Exploring carbon particle type and plasma treatment to improve electrochemical properties of stencil-printed carbon electrodes

Abstract

Stencil-printing conductive carbon inks has revolutionized the development of inexpensive, disposable and portable electrochemical sensors. However, stencil-printed carbon electrodes (SPCEs) typically suffer from poor electrochemical properties. While many surface pretreatments and modifications have been tested to improve the electrochemical activity of SPCEs, the bulk composition of the inks used for printing has been largely ignored. Recent studies of other carbon composite electrode materials show significant evidence that the conductive carbon particle component is strongly related to electrochemical performance. However, such a study has not been carried out with SPCEs. In this work, we perform a systematic characterization of SPCEs made with different carbon particle types including graphite particles, glassy carbon microparticles and carbon black. The relationship between carbon particle characteristics including particle size, particle purity, and particle morphology as well as particle mass loading on the fabrication and electrochemical properties of SPCEs is studied. SPCEs were plasma treated for surface activation and the electrochemical properties of both untreated and plasma treated SPCEs are also compared. SPCEs displayed distinct analytical utilities characterized through solvent window and double layer capacitance. Cyclic voltammetry (CV) of several standard redox probes, FcTMA⁺, ferri/ferrocyanide, and pAP was used to establish the effects of carbon particle type and plasma treatment on electron transfer kinetics of SPCEs. CV of the biologically relevant molecules uric acid, NADH and dopamine was employed to further illustrate the differences in sensing and fouling characteristics of SPCEs fabricated with different carbon particle types. SEM imaging revealed significant differences in the SPCE surface microstructures. This systematic study demonstrates that the electrochemical properties of SPCEs can be tuned and significantly improved through careful selection of carbon particle type and plasma cleaning with a goal toward the development of better performing electrochemical point-of-need sensors.

Graphical Abstract

1. Introduction

Carbon electrodes are widely employed in sensing applications due to their favorable electrochemical activity, high conductivity, extended solvent window, biocompatibility, rich surface chemistry, relative inertness and low cost [1–3]. More recently, a significant focus has been placed on the development of easy to fabricate and inexpensive carbon electrodes for point-of-need (PON) sensing. Carbon electrodes such as glassy carbon (GC), pyrolytic graphite, or boron doped diamond (BDD) are difficult to fabricate, thus, their use is limited in PON sensing. In order to improve ease of fabrication and lower costs, carbon composite electrodes (CCEs), made from conductive carbon particles held together by an inert binder have been the electrode of choice [2, 4, 5]. A variety of binder materials have been used for composite electrode fabrication including high molecular weight polymers such as poly(methyl methacrylate) and poly(caprolactone), referred to as thermoplastic electrodes (TPEs) [2, 4, 5], low molecular weight hydrocarbons such as mineral oil referred to as carbon paste electrodes (CPEs) [6–8] and mixtures of materials to generate printable inks referred to as screen/stencil printed carbon electrode (SPCEs) [9, 10].

A variety of CCEs exist which are differentiated by the physical properties the binder material imparts on the final electrode [2, 4, 8, 11–13]. For example, after homogenizing a dissolved polymer and graphite powder, TPEs form a material with a “gum-like” consistency that can then be pressed into a template until excess solvent has evaporated, forming a solid electrode. The solid TPE is mechanically robust and able to withstand abrasive polishing and sanding [2]. Alternatively, the low molecular weight hydrocarbons used in CPEs remain a liquid throughout the lifetime of the electrode, rendering CPEs mechanically fragile. After homogenization, CPEs are packed into electrode bodies by hand, reducing interelectrode reproducibility. The surface can be easily refreshed by removing outer layers of the paste, eliminating more complicated pretreatments that may be required for surface renewal after electrochemical cycling. While both CPEs and TPEs are inexpensive, can be fabricated simply and have high electrochemical activity, their use in PON electrochemical sensing has remained limited by lack of disposability, mass producibility and facile integration with portable sensing platforms such as paper-based analytical devices.

Screen or stencil-printed electrodes (SPCEs) have transformed the development of electrochemical PON sensors. SPCEs are characterized by low cost, small size, low sample volume requirement, disposability - precluding the need for treatments to return an electrode to its original state - good sensitivity, and mass producibility [14, 15]. SPCEs are a composite of various substances including graphite, carbon black, solvents, and a polymeric binder such as polyvinyl chloride, however, the exact composition of commercial inks is proprietary [15, 16]. Prefabricated, commercial SPCEs are used in a significant portion of sensor development due to the availability of commercial flow cells and ease of integration with portable instrumentation allowing for more seamless commercialization and field use. SPCEs are also printed in-house, most commonly using commercially available conductive carbon inks since their thixotropic properties are required for screen or stencil printing [17–19]. A squeegee is used to print inks onto the desired substrate either through a mesh screen or stencil. After printing, the electrodes are typically cured in an oven, generating solid electrodes. In this way, 100s of electrodes can be produced quickly using low cost materials. In-house printing of SPCEs is critical in portable sensor development because the inks can be bulk modified with electrocatalysts or biorecognition elements prior to printing, precluding the need for post-fabrication modifications which add complexity and cost to sensor development. Moreover, the inks can be printed in many planar geometries onto microfluidic paper-based analytical devices, which can carry out tasks such as mixing, timing, and sequential fluid delivery, further improving assay automation in PON sensing [20–23]. While SPCEs have been a critical component in the development of sophisticated PON sensors, SPCEs suffer from poor electrochemical activity, higher detection limits, and lower sensitivity than other CCE types [2, 13, 24, 25].

SPCE electrochemistry can be improved through pretreatment or surface modification with nanomaterials. Pretreatments include plasma treatment [26, 27], soaking in organic solvents [28], exposure to UV/ozone [29], and electrochemical cycling in various solvents [30–32]. Nanomaterial modification of SPCEs via drop-casting is widely used to control the electrochemical active portion of the electrode surface and resulting electrochemistry. Common nanomaterials include metallic nanoparticles [33–35], carbon nanotubes [36, 37], carbon black [38, 39] and graphene/graphene oxide [40, 41]. While these methods have been successful, pretreatment and/or post-fabrication modification to improve the basic electrochemical functionality of a disposable, single-use, electrode is not ideal. On the other hand, it is well known that particle characteristics including size, purity, and aspect ratio as well as the carbon to binder ratio are related to CCE performance, with some electrode compositions providing as good or better electrochemistry than commercial electrodes such as GC, Au, or Pt macro-electrodes [2, 24, 42]. The electrochemical activity of both prefabricated SPCEs and SPCEs made in-house with commercially available inks from a variety of vendors have been characterized, however, little is known about the exact composition of commercial SPCEs and its relationship to electrochemical and physical quality [14–16, 40, 43]. Recently, we reported the fabrication of glassy carbon microparticle SPCEs (GC-SPEs) and applied them to detecting Cd and Pb. In this work, GC-SPEs demonstrated dramatic improvement for ASV of Cd and Pb compared to graphite SPCEs due to the relationship between carbon particle and electrochemical response [44].

Here, we report a systematic study on the relationship between SPCE bulk composition, specifically carbon particle type and carbon particle mass loading, and the physical and electrochemical properties of SPCEs. Different carbon particles were added to a starting commercial carbon ink that contains an unknown carbon nanomaterial, but has required the addition of graphite for use in electrochemical sensing in previous works. Such a study is relevant to further understand and improve the quality of SPCEs without increasing time or fabrication cost. We explore graphite particles, glassy carbon microparticles and carbon black which vary in size, microstructure, purity and vendor for SPCE fabrication. Plasma treatment is carried out and optimized for each SPCE type. Plasma treatment was chosen since it is effective in activating the electrode surface, yet it is simple, fast and can be performed on several electrodes simultaneously. Electrochemical characteristics of both untreated and plasma treated SPCEs was determined with several inner and outer sphere redox species, double-layer capacitance, and aqueous solvent window. Herein, a benchmark for researchers wishing to work with SPCEs which have significantly higher electrochemical activity than previous SPCE iterations, while maintaining low background currents and wide solvent windows, without complex and time-consuming modification procedures is provided.

2. Experimental

2.1. Solutions

All solutions were prepared fresh daily using 18.2 M Ω・cm water purified using a Milli-Q system (MilliporeSigma, USA). To test the electrochemical characteristics of all SPCEs, solutions were prepared in 0.1 mol L⁻¹ potassium chloride (KCl, Fisher Scientific, New Jersey, USA) with 1 mmol L⁻¹ potassium hexacyanoferrate(III) (K₃Fe(CN)₆) and potassium hexacyanoferrate(II) trihydrate (K₄(CN)₆・3H₂O (Sigma-Aldrich) as well as 1.0 mmol L⁻¹ ferrocenylmethyl trimethylammonium hexafluorophosphate (FcTMA⁺, synthesized in-house) [45]. 0.5 mmol L⁻¹ p-aminophenol (pAP, Sigma-Aldrich) 1.0 mmol L⁻¹ ascorbic acid (AA, Sigma-Aldrich), 1.0 mmol L⁻¹ beta-nicotinamide adenine dinucleotide disodium salt (NADH, Sigma-Aldrich), 1.0 mmol L⁻¹ uric acid (UA, Sigma-Aldrich), and 1 mmol L⁻¹ dopamine hydrochloride (DA, Sigma-Aldrich) were all prepared in 0.1 mol L⁻¹, pH 7.4 phosphate buffer. Phosphate buffer was prepared by titrating a 0.2 mol L⁻¹ potassium phosphate monobasic (KH₂PO₄, Sigma-Aldrich) solution with 1.0 mol L⁻¹ sodium hydroxide (NaOH, Sigma-Aldrich) to a pH of 7.4. All chemicals were used as received and were of analytical grade.

2.2. SPCE materials and fabrication

Five carbon types, listed in Table 1, were used to fabricate SPCEs. 0.4 to 12 μm glassy carbon (GC) microparticles (spherical, Type 1) were purchased from Alfa-Aesar (Massachusetts, USA). 2 to 12 μm GC microparticles were purchased from Sigma-Aldrich, acetylene carbon black powder was purchased from Strem Chemicals (Massachusetts, USA), graphite powder (< 20 μm, sigma graphite) was purchased from Sigma-Aldrich, and 3569 graphite powder was purchased from Asbury Carbons (New Jersey, USA). To prepare the composites, the carbon types were hand-mixed with commercial carbon ink (E3178) purchased from Ercon Inc. (Massachusetts, USA) in the ratios indicated in Table 1. To prepare the stencil, a poly(ethylene) terephthalate (PET) sheet, purchased from 3M (Minnesota, USA) was laser cut using a 30 W Epilog Engraver Zing CO₂ laser cutter and engraver (Colorado, USA). The PET stencil was taped to another PET sheet and a squeegee was used to print the carbon inks. After printing, the SPCEs were cured in a 65 °C oven for 30 min. A previously reported three electrode design was used for the stencil [44, 46]. The solution reservoir was fabricated by sandwiching a PET sheet between two layers of double sided adhesive (3M) and laser cutting. The reservoir was attached to the SPCEs with the double sided adhesive and aligned with the connection pads to define the electrode area reproducibly at 0.112 cm² as reported previously [44]. SPCE plasma treatment was carried out with an LF-5 system purchased from Mercator Control Systems INC. (CA, USA) at a power of 125 W and a pressure of 500 (± 50) mTorr for various times specified in the relevant figures.

Table 1.

Carbon particle types and their characteristics used to fabricate SPCEs.

Carbon Particle Type	Particle Size	Purity	Manufacturer	Ink to Carbon Ratio (g)
Alfa GC	0.4 to 12 μm	Ash content 0.0042%	Alfa-Aesar	0.8 to 1.0
Sigma GC	2 to 12 μm	99.95%	Sigma	0.9 to 1.0
Carbon Black	42 nm	99.9%	Strem Chemicals	3.7 to 0.5
Sigma Graphite	< 20 μm	Unavailable	Sigma	3.5 to 1.5
3569 Graphite 1 & 2	150 μm	99.8%	Asbury Carbon	3.5 to 1.5 (1) & 3.5 to 2.5 (2)

2.3. SPCE electrochemical characterization and imaging

All electrochemical experiments were performed using a CH Instruments 660B potentiostat (Texas, USA) at room temperature (22 ± 1 °C). The reference electrode (RE) was either a saturated calomel electrode (SCE) or Ag/AgCl ink (Gwent Group, Torfean, UK) painted directly onto the SPC RE and allowed to dry at room temperature. The RE used is denoted in each figure. Cyclic voltammograms (CVs) were recorded in 1.0 mmol L⁻¹ solutions of FcTMA⁺, Fe(CN)₆⁴⁻/Fe(CN)₆³⁻, AA, DA, UA, and NADH and in 0.5 mmol L⁻¹ solutions of pAP at a scan rate of 100 mV s⁻¹. Solvent windows were recorded in 0.1 mol L⁻¹ KCl (pH 6.5) at a scan rate of 200 mV s⁻¹. Solvent windows were also recorded in 0.1 mol L⁻¹ KCl (pH 6.5) at a scan rate of 100 mV s⁻¹ for quantification of the potential range. Capacitance was quantified by recording CVs in 0.1 mol L⁻¹ KCl (pH 6.5) between +0.1 and −0.1 V for a total of 5 cycles. Scanning electron microscopy images of the electrodes were acquired with a JSM-6500F field emission scanning electron microscope (JEOL, Tokyo, Japan) with a 2 kV acceleration voltage.

3. Results and discussion

3.1. SPCE fabrication and imaging

To explore the effects of carbon particle type on the electrochemical response of SPCEs several carbon types were selected for SPCE fabrication. Carbon particles were selected in a range of particle sizes, microstructures and purities as indicated in Table 1. Two different graphite particles, Sigma graphite and 3569 graphite were selected since graphite is by far the most common carbon particle type used for CCE fabrication. The Sigma graphite SPCE used here is the same composition of SPCE employed in several previous works so as to compare the performance to the 3569 graphite SPCEs developed here [46–49]. Carbon black is often used to modify SPCE surfaces; therefore, it was selected to determine its suitability in bulk SPCE formulations. Two GC microparticles, Alfa GC (0.4 to 12 μm) and Sigma GC (2 to 12 μm) were selected due to GC’s distinct physical and electrochemical properties and recent successful use in SPCEs [44]. 10 to 20 μm Afla GC microparticles were characterized in our previous work, and were not studied further here because of their poor electrochemical performance in SPCEs [44].

Prior to further characterization, the carbon to binder ratio was optimized, since conductive particle loadings have significant implications on composite electrode performance [2, 4]. The optimal carbon to binder ratio is physically limited by the amount of carbon particle that can be homogenized with the ink as well as the final consistency of the composite for stencil printing. If too thin, the ink will easily leak between the stencil boundary and the substrate. On the other hand, if the ink is too thick, it is difficult to achieve a defect-free surface. Particle mass loading is also important for improving the conductivity of composite electrodes since the contact resistance between particles as well as the number of contact points delimits composite electrode conductivity [2, 6, 15]. Since commercial inks are typically employed for SPCE fabrication, carbon particle mass loading and the effects on electrochemical properties of SPCEs has not been studied. However, ohmic resistance characterized by slow apparent electron transfer rates and substandard electrochemical properties is a significant drawback to working with conventional SPCEs [15, 43, 50]. For these reasons, particle mass loading was optimized in order to maximize carbon particle loading while retaining stencil-printing capabilities. Among the graphitic carbon and carbon black particles, particle size significantly contributed to mass loading capabilities. As indicated in Table 1, carbon black particles are 42 nm while the 3569 graphite particles are 150 μm. The maximum carbon black to binder ratio was limited to 0.5 g to 3.7 g. However, the maximum mass loading of the 3569 graphite was 2.5 g graphite in 3.5 g binder (3569 graphite 2). The Sigma graphite particles, with an intermediate size of ≤ 20 μm, exhibited maximum mass loading at a graphite to binder ratio of 1.5 g to 3.5 g. A 3569 graphite fabricated with a mass ratio of 1.5 g carbon to 3.5 g ink was used to compare to the Sigma graphite SPCE with the same mass loading and is referred to as 3569 graphite 1. As particle size increases, higher mass loadings can be achieved. The dependence of mass loading on particle size has been seen with other CCEs as well [2]. In contrast to the graphite and carbon black particles, GC microparticles required less binder, exhibiting optimal particle to binder ratios of 1.0 g to 0.8 g for Alfa GC and 1.0 g to 0.9 g for Sigma GC microparticles.

Following fabrication, scanning electron microscopy (SEM) was used to image SPCE surfaces. Representative SEM images are shown in Fig. 1. Here, the different carbon particles and the resulting SPCEs are characterized by significant morphological differences. Graphite flakes can be seen clearly in the 3569 graphite (1 and 2) and the Sigma graphite images with a relatively thick coating of the binder present on the graphite surfaces. The 3569 graphite 1 SPCE did not show any visible differences at lower magnifications compared to the 3569 graphite 2 SPCE and is provided in the supporting information, Fig. S1. However, at higher magnifications, the lower mass loading 3569 graphite 1 SPCE exhibited more charging effects, voids between carbon particles, and appears to be less densely packed (Fig. S8), consistent with the electrochemical characterizations discussed below. An additional difference between the 3569 graphite 1 and 2 (larger) and Sigma graphite (smaller) particles is the orientation of particles observed in the SEM images. The 3569 particles in both the high and low mass loading versions appear horizontally aligned and densely packed together while the Sigma graphite particles appear randomly oriented and less densely packed. This is further supported in CV experiments discussed below, in which the 3569 graphite 1 and 2 SPCEs exhibited less ohmic resistance, indicating a higher number of particle contacts than the Sigma graphite SPCEs. The morphology of the carbon black SPCE surface is characterized by significant cracking as indicated in the SEM image. Some cracks in the electrode surfaces were also visible with the naked eye. At higher curing temperatures, some commercial inks used for SPCEs have generated electrode surfaces with cracks due to the rate of evaporation of solvent components, thus, the carbon black SPCEs were cured at room temperature as opposed to 60 °C to determine if a defect free surface was achievable [51]. Decreasing the curing temperature did not reduce cracking and this affect is likely a result of the small particle size.

Fig. 1.

SEM images of 3569 graphite 2 (500x), Sigma graphite (1,500x), carbon black (80x), Alfa GC (2,000x), and Sigma GC (2,000x) SPCE surfaces. SPCE types are denoted in the images.

In contrast to the graphite particles, GC microparticles SPCE surface contained much less binder on the surface. This is expected since GC is non-porous and little to no penetration of particles with binder occurs resulting in weaker particle-binder interactions [52, 53]. However, graphite is soft and porous, so binder components can physically penetrate and stick to graphite particles, resulting in the thick binder coating observed. A similar phenomenon was observed when GC and graphite paste electrodes were compared [54]. This observation is also consistent with the higher mass loading necessary for GC particles compared to graphite particles. While the Alfa GC (0.4 to 12 μm) and the Sigma GC (2 to 12 μm) particles have a similar manufacturer reported size distribution, SEM imaging (Fig. S2) revealed that the Sigma GC particles have a narrower size distribution with less frequency of particles greater than 10 μm. The effects of this difference in particle size distribution on the electrochemical characteristics of the GC-SPEs is discussed in the following sections.

3.2. SPCE Plasma Treatment and Stability

SPCEs typically suffer from poor electrochemical characteristics due to the presence of non-electroactive components of the binder as well as other impurities on the electrode surface.[28, 30] Several pretreatments have been adopted to etch polymeric binder and other impurities from the SPCE surface to further expose the electrochemically active particles including oxygen [26] or argon [27] plasma treatment, electrochemical (galvanostatic or cycling) [30, 55], and soaking in NaOH [32] or organic solvents [28] for tens to hundreds of minutes. Both electrochemical and chemical pretreatment are time consuming, labor intensive and carry a risk of contamination of the electrode surface via adsorption of impurities [56, 57]. Therefore, we selected plasma treatment as a simple and fast method to etch surface binder and/or impurities and expose the underlying carbon particles at SPCE surfaces.

To optimize plasma treatment parameters, SPCEs were plasma treated at the same pressure (500 mTorr) and plasma power (125 W) for varying time intervals of 1, 1.5 and 2 min. Carbon black SPCEs were plasma treated for time intervals of 0.5, 1 and 1.5 min due to significant increases in surface cracking at longer treatment times. The optimal treatment time was selected based on the voltammetry of FcTMA⁺, Fe(CN)₆⁴⁻/Fe(CN)₆³⁻ (ferri/ferrocyanide) and pAP assessing anodic peak to cathodic peak separation (ΔE_p) and peak current (Fig. S3 – S5). Based on these parameters, the optimal plasma treatment time was two min for the Sigma GC, Alfa GC, 3569 graphite 1 and 2, and Sigma graphite SPCEs and 30 s for carbon black SPCEs. The carbon black SPCE electrochemistry became worse at longer treatment times, likely due to excessive etching of binder verified by visibly larger surface cracks.

CV was carried out with FcTMA⁺, ferri/ferrocyanide and pAP at increasing time intervals after plasma treatment to ensure the electrochemical responses were stable, with the longest time tested being 10 days (Fig. S6 – S7). Since the electrodes were stored in ambient conditions, some deactivation resulting in larger ΔE_p is expected due to adsorption of small hydrocarbons to the surface [3, 58]. Increased ΔE_ps were obtained for nearly all plasma treated SPCEs with FcTMA⁺, ferri/ferrocyanide and pAP after 10 days of aging. However, the Sigma GC-SPE only exhibited a significant increase in ΔE_p for pAP. While ΔE_p increased to varying degrees for each electrode, the peak currents for all redox mediators remained stable for at least 10 days, with any discrepancies attributable to batch-to-batch reproducibility (Table S1 and S2). Plasma treated SPCEs can be stored for long periods of time without requiring a reactivation procedure due to airborne hydrocarbon contamination, however, care should be taken in some electrochemical experiments to ensure currents are being measured at the appropriate potential. Attention can also be given to electrode storage conditions to reduce any surface contamination of both bare and plasma activated SPCEs from exposure to air including storage in an air tight container or at low temperatures [58]. Fig. S8 shows SEM images of the plasma treated SPCEs compared to their untreated counterparts. Briefly, binder is visibly etched from the graphitic and carbon black SPCE surfaces. However, the untreated GC surfaces are not as significantly coated with the binder (Fig. 1), and the binder is not visibly removed after plasma treatment. The resulting activation of the GC-SPEs then is likely due to removal of other adsorbed impurities.

3.3. Capacitance and potential window

An advantage to working with composite electrodes is the small contributions double layer capacitance (C_dl) makes to background currents. C_dl is determined by the fraction of electroactive carbon exposed to solution; typically a small fraction of the geometric surface area of native CCEs [3]. To determine optimal carbon particle type, the effects of carbon particle type and plasma treatment on SPCE C_dl were assessed. Fig. 2 shows CVs recorded in 0.1 mol L⁻¹ KCl between −0.1 V and +0.1 V for untreated (dashed lines) and plasma treated (solid lines) SPCEs.

Fig. 2.

Averaged CVs (N = 3) recorded in 0.1 mol L⁻¹ KCl at a scan rate of 100 mV s⁻¹ over the potential range −0.1 V to 0.1 V for (a) 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon black (grey) and (b) Alfa GC (red) and sigma GC (blue) at untreated (dashed) and plasma (solid) (ai and bi) and untreated only (aii and bii).

C_dl values were estimated for all SPCE types and plasma treatment times using Equation 1 [59].

$$C_{dl}=\frac{i_{\mathit{average}}}{v{A}_{\mathit{geometric}}}$$ Equation 1

Here, i_average is the absolute value of current (A) recorded between 0.1 V and −0.1 V, v is scan rate (V s⁻¹) and A_geometric is geometric surface area (cm²). Greater currents in the CVs shown in Fig. 2 for plasma treated electrodes are expected since etching and removal of binder coating exposes a higher surface area of carbon particles [3]. C_dl also increases with increasing plasma treatment times, but to varying degrees for each carbon type. C_dl values of 5.9 ± 1.1 and 6.8 ± 3.0 μF cm⁻² were determined for the untreated Alfa GC and Sigma GC-SPEs, respectively and 13.2 ± 3.0, 3.3 ± 2.1, 4.8 ± 2.2, and 9.0 ± 2.2 μF cm⁻² for the untreated Sigma graphite, 3569 graphite 1, 3569 graphite 2, and carbon black SPCEs, respectively. Untreated SPCE C_dl are lower than C_dl values for conventional glassy carbon electrodes (24 – 36 μF cm⁻²), edge plane graphite electrodes (60 μF cm⁻²), bare commercial DropSens SPCEs (37 μF cm⁻²), as well as CPEs (62 μF cm⁻²) [2, 3, 60]. Compared to the 3569 graphite 1 and 2 SPCEs, the Sigma graphite and carbon black SPCEs exhibit slightly larger C_dl. The larger C_dl of the Sigma graphite SPCE could be due to a higher proportion of edge plane character exposed due to the much smaller particle size. The larger carbon black C_dl is likely due to surface cracks significantly increasing the actual surface area in contact with solution [61].

An increase in C_dl (Fig. 2ai and 2bi) after plasma treatment was observed for all SPCEs, however the extent of the increase varied among the carbon types and plasma treatment times (Fig. 3). While low C_dl is advantageous in electrochemical sensing in order to reduce background currents, a 2 min plasma treatment time was selected since the electrochemical response of these SPCEs was the most reproducible (Fig. S6.3 – S6.5) as well as the most stable over time (data not shown). While C_dl does increase with increasing plasma treatment times, this is expected due to additional removal of binder exposing a larger fraction of electrochemically active carbon particles. C_dl increased by about 20x after 2 min plasma treatment to 230 ± 7, 73.2 ± 3.6, 72.3 ± 13.3 and 195 ± 13 μF cm⁻² for the Sigma graphite, 3569 graphite 1, 3569 graphite 2, and carbon black SPCEs respectively. C_dl of the GC-SPEs also increased by about 20x to 102 ± 13 and 75.6 ± 10.9 μF cm⁻² for the Alfa GC and Sigma GC SPEs respectively. The plasma treated 3569 1 and 2 SPCEs and Sigma GC-SPEs C_dl are closest to, but still greater than what is expected for typical planar carbon electrodes, pointing toward a rough electrode surface. The exact mechanism governing the much larger C_dl of the other SPCEs is unclear. However, it is likely a combination of surface roughness, as well as defect site type and density, and/or electroactive surface groups, both of which could be intrinsic to the carbon type and exposed after plasma cleaning, or introduced during plasma treatment [3, 62].

Fig. 3.

Calculated C_dl of untreated and plasma treated electrodes as a function of plasma treatment time for 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon black (grey), Alfa GC (red) Sigma GC (blue) SPCEs.

Aqueous potential window is defined by the potential limits at which water electrolysis occurs, where oxygen evolution (OER) occurs at positive potential extremes and hydrogen evolution (HER) occurs at negative potential extremes [63]. Since HER and OER can interfere with electroanalysis of certain analytes, it is important to establish a given electrode material’s potential window. The aqueous potential windows for both untreated and plasma treated electrodes were recorded to define operable potential windows for each SPCE, as well as determine any other faradaic background reactions. Potential windows recorded at a pH of 6.5 (0.1 mol L⁻¹ KCl) are shown in Fig. 4a and 4b with the CVs vertically offset for clarity. The potential window was defined as the anodic and cathodic potential limits where a current density greater than 0.5 mA cm⁻² at a scan rate of 100 mV s⁻¹ (CVs not shown) is passed due to water oxidation or reduction as shown in Table 2 [64].

Fig. 4.

Potential windows recorded in aerated 0.1 mol L⁻¹ KCl (pH 6.5) at a scan rate of 200 mV s⁻¹ from 2.0 V to −2.0 V for (a) 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon back (grey) and (b) Alfa GC (red) and Sigma GC (blue) at (i) untreated electrodes and (ii) plasma treated electrodes.

Table 2.

Estimated potential windows in 0.1 mol L⁻¹ KCl (pH 6.5) for untreated and plasma treated SPCEs.

Carbon Type	Untreated Potential Window (V)	Plasma Potential Window (V)
Alfa GC	2.32	2.15
Sigma GC	2.91	2.50
3569 Graphite 2	2.47	1.95
3569 Graphite 1	3.36	2.08
Carbon Black	2.63	2.11
Sigma Graphite	2.45	1.93

A few trends in the potential window widths are evident. First, after plasma treatment, the potential window of all SPCEs decreases. This is indicative of increased exposure of catalytic sites on the electrode surface after etching of the binder. Another trend, particularly evident with the plasma treated SPCEs, is the potential window dependence on particle purity. Impurities such as carbonaceous substances and metallic impurities are typically present in manufactured carbons [65]. Common metallic impurities such as Ni, Cu, Fe, and Mn are catalytic toward HER. The Sigma GC particles have the lowest metal content of any particle at 99.95% purity, and exhibit the widest potential window of the plasma treated SPCEs by about 0.4 V. This is evident in the CVs (Fig. 4aii and 4bii) where HER is hindered at negative potential extremes at the Sigma GC-SPE. The Alfa GC is not analyzed for metallic content, therefore a lower purity and higher concentration of metallic impurities could lead to the more facile HER observed at Alfa GC-SPEs. While the GC-SPEs are more active toward HER than the 3569 graphite with onset potentials of about −1.0 V and −1.6 V respectively, the 3569 graphite appears to be more active toward OER, passing much higher currents for this process, indicating the catalytic sites types vary between SPCE surfaces.

The Sigma graphite SPCE, carbon black SPCE and Sigma GC-SPE show evidence of a reduction peak between −1.3 and −1.8 V and −1.4 and −2.0 V. This peak is attributed to reduction of dissolved oxygen, a common faradaic reaction that occurs at graphitic carbon electrodes [63, 64]. Since oxygen reduction can interfere with the detection of some analytes at negative potentials, and is difficult to purge in field settings, this should also be considered when selecting an appropriate carbon for SPCE fabrication. Another reduction peak is evident in the CVs of the Sigma GC-SPE, 3569 graphite and Sigma graphite SPCEs at about +0.2 V for the untreated SPCEs which shifts to a potential of about +0.5 V after plasma treatment. This is likely attributable to the reduction of surface quinone groups [66, 67]. The shift to a more facile potential after plasma treatment could be due to increased electrocatalytic activity and/or an increase in the quinone coverage. In the oxidative potential region, a redox couple is evident at the Sigma graphite SPCEs at a potential of about +1.0 V and 0.0 V for the untreated and plasma treated electrodes respectively. The reaction that produces this faradaic current is unclear and could be due to impurities in the graphite or further oxidation of the graphite [63].

3.4. Electrochemically reversible redox probes

To further assess the effects of both carbon type and plasma pretreatment on SPCE performance, CV was carried out with 1.0 mmol L⁻¹ solutions of FcTMA⁺, Fe(CN)₆⁴⁻/Fe(CN)₆³⁻ and pAP as shown in Fig. 5ai–iii (graphite and carbon black SPCEs) and Fig. 5bi–iii (GC-SPEs). These redox species were selected because they typically exhibit electrochemical reversibility at sufficiently conductive and clean electrode surfaces.

Fig. 5.

Averaged CVs (N = 3) recorded with (a) 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon black (grey) and (b) Alfa GC (red) and Sigma GC (blue) at untreated (dashed lines) and plasma treated (solid lines) for (i) the oxidation of 1.0 mmol L⁻¹ FcTMA⁺, (ii) the oxidation of 1.0 mmol L⁻¹ ferri/ferrocyanide and (iii) the oxidation of 0.5 mmol L⁻¹ pAP in 0.1 mol L⁻¹ KCl at a scan rate of 100 mV s⁻¹.

FcTMA⁺ was tested first to gauge ohmic resistance since it undergoes an outer sphere electron transfer process exhibiting Nernstian behavior (ΔE_p 59 ± 10 mV) under diffusion limited conditions (Fig. 5ai and 5bi insets) [63]. All untreated electrodes exhibit irreversible ΔE_p for FcTMA⁺, indicating ohmic resistance (Table 3) [2, 59]. The 3569 SPCEs exhibit the smallest ΔE_p of 72 ± 2 mV and 76 ± 4 mV for the high and low mass loading SPCEs respectively. The Sigma graphite SPCE exhibited a less reversible ΔE_p of 94 ± 7 mV. Here, an increase in ohmic resistance is believed to be a direct result of the less dense packing of Sigma graphite compared to the 3569 graphite SPCEs observed in Fig. 1. Carbon black exhibited the largest ΔE_p of 125 ± 8mV, which is consistent with the lowest mass loading of particles as well as the presence of cracks on the electrode surface. The GC-SPE types exhibited slightly larger ΔE_p of 110 ± 6 mV and 92 ± 4 mV for the Alfa GC and Sigma GC-SPEs respectively. While both particle types have a similar manufacturer reported size distribution, SEM imaging revealed the Sigma GC microparticles are characterized by a narrower size distribution (Fig. S2). The narrower size distribution and smaller average particle size of the Sigma GC microparticles likely results in additional particle to particle contacts throughout the bulk composite compared to the Alfa GC.

Table 3.

Peak-to-peak separation for CV of 1.0 mmol L⁻¹ FcMTA⁺, ferri/ferrocyanide, and 0.5 mmol L⁻¹ pAP at untreated and plasma treated SPCEs.

Carbon Particle Type	FcTMA+ ΔEp (mV)	FcTMA+ ΔEp (mV) plasma	Ferri/Ferrocyanide ΔEp (mV)	Ferri/Ferrocyanide ΔEp (mV) plasma	pAP ΔEp (mV)	pAP ΔEp (mV) plasma
Alfa GC	110 (±6)	110 (±3)	532 (±46)	219 (±8)	222 (±9)	116 (±3)
Sigma GC	92 (±4)	80 (±2)	719 (±163)	183 (±3)	350 (±14)	67 (±2)
3569 Graphite 2	72 (±2)	53 (±1)	572 (±96)	105 (±13)	376 (±17)	39 (±2)
3569 Graphite 1	76 (±4)	56 (±2)	753 (±16)	204 (±16)	455 (±8)	48 (±4)
Carbon Black	125 (±8)	102 (±6)	398 (±50)	174 (±3)	415 (±27)	92 (±4)
Sigma Graphite	94 (±7)	69 (±2)	297 (±8)	127 (±4)	222 (±20)	62 (±1)

After plasma treatment, all electrodes except for the Alfa GC-SPE exhibited improved reversibility (Table 3). While FcTMA⁺ voltammetry is not sensitive to surface chemistry or microstructure, the removal of the binder, which acts as a barrier to electron tunneling, slightly improves electron transfer kinetics [3]. Of significance, reversible voltammograms were obtained for both plasma treated 3569 graphite SPCEs with ΔE_p of 53 ± 1 mV and 56 ± 2 mV for the high and low mass loading respectively. This reversibility is typically not observed at SPCEs for outer sphere redox couples, indicating that there is a relationship between both the bulk composition and surface cleanliness and the quality of the electrochemical response at SPCEs [68]. Expected peak currents for FcTMA⁺ were estimated using Equation S6.1 with the geometric surface area of the SPCEs. All untreated SPCEs, excluding carbon black, exhibited greater peak currents than the expected peak current of 22.26 μA, indicating SPCE surfaces are rough. After plasma treatment, the peak currents for all electrodes increased slightly (Table S3), indicating a greater surface roughness [26].

Ferri/ferrocyanide is an inner-sphere redox couple commonly applied to electrode characterizations [3, 40, 69]. The electron transfer kinetics of ferri/ferrocyanide are sensitive to surface chemistry, microstructure, density of electronic states and electrode cleanliness [53, 70]. Therefore, it was used to assess variation in SPCE surface conditions before and after plasma pretreatment (Fig. 5aii and 5bii). Here, the surface sensitive nature of ferri/ferrocyanide is demonstrated by large ΔE_p (Table 3) and low peak currents for all untreated SPCEs. Upon plasma treatment, the electrocatalytic activity of the SPCEs toward ferri/ferrocyanide increases drastically, indicated by decreased ΔE_p and increased peak current, however, the magnitude of improvement depended on carbon particle type. The 3569 graphite 2 SPCE exhibited a decrease of ΔE_p from 572 ± 96 mV to 105 ± 13 mV after plasma treatment. In contrast, the Sigma graphite SPCE exhibited a decrease of ΔE_p from 297 ± 8 mV to 127 ± 4 mV. The untreated Sigma graphite SPCE exhibits a much smaller ΔE_p than the untreated 3569 graphite SPCE. This is possibly a result of the smaller particle sizes resulting in a higher density of catalytic edge planes and defect sites on the untreated electrode surface which would affect the capacitance discussed previously as well. However, upon plasma treatment, the 3569 graphite 2 SPCE exhibits a smaller ΔE_p by about 20 mV. These results are consistent with FcTMA⁺ voltammetry, suggesting Sigma graphite SPCEs are more resistive than high and low mass loading 3569 graphite SPCEs.

While the electrocatalytic activity of GC-SPE types toward ferri/ferrocyanide improves with plasma treatment, the kinetics were sluggish compared to the graphite SPCEs. ΔE_p of 219 ± 8 m and 183 ± 3 mV were obtained for the plasma treated Alfa and Sigma GC-SPEs respectively. This is likely influenced by the microstructure of GC microparticles compared to graphite particles, where graphite contains a high proportion of edge plane like defects to support fast electron transfer of ferri/ferrocyanide [70]. The theoretical ΔE_p of ferri/ferrocyanide is 59 ± 10 mV, indicating electron transfer is still slightly impeded for this couple at all plasma treated SPCEs. Surface oxygen functionalities are introduced during plasma treatment and surface oxides are believed to impede electron transfer of ferri/ferrocyanide at carbon electrodes, possibly by blocking of adsorption sites or electrostatic repulsion between negatively charged oxides and negatively charged ferri/ferrocyanide molecules [69, 70]. Nearly reversible voltammetry (ΔE_p ~ 70 mV) was obtained for ferri/ferrocyanide at argon plasma treated SPCEs [27], while irreversible voltammetry (ΔE_p ~ 156 mV) [26] was obtained for oxygen plasma treated SPCEs and quasireversible voltammetry (ΔE_p ~ 86 mV) was obtained at alumina polished GC electrodes [71], consistent with the data provided here.

Sensor development for the electrochemical detection of p-aminophenol (pAP) is important because pAP serves as the electroactive product in immunoassays and is used in several industries leading to environmental pollution [72, 73]. Typically, sluggish electron transfer kinetics resulting in low sensitivity are obtained at conventional and pretreated SPCEs [74]. Therefore, voltammetry of pAP at the SPCEs developed in this work was assessed. CVs for pAP at untreated and plasma treated SPCEs are shown in Fig. 5aiii and 5biii. Again, significant improvement in the electrocatalytic activity of plasma treated SPCEs is observed. pAP undergoes a two electron, two proton oxidation with a theoretical peak separation of 29.5 mV (59/n mV). A wide range of ΔE_p values have been reported for pAP at carbon electrodes with most clean/activated electrodes exhibiting quasireversible voltammetry. ΔE_p of about 60 mV for TPEs [5], 140 mV for bare GC macro electrodes [75], and 64 mV for electrochemically activated SPCEs [74] have been reported. Here, as expected, the untreated SPCEs exhibit irreversible voltammetry (Table 3). However, remarkably, the plasma treated 3569 graphite 2 SPCEs exhibit nearly reversible voltammetry with ΔE_p of 39 ± 2 mV. 3569 graphite TPEs, which also require an activation step to remove excess polymer from coated particles, did not exhibit this level of reversibility toward pAP [5]. These results indicate that optimizing SPCE composition can drastically improve electrochemical performance toward relevant analytes to be on par with electrodes typically considered to be higher quality [2, 5]. At the GC-SPEs, pAP voltammetry remained consistent with the trend observed for FcTMA⁺ and ferri/ferrocyanide, where the Sigma GC-SPEs provided better reversibility than the Alfa GC-SPEs (Table 3).

To further understand any ohmic drop effects due to conductivity difference of different SPCE formulations CV of FcTMA⁺ at varying scan rates was carried out. Fig. 6 shows ΔE_p as a function of scan rate for all SPCEs. Here, ΔE_p increases to varying degrees for each SPCE with increasing scan rate. The low and high mass loading 3569 graphite SPCEs exhibited the smallest increase in ΔE_p, while the Sigma graphite SPCE exhibited a slightly larger ΔE_p increase, consistent with the higher resistance of this SPCE due to lower mass loading of particles and particle orientation. Sigma GC-SPE ΔE_p exhibits a smaller scan rate dependence than the Alfa GC-SPE, further supporting the hypothesis of increased particle-to-particle contacts. The carbon black SPCE ΔE_p exhibited the greatest scan rate dependence, consistent with the low mass loading of carbon black particles and the cracking of the SPCE surface.

Fig. 6.

Peak-to-peak separation for 1.0 mmol L⁻¹ FcTMA⁺ recorded with CV as a function of scan rates in 0.1 mmol L⁻¹ KCl with 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon black (grey), Alfa GC (red) and Sigma GC (blue) SPCEs (N = 3).

3.5. Surface sensitive and irreversible redox probes

To further elucidate the effects of carbon particle type and plasma treatment on the electroanalytical utility of SPCEs, several biologically relevant probes were studied. Fig. 7a and 7b show the averaged CVs recorded for uric acid (UA, Fig. 7ai and 7bi) and nicotinamide adenine dinucleotide (NADH, Fig. 7aii and 7bii) at both untreated and plasma treated SPCEs. Peak current and peak oxidation potential are listed in Table S4 and S5. Plasma treatment was found to dramatically improve the response at SPCEs, with increased peak currents and lowered detection potentials in both cases. Since plasma treatment is carried out with ambient air, it is expected to etch binder as well as increase surface oxide coverage on the SPCE surfaces [2, 26, 69]. UA is known to be highly sensitive to surface oxides, therefore, the increased catalytic activity of plasma SPCEs toward this probe is attributed to additional surface oxides. The plasma treated Sigma GC-SPE provided the highest catalytic activity of the two GC-SPE types, consistent with previous results. The plasma treated 3569 graphite 2 SPCE exhibited the highest catalytic activity toward UA compared to the other graphite and carbon black SPCEs. Again, the carbon black SPCEs exhibited the poorest voltammetry at both the untreated and plasma treated versions.

Fig. 7.

Averaged CVs (N = 3) recorded with (a) 3569 graphite 2 (green), 3569 graphite 1 (black), Sigma graphite (orange), carbon black (grey) and (b) Alfa GC (red) and Sigma GC (blue) at untreated (dashed lines) and plasma treated (solid lines) for (i) the oxidation of 1mM UA and (ii) the oxidation of 1 mmol L⁻¹ NADH in 0.1 mol L⁻¹ phosphate buffer (pH 7.4) at a scan rate of 100 mV s⁻¹.

The electrooxidation of NADH is of interest since has wide application in dehydrogenase based biosensors [76–78]. NADH oxidation proceeds at relatively high potentials at unmodified carbon electrodes (greater than +0.5 V), reducing specificity [78, 79]. Typically, SPCEs are modified with nanomaterials, such as carbon nanotubes, to increase edge plane coverage since untreated SPCEs are relatively inactive [80, 81]. Here, similar to other carbon electrodes, all untreated SPCEs exhibit a peak potential greater than +0.5 V for NADH oxidation, however, the catalytic activity of the SPCEs significantly improves after plasma treatment. Upon plasma treatment, the 3569 graphite 2 SPCE exhibits the lowest peak potential of 350 ± 15 mV, a decrease of about 175 mV. The 3569 graphite 1 SPCE demonstrated lower activity than its higher mass loading counterpart, likely due to a combination of increased resistance and fewer catalytic sites at the electrode surface. The Sigma graphite SPCE again exhibited a slightly higher peak potential and lower peak current than the high mass loading 3569 SPCE, however, this is probably related to the lower mass loading and particle orientation resulting in additional resistance as discussed previously. The GC-SPEs did not perform as well as their graphite SPCE counterparts for this probe. This is consistent with literature where peak potentials of ~ 400 mV (vs. SCE) and ~ 600 mV (vs. SCE) were obtained at an edge plane pyrolytic graphite (EPPG) and GC electrode respectively [80]. The peak current for the plasma treated 3569 graphite 2 SPCEs was closest to the theoretically predicted peak current of 22.6 μA estimated using Equation S6.2.

3.6. Surface adsorption dependent species and electrode fouling

The electrochemical detection of dopamine (DA) is of substantial interest. DA is a neurotransmitter with excesses and deficiencies linked to neurological illnesses including Parkinson’s, autism and schizophrenia [82, 83]. The oxidation of DA is a surface adsorption dependent two-electron, two-proton process after which several side reactions can occur generating dopaminergic products that can adsorb to and block the electrode surface [84]. It is widely accepted that both the kinetics of DA oxidation and the susceptibility to surface passivation is highly dependent on carbon electrode surface microstructure [84]. Therefore, the oxidation and fouling characteristics of DA were studied on untreated and plasma treated SPCEs by recording successive CVs for a total of seven cycles. CV cycle #1 and #7 are shown in Fig. 8a for (i) 3569 graphite 2 SPCEs and (ii) Sigma GC-SPEs. Fig. 8b compares the change in peak current for successive voltammetric cycling for (i) 3569 graphite 2 SPCEs and (ii) Sigma GC-SPEs. The peak current data is shown as the ratio of the n^th cycle to the first cycle (i_n/i_initial). Consistent with previous results, plasma treatment significantly enhanced electron transfer kinetics for this surface sensitive probe indicated by the decreased peak separation and increased peak currents for both SPCEs.

Fig. 8.

Averaged (a) CVs (N = 3) for cycle #1 and cycle #7 of 1.0 mmol L⁻¹ DA in 0.1 mol L⁻¹ phosphate buffer (pH 7.4) at (i) 3569 graphite 2 SPCE and (ii) Sigma GC-SPE at untreated (black) and plasma treated (red) SPCEs. Plot (b) of the ratio of a given cycle # peak current (i_n) to initial peak current (i_initial) for (i) 3569 graphite 2 SPCE and (ii) Sigma GC-SPE at untreated (black) and plasma treated (red) SPCEs.

Initially, the 3569 graphite 2 SPCEs exhibit greater reversibility and sensitivity than Sigma GC-SPEs with ΔE_p 65 ± 6 mV and 90 ± 4 mV respectively. Commercial graphite SPCEs exhibited significantly more irreversible voltammetry of DA with ΔE_p of 232 mV after undergoing oxygen plasma treatment, providing further evidence of the enhanced quality of SPCEs fabricated in this work [26]. However, after seven cycles, ΔE_p at the 3569 graphite 2 SPCE decreases by 28% to 91 ± 6 mV while the Sigma GC-SPE decreases by only 14% to 103 ± 8 mV. This trend is also apparent in Fig. 8bi and 8bii, where the peak current decreases less upon voltammetric cycling of DA with the Sigma GC-SPE compared to the 3569 graphite 2 SPCE. The peak current decreases, stabilizing after 3 cycles to about 72% of the initial peak current for the GC-SPE while peak current stabilization occurs after cycle four at the 3569 graphite 2 SPCE to about 66% of the initial peak current. These results are consistent with literature reports of GC electrodes exhibiting less susceptibility to fouling and lowered sensitivity than both EPPG and BPPG electrodes at these DA concentrations [84]. Fouling has been found to be more severe at BPPG than both EPPG and GC, suggesting edge planes and defect sites are more resistant to fouling than basal planes. However, the graphite used here is highly heterogeneous consisting of both edge and basal planes and cannot be directly compared to EPPG or BPPG; therefore, the exact mechanism for differences in fouling severity at 3569 graphite 2 SPCEs and GC-SPEs is unclear at this time [84]. However, these results indicate that fouling severity at SPCEs can be tuned by careful selection of carbon particle. Surface oxide coverage also plays a role in stability toward DA cycling, with increased surface oxygen functionalities thought to enhance stability [85]. Since plasma treatment increases surface oxide coverage, this is likely partially responsible for the increased stability toward DA cycling of the plasma treated SPCEs.

4. Conclusions

The work presented in this chapter demonstrates that there is a significant relationship between bulk SPCE composition and electrochemical and physical properties of the final electrode. The carbon particle types used in this work result in SPCEs with widely different electron transfer properties toward a range of redox probes, with the responses related to binder content, surface oxides, carbon particle microstructure, particle purity, and particle size. Of significance, it was found among graphite types that larger carbon particles can achieve higher mass loadings in the binder, resulting in more particle to particle contacts and less resistive effects typically associated with SPCEs. The electrochemical properties and electron transfer rates of the 3569 graphite SPCEs were comparable to other CEs typically considered to be higher quality than SPCEs (e.g. TPEs). Carbon nanomaterials, such as carbon black, are limited to low carbon particle to binder ratios resulting in more resistive and mechanically unstable SPCEs. Therefore, it is recommended that carbon nanomaterials, such as carbon black, carbon nanotubes and graphene are better suited for use as surface modifications of SPCEs as opposed to serving as the main conductive component of SPCE formulations. In terms of GC microparticle SPEs, SEM imaging revealed that a narrower size distribution of particles results in a less resistive and higher quality GC-SPE. However, due to morphological differences, GC-SPEs are still more resistive than their graphite counterparts and are better suited to applications in which a surface with homogeneous electrochemical activity is beneficial as demonstrated in our previous work for anodic stripping voltammetry of heavy metals. Several redox probes were studied on all SPCE surfaces providing insights into the electrochemical reactivity of SPCEs fabricated with different carbon particles. Results indicate that SPCE properties can be tuned and improved with the careful selection of carbon particle type. Plasma treatment, which is fast and simple, was found to significantly enhance electrochemical characteristics of SPCEs through removal of surface binder and impurities, exposing the underlying, electrochemically active carbon particles. Increased double layer capacitance and decreased solvent window after plasma treatment can be minimized by employing carbon particles with a known high purity. Here, we present a systematic guide to SPCE composition that can be taken into account when developing new, high performance electrochemical PON sensors with printable carbon inks. Significantly, the improvement in electrochemical responses at SPCEs was achieved without complex surface modifications, which has broad implications for improving electrochemical PON sensors. Furthermore, these inks can be mass produced and printed into several geometries onto a variety of substrates including paper, PET films, glass, and plastics providing adaptability to a wide array of PON sensing applications.

Highlights.

Screen- and stencil-printed electrodes are widely used in electroanalytical chemistry but are typically based on commercial inks. In this work we sought to answer the question of how different carbon additives affected electrochemical performance of these electrodes with a goal of generating improved performance. We found that by controlling the amount and type of carbon additive as well as controlling surface pretreatment methods, significant improvements in performance could be achieved using the modified inks.