RESEARCH ARTICLE www.small-journal.com The Cleanroom-Free, Cheap, and Rapid Fabrication of Nanoelectrodes with Low zM Limits of Detection Gabriel Maroli, Vernalyn Abarintos, Andrew Piper,* and Arben Merkoçi* dropping mercury electrode,[4] where Nanoscale electrodes have been a topic of intense research for many decades. mercury is dropped through a small Their enhanced sensitivities, born out of an improved signal-to-noise ratio as capillary to constantly create a fresh electrode dimensions decrease, make them ideal for the development of electrode surface of a defined area at low-concentration analyte sensors. However, to date, nanoelectrode which measurements can be taken. Due to safety concerns over working with fabrication has typically required expensive equipment and exhaustive, mercury, these are rarely used nowadays. time-consuming fabrication methods that have rendered them unsuitable for More modern alternatives include inkjet widespread use and commercialization. Herein, a method of nanoband and screen-printed electrodes.[5–8] In such electrode fabrication using low cost materials and equipment commonly electrodes, nano-/microparticle inks are found in research laboratories around the world is reported. The materials’ patterned onto a substrate. To prevent ag- gregation, the nanoparticles need to be cost to produce each nanoband is less than €0.01 and fabrication of a batch stabilized in solution by adding bulky or- takes less than 1 h. The devices can be made of flexible plastics and their ganic ligands to make them stable col- designs can be quickly and easily iterated. Facile methods of combining these loids. The presence of these ligands can nanobands into powerful devices, such as complete three-electrode systems, impede subsequent electrode performance are also displayed. As a proof of concept, the electrodes are functionalized for and their removal post-printing is diffi- the detection of a DNA sequence specific to SARS-CoV-2 and found to display cult given their size and covalent attach- ment to the nanoparticles. The printed single molecule sensitivity. electrodes are incredibly rough and irre- producible, which contributes to irrepro- ducibility in the intended applications. In most cases, the devices are not conductive 1. Introduction immediately after printing and need to be “sintered,” a process by which the conductive particles morph to For many years now, electrochemists have been limited by the increase the contact area between them.[9,44] Printed electrodes choice of available electrodes. Traditional gold, silver, and carbon can be purchased commercially for several euros per electrode, electrodes are routinely used around the world and can be pur- with the exact price depending on the electrode type. chased for a few hundred euros each. However, these electrodes Alternatively, electrodes can be made in clean rooms. These require polishing by hand in alumina slurries, chemical cleaning facilities are expensive to build and their maintenance and in harsh, environmentally unfriendly solvents (such as piranha running costs are prohibitively expensive for many research solution), sonication, and electrochemical cycling in strong acids and commercial applications. Inside the clean room, conduc- (e.g., H2 SO4 ), in order to clean them before use.[1–3] This is all tive materials can be deposited on substrates in a highly con- done to try and get the electrodes as clean, flat, and reproducible trolled fashion with sub-nanometer resolution using chemical as possible. The oldest alternative to these electrodes is the or physical vapor deposition, atomic layer deposition (ALD), or electron beam evaporation (E-beam).[10,11] ALD is a surface- G. Maroli, V. Abarintos, A. Piper, A. Merkoçi controlled and self-limiting method for depositing thin films Catalan Institute of Nanoscience and Nanotechnology (ICN2) UAB Campus from gaseous precursors. This self-limiting growth mechanism Bellaterra, Barcelona 08193, Spain allows sub-nanometer-level control of film thickness with an E-mail: andrew.piper@icn2.cat; arben.merkoci@icn2.cat excellent conformity.[12] With E-beam, an electron beam is fo- G. Maroli cused onto a target metal under high vacuum, vaporizing the UIDI-CONICET Universidad Tecnológica Nacional metal in a high vacuum chamber where it can be deposited Buenos Aires C1041AAJ, Argentina in a uniform thin film on a chosen substrate.[13,14] The cost The ORCID identification number(s) for the author(s) of this article of these tools is in the range of hundreds of thousands of eu- can be found under https://doi.org/10.1002/smll.202302136 ros. Electrodes made by this method offer the best electrode © 2023 The Authors. Small published by Wiley-VCH GmbH. This is an purity and smoothness. Likewise, by fabricating in a clean- open access article under the terms of the Creative Commons Attribution room, photolithographic processes can be used to make a vari- License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. ety of electrode designs and architectures with incredibly high resolution. DOI: 10.1002/smll.202302136 Small 2023, 2302136 2302136 (1 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Figure 1. A) Schematic representation of the nanoband electrode fabrication showing i) the cleaned prepared substrate ii) after sputtering a nanometer thin layer of conductive electrode material through a mask iii) application of a capping layer iv) cutting to relieve a nanometer high edge electrode v) the completed device with a cross-section schematic (not to scale). B) Photograph of a device at step (iv), in this photograph the device was made with a PET substrate and Kapton tape capping layer. C,D) Scanning electron micrograph images of nanoband electrodes made with Au sputtered onto Kapton and capped with Kapton tape, cut with surgical grade scissors E) is an AFM image of the edge of one of the sputtered Au devices on Kapton (uncapped) with five corresponding step height measurements from left to right across the ridge in (F) showing that the Au thickness is ≈30–40 nm and the roughness of the Au surface is the same as the Kapton (single nanometers). G) CV of nanoband electrode, in 5 × 10−3 m potassium ferricyanide, 5 × 10−3 m potassium ferrocyanide, 100 × 10−3 m KNO3 recorded at a scan rate of 10 mV s−1 versus a Ag/AgCl reference electrode and Pt wire counter electrode. A nanoelectrode is defined as any electrode that has at least Here, we report a cleanroom-free method of nanoelectrode fab- one of its dimensions on the nanoscale.[15,16] Nanoelectrodes rication using sputter coaters. With this method, we are able to have many advantages over conventional, macroscopic elec- deposit metallic thin films with sub-nanometer resolution, out- trodes. Chief among which is their superior sensitivity.[17–20] This side of a cleanroom, using cheap equipment found in most re- improvement in sensitivity as the electrodes get smaller can search laboratories onto any substrate. These thin films are sand- be attributed to three factors: The first comes from their en- wiched between the substrate and a capping layer before being hanced mass transport, which allows them to measure a steady- cut to relieve a nanoband electrode, see Figure 1. These elec- state current; second, they have smaller double layers, which trodes are extremely cheap, with raw materials costing less than means that a greater proportion of the current is Faradaic; €0.10 per device, and can be cut multiple times in a matter of and finally, they have relatively smaller iR drops than larger seconds to relieve brand new, pristine electrodes that require no electrodes.[17,20,21] Recent studies have also shown that nanoscale cleaning prior to use. By cutting each device multiple times, the electrodes can be functionalized through the spontaneous for- cost of each electrode can be reduced to below €0.01. We show mation of thiolate self-assembled monolayers (SAMs), the most that these electrodes possess the characteristic electrochemical common method of electrode functionalization, in a matter of properties of nanoelectrodes, including better signal-to-noise ra- seconds; rather than the hours taken to functionalize macroscale tios and faster functionalization times than larger electrodes. As electrodes.[3] This means that sensor development and fabrica- a proof of principle, we show the superior performance of these tion can be rapidly sped up if done on nanoscale electrodes. electrodes in the detection of DNA with the same sequence tested At the time of writing, nanoelectrodes are typically either made for in the quantitative reverse transcription polymerase chain re- using complementary metal–oxide semiconductor fabrication action (PCR) diagnosis of SARS-CoV-2; where they exhibited su- techniques,[13,22] finely controlled chemical growth,[23] made in perior sensitivities and limits of detection to conventional elec- the form of nanowires[24,25] or grown on nanoscale templates.[26] trodes. These methods are either expensive or unsuitable for large-scale These electrodes are cheap, easy to fabricate, can be made in electrode manufacturing. any lab in the world with commonplace, affordable equipment, Small 2023, 2302136 2302136 (2 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com are more sensitive than the commonly used electrodes, and can to make five devices on each substrate, see Figure S2 in the Sup- be functionalized to make sensors in minutes. They therefore porting Information, the sheet resistances of the sputtered metal have the potential to revolutionize any research that uses elec- films on different substrate have been provided in Table S1 in the trodes. They are superior in every way to electrodes made by other Supporting Information, as well as that of E-beamed Au on Kap- methods and have the potential to replace existing electrodes im- ton, for reference. All the measured values are the same order of mediately, with little to no upfront capital costs. The faster func- magnitude (mΩ sq−1 ) with the E-beamed Au having a sheet resis- tionalization times and lack of required cleaning means that as- tance about half that of the sputtered Au. There is no significant say development can be rapidly sped up on these electrodes. Fi- difference between the resistances of the gold sputtered onto any nally, it is intended that the electrodes will allow more sensi- of the different substrates, this is pleasing as it evidences that the tive point of care sensors to be developed, allowing the diagno- substrate does not affect the electrochemical performance of the sis and prognosis of diseases with biomarkers too low in con- devices. centration for conventional sensors to detect, the earlier diagno- In order to create nanoelectrodes, the sputtered metal needed sis of diseases (when the concentration of biomarkers is lower), to be “capped” and cut. Several capping methods were tested, in- and perhaps even the amplification-free detection of nucleic acid cluding: covering with Kapton tape, spray coating with acrylic, biomarkers. and laminating with an office laminator. All were found to work, Figures S3–S6 in the Supporting Information, and may be suited 2. Results and Discussion to different applications. The requirements for a good capping layer are that it strongly adheres to both the electrode material 2.1. Electrode Fabrication and Characterization and the substrate, is chemically inert in the media in which it will be tested, forms a good physical and dielectric barrier, and does The nanoelectrodes in this project were fabricated by sputtering not damage the underlying electrode. It should also be easy to cut gold, using a conventional sputter coater, through a mask (sten- with the intended cutting method. In this project, we favored us- cil) onto various flexible plastic substrates. The substrates tested ing the Kapton tape because it was quicker to make devices with were polyimide (Kapton), polyethylene terephthalate (PET), and this than the other methods which need to be left to dry (acrylic) polyethylene naphthalate (PEN). These substrates were chosen or heated (lamination). because they are cheap, widely available, flexible, and easy to cut. Once the devices are capped, they must be cut to relieve The ability to form these electrodes on any vacuum stable sub- nanoband electrodes. The cutting was performed with scissors, strate, without any optimization of the fabrication, is a major ad- scalpels, or a paper guillotine, see Figure S4 in the Supporting vantage of this work over alternative methods. The masks used Information, all were capable of producing nanoband electrodes. in this project as stencils were made of aluminium or any of In order to make functioning electrodes, the blades in each of the the aforementioned plastics, theoretically they can be made out cutting methods need to be as sharp and clean as possible (e.g., of any material that does not interfere with the sputtering. The it is necessary to use surgical grade or microscopy scissors rather stencils are required to have sufficient weight to sit flush on the than office scissors). If the blade was blunt, it would cause a de- substrate and keep it in place during the evacuation of the sput- formation of the layers, Figure S7 in the Supporting Information. ter chamber. This is important to make electrodes reproducibly, Finally, the capping layers do not all set hard and should be cut as any lifting or deformation of the mask can change the sput- face down to drag the adhesive away from the electrode, so as not tered electrode dimensions. The ability to rapidly design, cut, and to cover it. use masks with different layouts allows a fast iterative approach Cyclic voltammograms (CVs) of each substrate type, with each for electrode design. This is not possible using photolithography capping layer and cutting method have been included in Figure where new masks take much longer to design and fabricate. In S6 in the Supporting Information. It is clear from the wave-like this study, a simple design motif of 1.2 cm wide and 2.4 cm long shapes of the CVs and the low capacitances of the electrodes that rectangles was used, Figure 1A,B and Figures S1 and S2 in the they are all nanoelectrodes. From Figure S6 in the Supporting Supporting Information. Information, the most reproducible cutting method was the pre- The rate of Au deposition by sputtering is well reported in the cision cutter, followed by the scissors and finally the scalpel. This literature, and our measurements of the electrode thickness as trend can be explained by the methods that allow for the most hu- a function of time and deposition current were found to be in man error causing the most irreproducibility. When cutting with agreement with other publications.[27,28] This method allows for the scalpel, the angle, pressure, and straightness of the cut are the controlled deposition of nanometer thin metallic films with more irreproducible than with the other cutting methods. sub-nanometer precision; using cheap equipment that is rou- Regarding the capping layers, the acrylic is the most repro- tinely found in most laboratories around the world, since sputter ducible owing to the fact that it sets the hardest. The softer setting coaters are used in scanning electron microscope (SEM) sample adhesives of the tape and melt layer of the laminate are prone to preparation. In this work, we sputtered metallic layers 50–100 nm being smeared on the electrodes during cutting. This is why it is thick. The rationale behind this was to make the devices thick important to cut the electrodes in a fashion that drags the adhe- enough to ensure a good connection through a crocodile clip on sive away from the electrode. Although some damage is still pos- the bond pad, to ensure a homogenous coverage of the substrate sible, which is why the precision cutter gave not only the most re- and to have the same thicknesses as similar cleanroom-fabricated producible cuts, see Figure S6 in the Supporting Information, but nanoelectrodes present in the literature.[3,17,18,27–29] also consistently yielded electrodes with a higher area; as there is The deposited metal (in this case Au) was initially tested with a less adhesive on the electrodes. Examples of the adhesive smear- four-point probe to assess its conductivity. The mask allowed us ing can be seen in Figure S8 and the video in the Supporting Small 2023, 2302136 2302136 (3 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Information. In all experiments, the substrates tested had no dis- where C is the capacitance in Farads, 𝜖 0 is the dielectric permit- cernible impact on the final electrode quality. tivity of a vacuum, 𝜖 r is the dielectric permittivity of the mea- The time taken to make a single batch of electrodes, includ- surement solution, A is the area of the electrode, and H is the ing sputtering, capping, and cutting was less than 1 h. Once the separation distance between the plates in the model. In aqueous devices are made, fresh electrodes can be made by cutting in systems, H can be taken as the inner Helmholtz plane (the di- a matter of seconds. This is a crucial advancement over other ameter of a water molecule in pure water).[31] It is possible to ap- cleanroom-free nanoelectrode fabrication methods such as those proximate the electrochemical surface area of the nanoband elec- reported by the White group,[30] in which substrates are sput- trodes using Equation (1), since the Cdl follows the electrode sur- ter coated through a mask and then encapsulated in resin which face very accurately and the other values can all be taken from the takes 24 h to dry, before being mechanically polished for hours literature for pure water (𝜖 0 = 8.854 × 10−12 F m−1 ; 𝜖 r = 78.3; H to create pristine flat electrodes. A video showing the entire fab- = 2.75 Å).[31] In these experiments, 75 nm thick electrodes were rication process has been made to accompany this publication. sputtered, with a geometric electrode surface area of 9 × 10−4 The equipment used is commonly found in most laboratories mm2 . We calculated the area of our electrode, from its voltammet- around the world but the up-front capital costs are also very low ric capacitance, Figure S9 in the Supporting Information, to be if they do need to be purchased. Table S2 in the Supporting Infor- 13.5 × 10−4 ± 0.6 × 10−4 mm2 , within experimental error the mation summarizes these capital costs and the cost of consum- same as the geometric area. These data further evidence the ables used for electrode fabrication. If the upfront equipment nanoscale nature of the electrodes. Likewise, the standard devi- costs are excluded, then the materials’ cost per electrode is below ation was obtained from three electrodes cut on a single device, €0.01, on the condition that the devices are cut a minimum of ten showing the reproducibility of the cutting method and the uni- times each, to yield at least ten electrodes per device. This makes formity of the sputtered Au over the entire device. The slightly these electrodes cheap and therefore inherently disposable, it is larger electrochemical surface area than geometric surface area cheaper to cut and create a new electrode rather than trying to is attributed to surface roughness and heterogeneity.[36] It is satis- clean and re-use them. The ability to cut one device several times fying that our so-called “roughness factor” is 1.5, which is better to make multiple electrodes is a novel advantage of these devices. than reported for other electrode types in the literature.[37] In this The devices can take up to an hour to make but cutting and mak- system, we postulate that as the capping layer adhesive is setting, ing a fresh nanoelectrode takes seconds. it may move (during handling), dragging the Au layer out of the To confirm that the fabricated electrodes are truly nano in na- device, much like the filling in a sandwich when it is pressed. ture, the bands were imaged by SEM, Figure 1C,D, the height This theory is based on the SEM data, Figure 1D, where films ap- of the uncapped sputtered metal edges was measured by atomic pear to protrude from the device. This is further supported by the force microscopy (AFM), Figure 1E,F, and the nanoscale elec- findings that the signals from the hardest setting capping layers trochemical responses of the bands were evaluated. The SEM (acrylic) gave the most reproducible electrodes. Also, anecdotally, images clearly show a nanoscale conductive band between the the electrode reproducibility could be improved by leaving any of substrate and capping layer. Using the SEM software (xT micro- the capping layer types for extended periods of time (days/weeks scope Control), the height of this band could be estimated and depending on the capping material), to allow them to set harder. was found to be between 40 and 45 nm, which was slightly below To further characterize the electrochemical response of the the 50 nm targeted sputtering thickness of these devices. This nanobands, they were analyzed by electrochemical impedance slight variation is attributed to experimental errors in the sput- spectroscopy. The Nyquist plot from these experiments has been tering since the AFM measurements of uncapped Au layers sput- provided in Figure S12 in the Supporting Information, within an tered onto Kapton also had a step height of about 40 nm. inset showing the equivalent circuit to which it has been fit. This Typical CVs of macroelectrodes have peaks that form as the is the established equivalent circuit for nanoelectrodes.[3,15,35] The current becomes limited by the mass transport of the redox fitted values of the fit, reported in Table S5 in the Supporting In- molecule to the electrode surface. The response of nanoelec- formation, include an Rct of 29 980 Ω ± 3.5% (error from the trodes is different,[24,31] because the electrodes are so small, fitting). This was used to determine an electron transfer rate of a hemispherical diffusion profile is rapidly established.[15,32] 8.03 cm s−1 , the derivation of which is included in Figure S13 in Therefore, waves rather than peaks are observed in the CVs the Supporting Information, which is in agreement with other of nanoscale electrodes.[33–35] The nanoband electrodes devel- nanoelectrodes in the literature.[32,35,38] oped in this project exhibit electrochemical responses typical of nanoscale electrodes, Figure 1G. Likewise, the current ranges ob- served are in good agreement with nanobands of similar dimen- 2.2. Different Electrode Designs sions in similar buffers reported by other groups.[30] Another interesting electrochemical feature of the nanoelectrodes is the One of the advantages of making electrodes using this methodol- extremely low capacitances that they display, in this system the ogy is the ability to rapidly iterate different electrode designs and only capacitance present is the double layer capacitance (Cdl ). The configurations. New plastic masks can be made quickly (in under magnitude of the Cdl is given by the Helmholtz equation, Equa- 20 min) and easily using cutting plotters (or by hand) to change tion (1)[31] the pattern and size of the metallic coatings to suit different ap- plications. Likewise, it is very easy to combine electrodes of differ- ent designs and types to create customized devices for different 𝜀0 𝜀r A applications. It is easy to fabricate a library of electrodes of differ- C = (1) ent patterns and thicknesses, sputtered from different materials. H Small 2023, 2302136 2302136 (4 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Figure 2. Schematics (not to scale), photographs, and SEM images of the different LEGO brick designs fabricated and tested in this project. A) Sticking completed devices on top of each other: i) a cleaned substrate, ii) is sputter coated with a nanometer thin film of Au, iii) before being capped, and iv) the bond pads cut so that each will open to the environment in the final 3D device, v) multiple separate devices are stacked and stuck to each other, vi,viii) before the three nanobands are exposed by cutting. viii) A photograph of one such completed device, ix) as well as a schematic and x) SEM of the Small 2023, 2302136 2302136 (5 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com These individual components can then be easily combined to de- there is a clear decrease in the limiting current, Figure 3. This velop new devices for different applications, including those with is because the charge transfer at the electrode is impeded by the different numbers of electrodes, such as two-electrode systems steric blocking of the redox molecules by the SAM film, as well as or for multiplexed sensing applications. Much like LEGO bricks, the electrostatic repulsion of the anionic Fe species by the nega- these can be attached together to create an endless possibility of tively charged phosphate backbone of the DNA. A control experi- designs. As a proof of principle for this “LEGO-brick” concept, ment was run to assess the stability of the SAM over the timescale complete three-electrode systems have been developed by a num- of an assay, Figure S11 in the Supporting Information, in this ber of different methods. The authors would like to stress that control experiment, the functionalized nanobands were dipped this is simply a proof of concept and that an endless number into solutions of DNA-free water. In the main experiment, the of designs are possible. Photographs, schematics, and SEMs of functionalized electrodes were first placed in a solution of deion- these different methods of combining devices have been shown ized water and then left to sit in the measurement solution for in Figure 2. First, it is possible to stick multiple capped devices 20 min to make sure the probe monolayer was also stable. Af- on top of each other using adhesives, see Figure 2A. Second, it ter which, it was placed consecutively in solutions containing in- is possible to build the devices up vertically by sputtering directly creasing concentrations of target DNA before being rinsed with, onto the capping layers, using masks to separate the different and measured in, the measurement solution. The washing was bond pads for the different layers, Figure 2B. Third, it is possible important to remove any nonspecifically bound DNA from the to design a mask in which the three electrodes are in the same 2D surface. As the target DNA sequence binds to the probe layer, it plane and simply separated from each other by the mask design, further impedes the charge transfer at the electrode surface caus- Figure 2C. The CVs of these devices, Figure S10 in the Supporting ing a decrease in current that is proportional to the amount of Information, show that there is a shift in the half-wave potential target DNA bound to the surface. (from +0.175 to +0 V) when switching from a Ag/AgCl reference The CV and square wave voltammetry data from these ex- electrode to a sputtered Au pseudoreference electrode, as would periments on macro and nanoband electrodes are provided in be expected. Other than this, there is no difference in the obtained Figure 3. The Langmuir isotherm was chosen as the simplest electrochemical signal, proving that functioning complete three- and most appropriate model to fit these data to and from the electrode systems can be produced by this “Lego-brick” method. fits, the technical limit of detection (LoD) of the macroelectrodes was found to be 92.4 × 10−12 ± 30.4 × 10−12 m. However, the 2.3. DNA Detection performance of the nanoband electrodes is significantly differ- ent, showing a clear sensitivity enhancement with a LoD of In order to assess the sensing performance of the electrodes, they 97.1 × 10−21 ± 23.7 × 10−21 m (1 zM = × 10−21 m) obtained were functionalized with an ssDNA probe for the detection of a in one of the repeats. While these limits of detection are very DNA sequence specific to SARS-CoV-2. The authors would like low, there is evidence in the literature of other nanoelectrode to stress that this is a proof-of-concept experiment, intended to and nanoelectrode ensembles exhibiting similar sensitivities, in- ascertain the performance of the nanoband electrodes and com- cluding for DNA detection.[39–42] Avogadro’s number is 6.022 × pare them to existing commercial electrodes. As such, it was 1023 , meaning that a 10 × 10−21 m solution will contain 6000 deemed suitable to detect a synthetic complimentary DNA se- molecules per liter. Therefore, a 2 mL solution of target DNA quence rather than RNA from real samples and the authors wish (as was being used here) should theoretically contain 12 target to stress that no claims are being made that an amplification- DNA strands. This means that with an LoD of 97 × 10−21 m, as free SARS-CoV-2 sensor has been developed. The mode of op- few as 117 molecules in the 2 mL sample can be detected. How- eration of the sensor is shown schematically in Figure 3. First, ever, the authors would emphasize caution on this front, these the electrochemical response of a clean (bare) electrode is mea- target DNA concentrations were made up by serial dilutions and sured in a solution of 5 × 10−3 m K[Fe(CN) 6 ]3− and 5 × 10−3 there is obviously an error introduced by the pipetting. Table S4 m K[Fe(CN) 6 ]4− in 1 × phosphate-buffered saline (PBS) with in the Supporting Information shows the error in the final DNA 10 × 10−3 m MgCl2 . Next, the electrodes were functionalized by concentration when different pipetting errors are taken into ac- the spontaneous formation of a mixed self-assembling mono- count. Even a 1% error (which is low for even the best of pipettes) layer from an aqueous solution containing 6-mercaptohexan-1-ol could result in significantly more (or less) DNA in the low target and a thiol-C6-modified ssDNA probe. This process takes a min- DNA concentrations than is quoted here. Therefore, the authors imum of 2 h on macroscopic electrodes but was possible in less would like to emphasize that the sensitivity of these electrodes than 2 min on the nanoband electrodes. This is in agreement with is thought to be in the low zM range, based on these data and previous publications and further evidences the nanoscale prop- are not claiming single molecule detection limits. There are also erties of these electrodes.[3] When placed back in the redox agent- concerns about whether or not 20 min is sufficient for such low containing solution and the electrochemistry was re-measured, amounts of analyte to diffuse to the electrode surface and bind to cross-section have also been provided. B) Sputtering directly onto the capping layers. i) A cleaned substrate, ii) is sputter coated with a nanometer thin Au coating through a mask that also patterns the contact pad, iii) it is then capped, iv–vi) before sputtering and capping are repeated, using different masks to separate the contact pads, vii) the devices are then cut to relieve the nanobands. It is possible to leave the top layer uncapped so when dipped in solution its area is much larger than the other electrodes, allowing it to be used as a counter electrode. ix) A photograph and x) cross-sectional schematics and xi) an SEM image of this type of device. C) 2D devices. i) A cleaned substrate is sputter coated ii) with Au through a mask that patterns separate devices on the substrate surface. These can then be iii) capped and iv) cut to relieve multiple electrodes with variable shapes and orientations in the same plane. v) Photographs and vi) cross-sectional schematics and vii) SEMs have been provided. Small 2023, 2302136 2302136 (6 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com 2302136 (7 of 11) www.advancedsciencenews.com Small 2023, 2302136 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com the probes. We propose that bulk solution movements, such as biomarkers too dilute to currently detect. This could include, but convection and the agitation of the solution when the electrodes is not limited to, biomarkers that cross the blood brain barrier or are inserted by hand, account for there being a sufficient amount cross the placenta from a fetal to a mother’s bloodstream. Like- of probe target hybridization in the 20 min tested. wise, as has been shown by the detection of DNA here, it may be What is clear, however, is that the sensitivity of these nanoband possible to replace PCR tests with amplification-free DNA sen- electrodes is far superior to that of the macroelectrodes and com- sors using this technology. parable to the best results for nanoelectrodes reported in the literature.[40] The variation in the nanoelectrode signal, at first glance, appears to be greater than that of the macroelectrodes. 4. Experimental Section However, it should be noted that the x-axis is logarithmic, and the pipetting errors are larger for the lower analyte concentrations, Materials and Equipment: A Mini Sputter Coater (model SC7620, Quo- therefore the spread of the data cannot be solely attributed to vari- rum Technologies) and the gold sputter target were purchased from ANAME Instrumentación Científica (Madrid, Spain). The three substrates ation in the electrodes which were previously shown to be very used were PET (thickness 75 μm), PEN (thickness 125 μm), and Dupont reproducible. The difficulty in reproducibly making the calibra- grade Kapton HN (thickness 75 μm), all were purchased from Goodfellow tion solutions with such low concentrations means that it is not GmbH (Spain). The capping layers were 3 m Kapton tape 5413(Digi-Key), possible to obtain linear ranges from the data shown in Figure 3F. Acrylic Varnish CRC – Clear Brilliant (RS Components Spain), and Fellowes The only conclusions that can be drawn are that these nanoband Enhance 80 micron lamination sheets (Amazon, Spain). The cutting was electrodes are extremely sensitive, capable of detecting DNA performed with a flintronic A4 Paper Cutter (Amazon, Spain), scalpels, down at low zM levels. Until methods of reproducibly making or surgical scissors purchased from VWR, Spain. Aluminum masks were made in house. A Graphtec ce6000-40 cutting plotter was used to fabricate solutions with such low concentrations of analyte are developed, masks out of the aforementioned Kapton that was also used as a substrate. accurate calibration curves will be impossible to obtain. Potassium hexacyanoferrate (III) (K3 [Fe(CN)6 ]) and potassium hexacyanoferrate (II) 3-hydrate (K4[Fe(CN)6]·3H2 O) were purchased from Panreac AppliChem. Tris(2-carboxyethyl)phosphine (TCEP), 2- 3. Conclusions mercaptohexanol, 2-propanol (IPA), acetone, and PBS, potassium nitrate (KNO 3 ), and magnesium chloride (MgCl2 ) were purchased from Merck Herein, we have reported the cleanroom-free fabrication of KGaA, Germany. Gold working electrodes (CH101), Ag/AgCl reference nanoband electrodes. The nanoscale nature of the electrodes has electrodes (CH111), and platinum wire counter electrodes (CH115) were been proven electrochemically as well as by SEM and AFM. These purchased from CH instruments. Au/Au/Au screen-printed electrodes devices are extremely cheap, with raw materials’ cost of €0.01 were purchased from DropSense, Spain. The 0.3 μm alumina low viscosity per electrode and can be made immediately in most labs around polishing slurry (ET034), 0.05 μm alumina low viscosity polishing slurry the world using commonplace, low-cost equipment. The elec- (ET033), micropolishing cloths (ET032), and glass polishing slides (ET031) were purchased from eDAQ (Poland). Concentrated sulfuric trodes exhibit superior performance compared to conventional acid (96%, ITW, Spain) and hydrogen peroxide 50% (Sigma Aldrich, electrodes, not just in terms of cost, but they are easy to manu- Spain) were used to make piranha solution for electrode cleaning. An facture in customizable design motifs, require no cleaning, can Ovan Ultrasonic 3L bath was purchased from Ovan (Spain). All aqueous be made out of a wide range of materials, and are extremely solutions were prepared using deionized water from a Milli-Q Advantage sensitive. In the proof-of-principle DNA sensor shown here, the A10 Water Purification System with 0.22 μm filters MPGP04001 (18.2 MΩ nanoband electrodes were able to achieve low zM levels of detec- cm, Merck-Millipore, Spain). The electrochemical measurements were carried out with a Metrohm Autolab PGSTAT12 with NOVA 2.1 software. tion; markedly better than the pM detection limits of the same All the data analysis was performed in the NOVA 2.1 software or Origin system on a macroscopic electrode. We believe that given these 2018, unless stated otherwise. advantages, this work and these electrodes have the potential Nanoelectrode Fabrication—Sputtering: The chosen substrate to revolutionize the fields of electrochemistry and point-of-care (Kapton, PEN or PET) was cut into 6 cm x 6 cm squares with the sensing. The low LoDs of these devices means that they can be cutting plotter. These dimensions were chosen to fit underneath used to develop sensors for biomarkers present in samples at very the mask and inside the sputter coater. The mask pattern was cre- ated using AutoCAD software, version 2021 from Autodesk. The low concentrations. Methods of making the electrodes suitable design was created with 0 mm strokes to generate a vector file. for real world sample analysis, such as microfluidics and antifoul- The Kapton masks were produced using a Graphtec ce6000-40 ing strategies will be the focus of future work. plotter, while the aluminum masks were machined using a Haas vf For decades now, commercial point-of-care biosensors have 1 CNC milling machine. The substrate was cleaned by soaking in acetone been limited to the detection of highly abundant biomarkers such for 5 min, rinsed under a stream of IPA to remove the acetone before as glucose, lactate, progesterone, and virions. It is hoped that the being completely immersed in a bath of IPA for another 5 min. Finally, lower LODs of these cheap nanoband electrodes will kick start the the substrate was rinsed with ultrapure water. Note, poor cleaning can affect the adhesion of the sputtered metal to the substrate. The development of biosensors for a wide range of low-concentration cleaned substrate and the mask were placed in the sputter coater with biomarkers. This could not only lead to the earlier diagnosis of a preloaded Au sputter target. Vacuum was applied until a pressure diseases, when the biomarkers are less abundant, but also to <0.8 mbar was achieved. The chamber was flushed three times with the development of point-of-care devices for disease states with Ar to remove any remaining oxygen. The Ar inlet was opened and a Figure 3. A) Schematic illustration of how the DNA sensing mechanism works. B) Overlayed CVs of the DNA sensing system response to different target DNA concentrations on a macroelectrode. Overlayed with Square Wave Voltammetry (SWV) data from the experiments, shown in (C). The results of the same experiment, with lower DNA target concentrations on nanoband electrodes are shown in (D) and (E). The conditions for these experiments are set out in detail in the experimental section. F) The calibration curves from the macro and nanoelectrode repeats with corresponding fits to the Langmuir isotherm. The reciprocal of the peak signal was normalized between 0 and 1 prior to plotting and fitting. Small 2023, 2302136 2302136 (8 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com potential was applied to create the argon plasma and begin sputtering. vacuum conditions. In all the figures presented, the electron beam inten- By controlling the gas pressure in the chamber (with the inlet valve), the sity, working distance, and magnification were stated. current could be altered, which determined the target metal deposition AFM images were obtained through Molecular Imaging’s PicoPlus rate. In this project, a 10 mA current was chosen and a deposition rate modular Scanning Probe Microscope (SPM) system in combination with of 5 nm min−1 was produced. The duration of the sputtering could the PicoScan Controller and magnetic MAC Mode. PicoView 1.20 software then be used with this deposition rate to make devices any desired was used during data acquisition but the analysis and corrections were thickness. Once the sputtering process was finished, the chamber was performed in Gwyddion 2.61. vented and the substrate was removed, see the Supporting Information Electrochemical Characterization: The electrodes were characterized video. by CV, square wave voltammetry, and electrochemical impedance spec- Capping—Adhesive tape: The freshly sputtered device was fixed onto a troscopy. Unless stated otherwise, the bare electrodes were tested in solu- flat surface. The tape was aligned and a small portion was stuck to the work tions of 5 × 10−3 m potassium ferricyanide, 5 × 10−3 m potassium ferro- surface to aid the alignment. The tape was then dragged across the surface cyanide, and 1x PBS versus a silver/silver chloride reference electrode and of the device by hand, under constant pressure to avoid the formation of Pt wire counter electrode. CVs were run between −0.1 and + 0.45 V at a bubbles. The excess tape was removed and the devices were taken onto scan rate of 10 mV s−1 , for a total of four cycles. The square wave voltam- the next step. metry was run in the same set up immediately after the CVs from −0.15 Spray Coating: The device was fixed onto a flat surface. A plastic mask to +0.6 V versus the open circuit potential with a 9 mV step, a modula- was used to define the area to be spray coated. The spray coating (in tion amplitude of 20 mV, and a frequency of 2.5 Hz. The electrochemical this project acrylic) was applied as per the manufacturer’s instructions, impedance spectroscopy was run with a dc voltage equal to the measured by shaking vigorously before use and spraying from 20 cm above the de- open circuit potential, with a 10 mV sinusoidal AC voltage over a frequency vice. The spray coating was repeated thrice to ensure a complete covering range of 100 kHz to 0.1 Hz, recording 10 frequencies spread logarithmi- of the devices before they were left to dry under ambient conditions for cally per decade. 20 min prior to use. DNA Sensing: The DNA probe was a thiol tagged primer approved by Lamination: The freshly sputtered devices were placed on a sheet of the Centre for Disease Control (CDC) for the detection of SARS-CoV-2. The paper. A pre-patterned lamination sheet (PET, with an ethylene vinyl ac- sequences were ordered from Merck (Spain) and were as follows: etate, EVA, inner coating) with pieces was cut out to leave exposed contact Probe 5´- ACCCCGCATTACGTTTGGTGGACC-(C6 SH) - 3’ pads on the finished devices, was aligned and placed onto the devices. The Target 5´- GGTCCACCAAACGTAATGCGGGGT - 3’ ensemble was then run through a Lamigator IQ (Renz, Spain), to laminate The electrodes were placed in an aqueous solution containing 30 × 10−6 the devices in a hot lamination process where the EVA acted as a melt ad- m 6-mercaptohexanol (MCH), 150 × 10−6 m TCEP, and 1.5 × 10−6 m of hesive. the probe DNA. The macroelectrodes were left in this solution for the self- Cutting: The different cutting methods investigated in this project assembled monolayer to form for 2 h, whereas the nanoelectrodes were were to cut with a scalpel, scissors, and a flintronic A4 paper cutter. left for 2 min, prior to rinsing with a stream of water to remove physisorbed The blades of each device were rinsed with IPA and dried with N2 prior thiols. to each cut. After repeated use, the blades could get blunt and either Before and after probe film formation, the electrodes were placed in a needed to be replaced or sharpened. It is vital to have the capping lay- solution of 1x PBS with 10 × 10−3 m MgCl2 , 5 × 10−3 m potassium hex- ers facing away from the direction of travel of the blade to avoid any acyanoferrate (III) and 5 × 10−3 m potassium hexacyanoferrate (II). CVs adhesive being dragged over the working area of the nanoband being were run between −0.1 and + 0.45 V versus a Ag/AgCl reference electrode created. (macro) or Au pseudoreference electrode (nano) and Pt wire counter elec- Scalpel Cutting: The devices were fixed face down (with tape) on a trode, at a scan rate of 10 mV s−1 , for a total of four cycles. The square Model craft cutting mat (RS components, Spain). A clean scalpel was used wave voltammetry was run in the same set up immediately after the CVs to cut the devices by hand by applying as even and uniform a pressure as from −0.15 to +0.6 V versus the open circuit potential with a 9 mV step, a possible with the blade perpendicular to the mat in a single swift stroke modulation amplitude of 20 mV, and a frequency of 2.5 Hz. using the markings on the cutting mat to guide the alignment of the cut. After the probe film was formed the functionalized electrodes were Scissor Cutting: Microscopy (surgical) grade scissors were needed as placed in solutions containing increasing concentrations of DNA target, conventional office scissors are not sharp enough for this application. The for 20 min each. Between target incubations, the electrodes were placed devices were placed face down with the region that was intended to be cut in the measurement solution, left for 20 min, and the electrochemical mea- resting on the bottom blade of the scissors. The upper blade was closed surements were repeated. The DNA target concentrations tested were dif- quickly and smoothly keeping it as perpendicular to the device as possible. fered for the macro and nanoelectrodes. For the macroelectrodes, the tar- Flintronic Paper Cutter: The devices were stuck face down on the cut- get DNA concentrations tested were: 0 m, 1 × 10−15 , 1 × 10−12 , 500 × ter surface, using the alignment markings on the surface to ensure the 10−12 , 1 × 10−9 , 500 × 10−9 , and 1 × 10−6 m. For the nanoelectrodes, the blade was cutting as straight across the device as possible. The blade was target concentrations tested were 0 m, 10 × 10−21 , 100 × 10−21 , 500 × pushed down and dragged through the device at a uniform speed and 10−21 , 1 × 10−18 , and 1 × 10−15 m. pressure. In this instance, it was held perpendicular to the device by the A control experiment was run where the functionalized nanoelectrodes design of the tool. were placed in MQ water rather than the increasing DNA concentrations, Sheet Resistance of the Metals: The sheet resistances of the Au sput- for the same number of repeats. A second control was run where the func- tered onto different substrates were measured using a Keithley DMM6500 tionalized electrodes were exposed to noncomplimentary DNA target se- multimeter in a 4-point probe mode. Pogo pins with a pressure of 120 g quences. and a separation of 2.54 mm were used. As is required for determin- Lego-Brick Concept: The different fabrication methods for the different ing the sheet resistance of thin layers, a correction factor of 4.53 and 2D stacking designs (Lego brick) assemblies of the nanoband electrodes a geometric correction factor of 0.78 were used when performing these are described below. In all these cases, the substrate cleaning, sputtering, experiments.[43] and capping were performed as described above. The e-beamed samples used as a control were sputtered onto a 4″ Kap- 2D Design: A mask was made out of Kapton, see Figure S2B in the ton substrate using the same mask as was used in the sputter coater. Supporting Information, in which three electrodes could be patterned on 35 nm of Au was deposited through the mask onto the substrate using a single device, running parallel to one another down the length of the an ATC-8E Orion evaporator (AJA International Inc., USA). device. The three electrodes were able to be used as a working electrode, AFM and SEM: The samples were mounted on SEM sample stubs pseudoreference electrode, and counter electrode. In the mask shown in using double sided conductive carbon tape (TED Pella, INC), these were Figure S2B in the Supporting Information, the electrodes were designed then loaded into a Thermo Fisher (formerly FEI) Quanta 650 FEG ESEM to be the same size, 3 mm wide and 21 mm long with 1.5 mm separations for image acquisition using the backscatter electrons detector under low between them. Small 2023, 2302136 2302136 (9 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Lego-Brick Concept 1: In this method, the individual devices were capped with the acrylic spray. Multiple devices could be stacked on top [1] G. Henihan, H. Schulze, D. K. Corrigan, G. Giraud, J. G. Terry, of each other with the acrylic acting as an adhesive between them. In this A. Hardie, C. J. Campbell, A. J. Walton, J. Crain, R. Pethig, K. E. system, each of the devices performed the function of a working, counter, Templeton, A. R. Mount, T. T. Bachmann, Biosens. Bioelectron. 2016, or reference electrode. In these experiments, the acrylic was needed to be left for longer to dry (overnight) under ambient conditions. The connection 81, 487. pads of each device could be cut with scissors to allow easy connection af- [2] D. K. Corrigan, H. Schulze, G. Henihan, A. Hardie, I. Ciani, G. Giraud, ter stacking. J. G. Terry, A. J. Walton, R. Pethig, P. Ghazal, J. Crain, C. J. Campbell, K. Lego-Brick Concept 2: In this method, new electrodes were sputtered E. Templeton, A. R. Mount, T. T. Bachmann, Analyst 2013, 138, 6997. directly onto the capping layer of the underlying device. Three mask de- [3] A. Piper, D. K. Corrigan, A. R. Mount, Electrochem. Sci. Adv. 2021, 2, signs (Figure S2C, Supporting Information) were used, one for each layer. e2100077. The masks were designed to isolate the connections of each layer in the [4] A. Frumkin, A. Donde, Z. Phys. Chem. 1926, 123, 339. final device. The final layer, sputtered to be the counter electrode, was not [5] K. Kuribayashi, H. Matsumoto, H. Uda, Y. Komatsu, A. Nakano, S. capped, so that when dipped into a solution, the electrode area would be Ikegami, Jpn. J. Appl. Phys., Part 1 1983, 22, 1828. far greater than that of the working electrode. [6] D. Mandler, S. Kraus-Ophir, J. Solid State Electrochem. 2011, 15, 1535. [7] D. H. Craston, C. P. Jones, D. E. Williams, N. Elmurr, Talanta 1991, Supporting Information 38, 17. Supporting Information is available from the Wiley Online Library or from [8] N. Komuro, S. Takaki, K. Suzuki, D. Citterio, Anal. Bioanal. Chem. the author. 2013, 405, 5785. [9] Y. Sui, C. A. Zorman, J. Electrochem. Soc. 2020, 167, 037571. [10] N. M. Hemed, A. Convertino, Y. Shacham-Diamand, Sens. Actuators, Acknowledgements B 2018, 259, 809. [11] R. K. Gupta, A. Periyakaruppan, M. Meyyappan, J. E. Koehne, Biosens. ICN2 is funded by CERCA programme, Generalitat de Catalunya Grant Bioelectron. 2014, 59, 112. SEV-2017-0706 funded by MCIN/AEI/ 10.13039/501100011033. Funding [12] B. C. Mallick, C. T. Hsieh, K. M. Yin, Y. A. Gandomi, K. T. Huang, ECS for this project was obtained from Grant PID2021-124795NB-I00 funded J. Solid State Sci. Technol. 2019, 8, N55. by MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe”. V.A. acknowledges the support and funding by the Department [13] D. E. Presnov, I. V. Bozhev, A. V. Miakonkikh, S. G. Simakin, A. S. of Science and Technology – Science Education Institute (DOST-SEI) of the Trifonov, V. A. Krupenin, J. Appl. Phys. 2018, 123, 054503. Philippines. G.M. would like to express his gratitude to the Carolina Foun- [14] N. Elfstrom, R. Juhasz, I. Sychugov, T. Engfeldt, A. E. Karlstrom, J. dation for financial support through the scholarship “Doctorado 2020.” Linnros, Nano Lett. 2007, 7, 2608. G.M. acknowledges Universitat Autonoma de Barcelona (UAB) for the [15] A. Piper, A. Mount, Electrochemical Characterisation of Microsquare possibility of performing this work inside the framework of Chemistry Nanoband Edge Electrode (MNEE) Arrays and Their Use as Biosensors, PhD Programme. V.A. acknowledges Universitat Autonoma de Barcelona University of Edinburgh, Edinburgh 2017. (UAB) for the possibility of performing this work inside the framework of [16] I. Kleps, A. Angelescu, R. Vasilco, D. Dascalu, Biomed. Microdevices Biotechnology PhD Programme. 2001, 3, 29. [17] A. Piper, B. Alston, D. Adams, A. R. Mount, Faraday Discuss. 2018, 201, 201. Conflict of Interest [18] E. J. F. Dickinson, R. G. Compton, J. Electroanal. Chem. 2011, 661, The authors have no competing interests to declare. 198. [19] M. Silvestrini, L. Fruk, P. Ugo, Biosens. Bioelectron. 2013, 40, 265. Author Contributions [20] H. L. Woodvine, D. A. R. Mount, Development and Characterisation of Microelectrode and Nanoelectrode Systems, The University of Edin- G.M. and V.A. contributed equally to this work. G.M. and V.A. contributed burgh, Edinburgh 2011. equally to the acquisition and analysis of the data as well as the paper [21] J. G. Terry, I. Schmueser, I. Underwood, D. K. Corrigan, N. J. Freeman, writing. A.P. conceived the original idea, supervised the acquisition of the A. S. Bunting, A. R. Mount, A. J. Walton, IET Nanobiotechnol. 2013, 7, data, acquired the data, analyzed the data, wrote the manuscript, and led 125. the project throughout. A.M. supervised the project and contributed to [22] N. Couniot, T. Vanzieleghem, J. Rasson, N. Van Overstraeten- drafting the paper. Schlogel, O. Poncelet, J. Mahillon, L. A. Francis, D. Flandre, Biosens. Bioelectron. 2015, 67, 154. Data Availability Statement [23] X. Zhu, Y. Qiao, X. Zhang, S. Zhang, X. Yin, J. Gu, Y. Chen, Z. Zhu, M. Li, Y. Shao, Anal. Chem. 2014, 86, 7001. The data that support the findings of this study are available from the cor- [24] D. W. M. Arrigan, Analyst 2004, 129, 1157. responding author upon reasonable request. [25] Y. Wang, J.-M. Noël, J. Velmurugan, W. Nogala, M. V. Mirkin, C. Lu, M. Guille Collignon, F. Lemaître, C. Amatore, Proc. Natl. Acad. Sci. USA 2012, 109, 11534. Keywords [26] S. M. D. Watson, H. D. A. Mohamed, B. R. Horrocks, A. Houlton, biosensors, low cost, nanobands, nanoelectrodes, single molecule detec- Nanoscale 2013, 5, 5349. tion [27] M. Schwartzkopf, G. Santoro, C. J. Brett, A. Rothkirch, O. Polonskyi, A. Hinz, E. Metwalli, Y. Yao, T. Strunskus, F. Faupel, P. Müller- Received: March 13, 2023 Buschbaum, S. V. Roth, ACS Appl. Mater. Interfaces 2015, 7, 13547. Revised: June 28, 2023 [28] M. Schwartzkopf, A. Hinz, O. Polonskyi, T. Strunskus, F. C. Lö, V. Kö, Published online: P. Mü Ller-Buschbaum, F. Faupel, S. V Roth, F. C. Löhrer, V. Körstgens, Small 2023, 2302136 2302136 (10 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202302136 by Readcube (Labtiva Inc.), Wiley Online Library on [30/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com P. Müller-Buschbaum, F. Faupel, S. V. Roth, ACS Appl. Mater. Inter- [37] E. F. Douglass Jr., P. F. Driscoll, D. Liu, N. A. Burnham, C. R. Lambert, faces 2017, 9, 5629. W. G. McGimpsey, Anal. Chem. 2008, 80, 7670. [29] C. S. Smith, C. M. Bowie, K. D. Song, H. Yoon, V. K. Varadan, W. K. [38] P. Sun, M. V. Mirkin, Anal. Chem. 2006, 78, 6526. Kim, Proc. SPIE 2011, 7980, 798011. [39] X. M. Bin, E. H. Sargent, S. O. Kelley, Anal. Chem. 2010, 82, 5928. [30] R. B. Morris, D. J. Franta, H. S. White, J. Phys. Chem. 1987, 91, 3559. [40] R. Gasparac, B. J. Taft, M. A. Lapierre-Devlin, A. D. Lazareck, J. M. Xu, [31] A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and S. O. Kelley, J. Am. Chem. Soc. 2004, 126, 12270. Applications, Wiley, New York 1980. [41] I. Ciani, H. Schulze, D. K. Corrigan, G. Henihan, G. Giraud, J. G. [32] H. L. Woodvine, J. G. Terry, A. J. Walton, A. R. Mount, Analyst 2010, Terry, A. J. Walton, R. Pethig, P. Ghazal, J. Crain, C. J. Campbell, T. 135, 1058. T. Bachmann, A. R. Mount, Biosens. Bioelectron. 2012, 31, 413. [33] K. Ngamchuea, S. Eloul, K. Tschulik, R. G. Compton, J. Solid State [42] A. P. Wettasinghe, N. Singh, C. L. Starcher, C. C. DiTusa, Z. Ishak- Electrochem. 2014, 18, 3251. Boushaki, D. Kahanda, R. McMullen, E. A. Motea, J. D. Slinker, ACS [34] E. J. F. Dickinson, R. G. Compton, J. Phys. Chem. C 2009, 113, 17585. Sens. 2021, 6, 2622. [35] I. Schmueser, A. J. Walton, J. G. Terry, H. L. Woodvine, N. J. Freeman, [43] S. Yilmaz, J. Semicond. 2015, 36, 082001. A. R. Mount, Faraday Discuss. 2013, 164, 295. [44] G. Maroli, S. Boyeras, H. Giannetta, S. Pazos, J. Gak, A. R. Oliva, [36] C. L. Alexander, B. Tribollet, M. E. Orazem, Electrochim. Acta 2015, M. A. Volpe, P. M. Julian, F. Palumbo, Frontiers in Electronics, 2023 3 173, 416. https://doi.org/10.3389/felec.2022.1060197. Small 2023, 2302136 2302136 (11 of 11) © 2023 The Authors. Small published by Wiley-VCH GmbH