VP) = 55.4 mmHg at 25°C; solvent and supporting electrolyte properties summarized in SI, Section S1, Table S1) compared to water (BP = 100°C, VP = 23.8 mmHg) or PC (BP = 242°C, VP = 0.045 mmHg), and is therefore more prone to local supersaturation (within the meniscus cell) and eventual crystallization of the supporting electrolyte. This was confirmed by optical microscopy (OM), post-experiment (Figure 2b, inset), where the scan area appears to be made up of a solid crystalline mass (i.e., [NBu4][PF6]), rather than individual, well-defined droplet footprints.
Aside from using a higher BP solvent (e.g., PC vs DMC), another way to avoid the complications illustrated in Figure 2b is to employ a supporting electrolyte that is a room temperature ionic liquid (IL). 25 Shown in Figure 2c is the Fc 0/+ process in  (Figure 2c, inset), where individual well-defined droplet footprints can be seen. Another possibility is to mix the low BP (high VP) solvent with a high BP (low VP) solvent, as is commonplace in Li-ion battery electrolytes, where linear carbonates such as DMC are usually mixed with cyclic carbonates such as ethylene carbonate, EC (BP = 248°C, VP ≈ 0.01 mmHg) to take advantage of the low viscosity of the former and high BP and dielectric constant of the latter. 35 Thus, shown in Figure 2d is the Fc 0/+ process in a "typical" Li-battery electrolyte, 1:1 (w/w) EC:DMC (1 M Li[ClO4]). Again, the Fc 0/+ process is highly-reproducible, giving rise to sigmoidal LSVs with E1/2 and ilim values of 0.411 ± 0.002 V vs Ag/AgClQRCE and ca. 8.8 ± 0.2 pA, respectively, indicating that the meniscus cell is also stable in this mixed-solvent electrolyte system. It is also worth noting that the ilim values increase in the order 1:1 (w/w) EC: ), in accordance with the relative viscosities (η) of these electrolytes (SI, Section S1), 35 as per the Stokes-Einstein equation (i.e., D ∝ 1/η). 33,36 Electrochemical meniscus cell stability on the scanning timescale. The experiments carried out above confirm the viability of various aprotic solvent/supporting electrolyte combinations for SECCM on the minutes timescale (i.e., 121 measurements takes ca. 7 minutes), which is on par with the fastest reported LSV hopping mode scans (≈0.25 s pixel −1 or ≈11 minutes for a 2500 pixel scan). 10,29 As "typical" SECCM scans of >1000 pixels are usually on the hours timescale, 5 a larger scan was carried out (4700 measurements over ≈4.5 hours), again considering the Fc 0/+ process [Eq. (1)] on nanocrystalline Au. Note that PC (0.1 M [NBu4][PF6]) has been taken as a "representative" aprotic electrolyte in this and all subsequent SECCM experiments. As shown in Figure 3a, the LSVs obtained from the Fc 0/+ process are highly reproducible on the hours timescale, with E1/2 and ilim values of 0.411 ± 0.002 V vs Ag/AgClQRCE and 3.69 ± 0.02 pA, respectively. Note that a smaller SECCM probe (dt ≈ 200 nm, Figure 1b-i) was employed during this set of experiments, which is why ilim is approximately one quarter of that in Figure 2a (dt ≈ 750 nm, Figure 1b-ii). Also note that the higher mass-transport rate (≈proportional to 1/dt) 6,28 achieved with this smaller probe also results in a sigmoidal LSV that approaches a true steady-state on the experimental timescale (e.g., compare Figures 2a and 3a). As shown in Figure 3b, ilim is relatively unchanged over the ≈4.5 hour scan, while E1/2 changes by ca. 5 mV from ≈0.414 V to ≈0.409 V. In steady-state voltammetry, a change in E1/2 is usually ascribed to differences in electrontransfer kinetics (i.e., a more positive E1/2 indicates slower electron-transfer for an oxidation, vide infra). 37 In the present case however, the waveshape of the LSV in Figure 3a indicates that Fc 0/+ process is reversible at all points during the scan, confirmed by calculating the |E3/4 -E1/4| values (where E3/4 and E1/4 are the potentials where i = 3·ilim/4 and i = ilim/4, respectively), shown in Figure 3c. Evidently, the |E3/4 -E1/4| values satisfy Tomes criterion of reversibility 37 at all points (pixels) within the scan area, with a scan-average value of 58 ± 2 mV (mean ± one standard deviation). Thus, the change in E1/2 seen in Figure 3b is attributed to slight drift of the QRCE potential on the order of ≈1 mV hr −1 , which is comparable to what has previously been reported for Ag/AgCl QRCEs in aqueous electrolyte media, also in the SECCM configuration. 38 Application: microfabrication of conducting polymer arrays. Aside from its primary application as a tool for high-resolution electrochemical mapping (vide infra), 3,4 the SECCM meniscus "nanocell" configuration can be used for local decoration (i.e., microfabrication) at electrode surfaces. 5 This has been shown to be a particularly powerful approach for the synthesis and screening of three-dimensional (nano)structures, where CPs 19,20 or metal NPs 21,22 are locally electrodeposited and subsequently characterized with high-resolution microscopy to explore the relationship between deposition conditions and material structure/function in a high-throughput manner. In order to explore the viability of this approach in aprotic media, the electropolymerization of pyrrole (Py) to form oxidized polypyrrole (denoted [PPy] n+ , herein) was explored, shown below in Figure 4. While CPs (e.g., [PPy] n+ ) have found many modern electrochemical applications, for instance as electrodes in sensing 39 and energy storage technologies, 40 the understanding of the interplay between structure/properties (e.g., conductivity) and synthesis conditions is still largely empirical, 41 making highthroughput synthesis/screening strategies invaluable.   Figure 4b. For the sake of simplicity, and to check for consistency, the same E−t waveform (Eapp = +1.2 V for 200 ms) was applied at all points during the scan. Due to this and the fact that the structure of the nanocrystalline Au electrode is effectively homogeneous on the scale of the SECCM probe (dt ≈ 1.7 μm, Figure 1b-iii), the i−t curves are highly reproducible, with an average charge (Q) of 6.5 ± 0.8 pC (proportional to the amount of [PPy] n+ deposited) passed during the 200 ms pulse. Note that while a uniform E−t waveform was applied herein, E or t can be varied on a per-hop basis to systematically study the interplay between electrodeposition parameters and deposit morphology. 21 The SEM image of the SECCM scan area, shown in Figure  4c, confirms the reproducibility of the [PPy] n+ deposits, which are visible as a rectangular array of uniform black "dots" across the Au electrode surface. Zooming in on an area of the scan, shown in Figure 4d, it is clear that the black [PPy] n+ deposits are surrounded by a circular halo, which represents the total area (footprint) wetted by the SECCM meniscus cell during the E−t pulse (vide supra). The diameter of the droplet footprints (ca. 2.5 μm, Figure 4d) are comparable to the diameter of the SECCM probe (dt ≈ 1.7 μm, Figure 1b-iii), indicating that surface wetting is minimal (wetting ratio, W.R. = dfootprint / dt ≈ 1.5). Interestingly, while previous SECCM studies in non-aqueous IL-based electrolytes have also demonstrated a stable meniscus cell with minimal surface wetting, 24,25 those carried out in aprotic electrolytes based on PC 27 and dimethyl sulfoxide 26 have shown severe surface wetting, with W.R. values of ≈10 and ≈15, respectively. Evidently, under the conditions investigated herein, the meniscus cell-metal contact is very stable, with scanning spatial-resolution on the order of the tip diameter (ca. 200 to 1700 nm, herein, depending on the experiment/application, see Figure 1b), enabling precise electrochemical mapping with high spatial resolution (vide infra). Application: mapping nanoscale structure−activity. As alluded to above, SECCM is primarily employed as a high-resolution electrochemical mapping technique in correlative surface structure−activity studies. 3,4 The feasibility of this approach in aprotic media was explored herein by investigating the electro-oxidation of iodide (I − ) at polycrystalline Pt. I − /I2 is a well-known inner-sphere electron-transfer process 43 that proceeds via a oneelectron per iodide-ion process on inert electrode materials such as Pt or glassy carbon: I − and I2 combine homogeneously to form the polyhalogen complex anion, triiodide (I3 − ): The equilibrium (stability) constant of this reaction is highly solvent dependent, ranging from ≈10 2.9 in water to ≈10 7.8 in PC, and has a strong bearing on the observed iodide oxidation mechanism. 44 The overall oxidation process shown in Eq. (2) may occur in one step, giving rise to a single voltammetric process (e.g., in water) or in two steps (via an I3intermediate), giving rise to two separate voltammetric processes (e.g., in PC): The I − /I3 − couple is the most commonly employed redox mediator system in dye-sensitized solar cells (DSSCs), where Eq. (4) takes place (usually) at a platinized counter electrode. 45 More recently, the addition of I − (in the form LiI) to lithium-oxygen battery electrolytes has been found to reduce the charging overpotential by influencing the cathode product distribution, achieved through the action of the I − /I3 − redox mediator sys-tem. 46 Thus, understanding the surface structure-dependent kinetics/mechanisms of the I − /I3 − /I2 redox processes in aprotic media (e.g., PC) is important for the optimization of these devices. A spatially-resolved electrochemical movie obtained from the oxidation of I − at polycrystalline Pt is shown in the SI, Movie S1 (caption in the SI, Section S4). Despite the small magnitude of the measured currents (<5 pA), the high signalto-noise ratio allows the I − /I3 − [Eq. (4)] and I3 − /I2 [Eq. (5)] processes to be readily distinguished at lower and higher potentials, respectively. Comparison with the EBSD map in Figure 5a reveals that I − /I3 − and I3 − /I2 are both structure (grain) dependent processes. All of the grains within the scan area [labelled (1) to (4) in Figure 5a] can be described as relatively "high-index", with average crystallographic orientations that are far from the low-index orientations of face-centred cubic (fcc) Pt (i.e.,  Figure 5d) is again ca. 10% of that expected at the same sized microdisk electrode (ca. 60 pA). 28,34 Similar trends in grain-dependent I − oxidation kinetics were also observed on other areas of the polycrystalline Pt surface, shown in the SI, Movie S2 and Figure S2.
Although the X − /X3 − /X2 processes [where X − is a general halide (Cl − , Br − or I − ) or pseudohalide (SCN − or SeCN − ) anion] have been the subject of intensive research in many different solvent media, 44 no consensus has been reached on the electrode reaction mechanism(s). For instance, X3 − redox has been proposed to occur via either direct oxidation/reduction routes [e.g., Eqs. (4) and (5)] 47 or generalized CE mechanisms [e.g., Eq. (3) followed by Eq. (2)] 48 in non-aqueous media. From the different grain dependencies of I − /I3 − and I3 − /I2 shown in Figure 5b and c, respectively, it may be inferred that the two processes take place preferentially at different active sites, suggesting that the direct oxidation route predominates. Alternatively, the unique structural dependencies of the two processes may result from the potential-dependent specific adsorption of I − /I2, 49 which may influence the I − /I3 − oxidation sites in a grain-dependent manner.
In addition to the crystallographic orientation of the grains themselves, the I3 − /I2 process additionally exhibited enhanced activity in the vicinity of the interface between grains (1)/(2) and (4) (i.e., the grain boundaries). This can be seen by comparing the E1/2 (I3 − /I2) map (Figure 5c) to the SECCM topographical map (which was collected synchronously with the spatially-resolved electrochemical data), shown in Figure 5e, in which the individual grains can be distinguished topographically, arising from the polishing and flame annealing processes (detailed in the SI, Section S1). Clearly, there is a line of high activity running along the grain boundaries between grains (1)/(2) and (4), visualized by taking line-scan profiles of E1/2 and z-height, shown in Figure 5e. In the vicinity of the grain boundary (i.e., at x-displacement = 2 μm, indicated by blue arrows in Figure  5e), E1/2 is shifted negatively by ca. 0.01 V and 0.03 V compared to grains (1) and (4), respectively, indicating enhanced activity towards the I3 − /I2 process. Note that similar enhancements in activity towards inner-sphere processes have also been noted in previous SECCM studies, [12][13][14] but the ability to now perform such studies in aprotic solvent media greatly expands the range of systems that can be mapped with high spatial resolution.
It is also worth noting that in other areas of the polycrystalline Pt surface, non-uniform I − /I2 activity was observed within individual grains (i.e., at the intra-grain level). For instance, comparing the electrochemical movie in the SI, Movie S3 to the corresponding EBSD map in Figure 6a, it is clear that the electrochemical activity of grain (1), which makes up >95% of the scan area (labelled in Figure 6a), is non-uniform with regards to both the I − /I3 − and I3 − /I2 processes. The areas of high electrochemical activity surrounding grains (2) and (3)/(4), particularly for the I3 − /I2 process, coincide with regions of grain (1) with relatively high crystallographic misorientation, evident by comparing the misorientation map in Figure 6b [areas labelled (A) and (B)] with the E1/2 maps in Figure 6c and d. Note that the misorientation map indicates the local misorientation angle relative to the grain average, with a value ≥10° indicating a grain boundary (indicated by black lines in Figure 6a and b). In the present case, the two areas of relatively high misorientation [i.e., (A) and (B) in Figure 6b] also coincide with local topographical deformation, as shown in the SI, Figure S3. The different patterns of reactivity in Figures 6c and d also imply that the I − /I3 − and I3 − /I2 processes take place at different active sites on Pt (vide supra).

CONCLUSIONS
Taking organic carbonates as a model class of aprotic solvent, SECCM was successfully adapted to operate in non-aqueous electrolyte media. Investigating the simple outer-sphere Fc 0/+ process, it was shown that high BP (low VP) solvents (e.g., PC) give rise to robust and reproducible electrochemistry on the scanning (hours) timescale, while low BP (high VP) solvents (e.g., DMC) need to be combined with suitable low MP supporting electrolytes (e.g., ILs) or high BP solvents (e.g., EC) to avoid complications associated with salt precipitation/crystallization. SECCM was applied for microfabrication, by synthesizing (depositing) a microarray of [PPy] n+ on a nanocrystalline Au electrode surface, which was subsequently characterized with SEM to confirm the reproducibility of the technique and stability of the meniscus cell during scanning. Finally, high-resolution electrochemical mapping of the I − /I3 − and I3 − /I2 processes across the surface of a polycrystalline Pt electrode revealed unique, structure-dependent patterns of surface reactivity at for example, individual grains, grain boundaries and areas of high local surface misorientation. Overall, this study expands SECCM as a premier technique for structure−function−activity studies in (electro)materials science, opening up possibilities in exciting new areas such as the microfabrication of water-sensitive materials (e.g., electrodeposition of reactive metals, which will be facilitated by housing SECCM in a glovebox) or electrochemical mapping of processes relevant to electrochemical energy storage (e.g., Li-ion batteries and supercapacitors) and electrocatalysis.

Supporting Information
The Supporting Information (SI) is available free of charge on the ACS Publications website. Supplementary experimental section (Section S1), pipet pulling parameters (S2), additional experimental data (S3) and movie captions (S4), as described in the main text (PDF) Electrochemical movie showing the electro-oxidation of I − at polycrystalline Pt, Movies S1 (AVI), S2 (AVI) and S3 (AVI)