Platinum Nanoparticle Size and Density Impacts Purine Electrochemistry with Fast-Scan Cyclic Voltammetry

Abstract

We demonstrate the density and shape of platinum nanoparticles (PtNP) on carbon-fiber microelectrodes with fast-scan cyclic voltammetry (FSCV) directly impacts detection of adenosine. Previously, we showed that metal nanoparticle-modified carbon significantly improves adenine-based purine detection; however, how the size and shape of the particles impact electrochemical detection was not investigated. Electrochemical investigations of how the surface topology and morphology impacts detection is necessary for designing ultrasensitive electrodes and for expanding fundamental knowledge of electrode-analyte interactions. To change the density and shape of the PtNP’s on the surface, we varied the concentration of K₂PtCl₆ and electrodeposition time. We show that increasing the concentration of K₂PtCl₆ increases the density of PtNP’s while increasing the electrodeposition time impacts both the density and size. These changes manipulate the adsorption behavior which impacts sensitivity. Based on these results, an optimal electrodeposition procedure was determined to be 1.0 mg/mL of K₂PtCl₆ deposited for 45 s and this results in an average increase in adenosine detection by 3.5 ±0.3-fold. Interestingly, increasing the size and density of PtNPs negatively impacts dopamine detection. Overall, this work provides fundamental insights into the differences between adenosine and dopamine interaction at electrode surfaces.

Introduction

Adenine-based purines, including adenosine and adenosine triphosphate (ATP), are important biological molecules involved in a wide range of signaling roles in the central and peripheral nervous system.^1,2 Adenosine plays a role in a wide range of functions, including sleep regulation³, inflammatory immune responses⁴, blood-brain barrier permeability⁵, and energy metabolism⁶. A key to measuring purinergic signaling in real-time is the measurement techniques must be on the necessary time scale to capture rapid fluctuations because the release and ultimate metabolic degradation of ATP to adenosine extracellularly happens on the millisecond times scale.⁷ Fast-scan cyclic voltammetry (FSCV) is an electrochemical technique that has been used for over the last decade to study rapid adenosine signaling in rat brains^8,9 as well as neurological studies in patients with brain tremors.¹⁰ Traditionally, FSCV is coupled to carbon fiber microelectrodes¹¹ and this has been widely used in both in vivo^12–14 and ex vivo preparations.^15–17 Despite its successful use for adenosine detection, bare carbon fiber surfaces remain significantly less sensitive to purines compared to more common neurotransmitter analytes detected with FSCV; most notably, the catecholamines. Previous work has shown that surface modifications on carbon-fiber microelectrodes can enhance adenosine and ATP detection.^18,19 Specifically, our lab has demonstrated that modifying the surface of carbon with metal nanoparticles, including gold and platinum, significantly improves purinergic detection.²⁰ Our prior work focused on improving ATP detection; however, we demonstrated that this approach could be beneficial for adenosine detection with FSCV. Despite this finding, our prior report focused on the analytical method of metal deposition on the carbon-fiber surface and its impact on detection; however, a fundamental investigation of how subtle manipulations in the nanoparticle (NP) density and size on the surface impacts detection was not explored. Improving our understanding of how specific analytes interact at changes in electrode surface structure and topology is critical for advancing our knowledge of the analyte-èlectrode interface and for improving sensitivity for ex vivo and in vivo detection with FSCV. In this paper, we investigate how changes in the platinum nanoparticle (PtNP) density on the surface and morphology of the NP’s impacts adenosine oxidation and adsorption at the electrode surface.

Using platinum nanoparticles as electrochemical sensors are of increased interest due to their conductivity and electrocatalytic properties.^21–23 Previous work has shown that aptamers modified with PtNPs acted as better redox mediators for electrochemical detection of adenosine.²⁴ Modifying carbon-based electrode surfaces with PtNPs changes the catalytic properties for a wide range of targets, thereby increasing electrochemical signalling.^22–25 Previously, it was shown that PtAu alloy nanoparticle morphology, composition, and size can be controlled by controlling the electrodeposition conditions, thereby increasing the electrocatalytic activity of the target.²² Nanoparticle size has been found to control oxygen reduction, causing a significant loss in catalytic activity with decreased PtNP size.^27,29,30 Controlling the nanoparticle dimensions is necessary to understand the physical and chemical properties of the PtNP and their impacts on electrochemical detection. Here, we demonstrate that manipulating the platinum particle size and morphology directly impacts electrochemical detection of adenosine and this impact is analyte dependent.

In this paper, we controlled the morphology and density of PtNP’s on the surface of carbon-fiber microelectrodes by manipulating the electrodeposition parameters to reveal the extent to which these properties differentially influence adenosine vs. dopamine detection. Although adenosine was our analyte of interest, a comparative analysis to dopamine demonstrates further that PtNP size/density on the surface and their impact on electrochemical detection is analyte specific. Overall, we show that increasing the size and density of PtNP on the surface positively impacts adenosine adsorption and electrocatalytic behavior; whereas, these changes negatively impact dopamine detection. Further analysis using surface-enhanced Raman scattering (SERS) revealed that adenosine has a higher affinity for Pt compared to dopamine which provides additional confirmation of the electrochemistry results.

Experimental

Chemical reagents.

Chemicals to make Tris buffer (15 mM Tris, 1.25 mM NaH₂PO₄, 2.0 mM Na₂SO₄, 3.25 mM KCl, 140 mM NaCl, 1.2 mM CaCl₂ dehydrate, and 1.2 mM MgCl₂) were purchased from Fisher Scientific (USA). Stock buffer solutions were prepared at pH 7.4 to use in all flow injection experiments. Stock solutions of dopamine (DA) and adenosine (AD) (Sigma-Aldrich, St. Louis, MO, USA) for electrochemistry experiments were prepared at 10 mM concentration, dissolved in 0.1 M HCl, and then stored at 4°C. Daily solutions were prepared by diluting the stock solution in Tris buffer. Potassium hexachloroplatinate (K₂PtCl₆) was purchased from Sigma-Aldrich. Stock K₂PtCl₆ solutions of 0.1 mg/ml, 0.5 mg/ml, 1.0 mg/ml and 2.0 mg/ml were prepared by dissolving in deionized water (Milli-Q, Millipore, Billerica MA). Solutions of dopamine (DA) for SERS experiments were prepared first at 300 mM in 0.1 M HCl and diluted as needed. Solutions of adenosine (AD) (Sigma-Aldrich, St. Louis, MO, USA) for SERS experiments were prepared at 100 mM, dissolved in 0.1 M HCl, and diluted as needed. Solutions of 1,2-bis(4-pyridyl)ethylene (BPE) (Sigma-Aldrich, St. Louis, MO, USA) for SERS experiments were prepared at 50 mM in 50% ethanol (v/v) (Decon Laboratories, Inc., King of Prussia, PA, USA), and diluted with water as needed. The piranha solution for cleaning the glass substrate used for SERS was made by making a 1:3 mixture of 30% hydrogen peroxide and concentrated sulfuric acid (Sigma-Aldrich, St. Louis, MO, USA). A 10% APTES solution was made by diluting (3-Aminopropyl)triethoxysilane (Sigma-Aldrich, St. Louis, MO, USA) in ethanol. 50 nm PtNPs for the substrates were purchased from nanoComposix (San Diego, CA, USA).

Fabrication of electrodes and surface modification.

Cylindrical carbon-fiber microelectrodes (CFME) were made from 7-μm in diameter T650 carbon-fibers (Gift from Mitsubishi Chemical Carbon Fiber and Composites Inc., Sacramento, CA, USA). Carbon-fibers were vacuum aspirated into a capillary glass tube (1.2 × 0.68 mm, A-M Systems, Sequim, WA) and pulled into two using a vertical micropipette puller (Narishige PE-22, Tokyo, Japan). Electrodes were cut 50 – 100 μm from the glass seal using a microscope (Fisher Education, USA). Electrodeposition of platinum nanoparticles was done by placing electrodes in K₂PtCl₆ solution, while scanning from −1.2 to +1.5 V at 5 V/s against an Ag/AgCl reference electrode. The concentration of deposition and the time of deposition was varied to study the impact of PtNP density and size on adenosine interaction.

PtNP substrate fabrication for SERS.

Glass microscope slides were cut to a size of 0.5×2.5 cm and cleaned using piranha solution. After 1 hr, the slides were removed from the piranha solution, washed in deionized water and stored in ethanol until use. The cleaned slides were immersed in a 10% APTES solution for 1 hour at 70° C, to generate an amine-functionalized glass surface which immobilize the nanoparticles. The slides were washed in ethanol, water and ethanol again before letting them air dry. Once dry, 20 μL of PtNPs were deposited on the APTES-treated surface and airdried overnight.

Fast-scan cyclic voltammetry (FSCV).

Fast-scan cyclic voltammograms were collected using a WaveNeuro potentiostat with a 1 MΩ headstage (Pine Instruments, Durham, NC, USA). Data collection was done using a National Instruments PCIe-6363 interface board (Austin, TX) and HDCV software (UNC-Chapel Hill, Mark Wightman). Electrodes were allowed to equilibrate for 10 minutes. A triangular waveform was applied from −0.4 V to +1.45 V, at 400 V/s and a frequency of 10 Hz. Analyte was injected using a home-built flow injection analysis system at the electrode for 5 s (Valco Instruments, Houston, TX, USA). Buffer was delivered at a rate of 1 mL/min with a syringe pump (Chemyx, Stafford, TX, USA) prior to and after analyte delivery. Nonfaradaic background current was removed just before analyte injection, to reveal background-subtracted cyclic voltammograms (CVs).

Scanning electron microscopy imaging.

Surface characterization of the density and size of PtNPs on CFME’s was done using scanning electron micrographs (SEM) using a FEI XL30 SEM coupled to an EDAX detector for energy dispersive spectroscopy (EDS), at a voltage of 10 kV. Images were taken at different magnifications to investigate surface topology. Image J Fiji was used in tangent with SEM to approximate the increased nanoparticle size after electrodeposition.

Raman measurements.

A Renishaw InVia confocal Raman microscope was used to collect the Raman spectra using a 633-nm He-Ne laser at 10% power and WiRE 3.4 as the data collection software (Gloucestershire, UK). The solutions used in these studies were water (control), 10-, 50-, and 100-mM adenosine, 15-, 25-, 50-, 75-, 100-, 150-, 175-, 200-, and 300-mM dopamine, and 0.1-, 1-, 10-, 100-, 500- μM, and 1-, 2-, 5-, and 10-mM BPE. Measurements were taken after depositing 20 μL of solution onto a substrate and waiting 25 minutes for adenosine and dopamine, and 10 minutes for BPE. Each sample was measured in three different spots with an integration time of 25 seconds for measurements on substrates and 20 seconds for capillaries. The measurements were taken focusing near the edge of the deposited PtNPs, where the signal was highest. As a control the solution were also dried on an aluminum slide, to detect the spontaneous Raman signal (no SERS). To show that there was a signal enhancement on the PtNP substrates, the signal of BPE was measured on a glass slide and compared with the PtNP substrates at the same concentration (2 mM). The spectra background for each measurement was subtracted using a lab-built MATLAB program that uses Savitzky–Golay frequency filter and a polynomial subtraction for the baseline.

Statistical analysis.

All statistical analysis was done in GraphPad Prism V. 9.0 (GraphPad Software Inc., La Jolla, CA, USA). Post modification peak current over pre modification peak current was calculated to determine the change in current after modification, with a ratio of 1.0 indicating no change. Statistical p values < 0.5 were considered significant (95% confidence level). The n represents number of electrodes tested.

Results and Discussion

Characterization of nanoparticle size and density on electrode surfaces

The concentration of the platinum solution and the time of deposition impacts the density and size of PtNP’s on the surface of carbon-fiber microelectrodes. Scanning electron microscopy (SEM) was used to characterize the extent to which deposition concentration and time impacted the size and density of PtNPs on carbon fiber microelectrodes. The concentrations tested ranged from 0.1 mg/mL (Figure 1D), 0.5 mg/mL (Figure 1C), 1.0 mg/mL (Figure 1B) to 2.0 mg/mL K₂PtCl₆ (Figure 1A). The deposition time was held constant at 30-s. At the lowest concentration (0.1 mg/mL), minimal PtNPs were observed on the surface (Figure 1D). Increasing the concentration of K₂PtCl₆ increases the density of PtNPs on the surface; however, particles are relatively uniform in size. At 2.0 mg/mL (Figure 1A), the electrode is almost completely covered with PtNPs. Energy-dispersive X-ray (EDX) spectroscopy was used to approximate the percentage of Pt on the surface. A trend of increasing % Pt on the surface was observed as a function of deposition concentration (Figure S1 A–D). The PtNP size was approximated using Image J and similar findings were observed to previous work.²⁰ On average, we observe that using a 0.1 mg/mL K₂PtCl₆ solution yielded an average NP size of 18.9 ± 0.85 nm, 22.25 ± 1.00 nm for 0.5 mg/mL, 25.51 ± 0.60 for 1.0 mg/mL, and 73.01 ± 2.5 nm for 2.0 mg/mL. Substantial increases in NP size was only approximated at the 2.0 mg/mL deposition solution; however, accurate particle sizing is difficult at such dense coverage on the electrode surface due to overlays of particles on the surface which could skew the estimated particle size.

Figure 1. The density and size of PtNP clusters change as a function of deposition parameters.

For all conditions, a higher magnification image (scale bar shown on image) is presented to the right. Changing the concentration from 2.0 mg/mL (A), 1.0 mg/mL (B), 0.5 mg/mL (C), and 0.1 mg/mL (D) with a 30-s deposition time resulted in an increase in the density of PtNPs on the surface. Changing the deposition time, while keeping the concentration unchanged at 1.0 mg/mL, from 60 s (E), 45 s (F), 30 s (G) and 15 s (H) resulted in changes in the density and size of the PtNP clusters on the surface.

The deposition time was varied while keeping the concentration of K₂PtCl₆ constant (at 1.0 mg/mL, Figure 1 E–H). The deposition times tested ranged from 15 s (Figure 1H), 30 s (Figure 1G), 45 s (Figure 1F), to 60 s (Figure 1E). Increasing the deposition time not only increases the density of Pt on the surface (Figure 1 and S1), but at longer deposition times, the morphology and size of the PtNPs on the surface changes. Subtle increases in PtNP size are evident when comparing a 15 s deposition to a 30 s deposition (Figure 1H and G, Figure S-2); however, changes in the actual morphology are evident at deposition times greater than 30 s. Clusters of PtNPs begin to form on the surface at 45 and 60 s deposition times (Figure 1F and E). Varying the deposition time yielded a mean NP size of 21.35 ± 1.4 nm for 15 s, 25.51 ± 0.60 nm for 30 s, 112.4 ± 5.6 nm for 45 s, and 156.3 ± 5.1 nm for 60 s. The treatments that formed clusters, 45 s and 60 s, were statistically significant in mean NP size above all other treatments (Figure S-2B, one-way ANOVA with Bonferroni, p < 0.0001, n = 65–235 particles). In addition to NP size, significant differences were observed with EDX analysis as a function of deposition time (Figure S1 E–H). The following section discusses how these changes in PtNP density and size impact electrochemical detection of adenosine and dopamine with fast-scan cyclic voltammetry.

Dopamine (DA) interaction at carbon-fiber microelectrodes has been well-characterized.^13,31–33 Carbon-based surfaces are known to be excellent electrode materials for DA detection. Although detection of adenine-based purines on carbon-based surfaces is possible, significantly less oxidative current for adenine-based purines is observed at carbon-fibers compared to DA (Figure 2). Adenosine (AD) is an important adenine-based purine often detected with FSCV in the brain.^8,9 Improvement in our understanding of adenosine’s interaction at electrode surfaces is necessary for rendering electrode surfaces with higher affinities for AD. AD undergoes three 2-electron oxidation reactions (Figure S-3); however, typically only the first 2 oxidation products are detected at carbon-fiber electrodes. In Figure 2, we demonstrate a representative color plot and cyclic voltammogram (CV) for 5 μM AD and 1 μM DA. The primary oxidation peak for AD appears at 1.4 V on the cathodic scan with a secondary oxidation peak at 1.2 V. Despite testing 5-times the concentration of AD, higher oxidative current is always observed for DA. We previously demonstrated that PtNP-modified carbon-fiber microelectrodes are beneficial for adenine-based purines²⁰; however, a comprehensive analysis of how their size, shape, and coverage impact electrochemical detection for adenine-based purines was not shown.

Figure 2. Significantly less oxidative current observed for adenosine compared to dopamine at unmodified carbon-fiber microelectrodes.

(A) False-color plot with time on the x-axis, voltage on the y-axis, and current on the z-axis for 5 μM AD. (B) Representative cyclic voltammograms (CV) for AD. (C) False-color plot for 1 μM DA and (D) Representative CV for DA.

Density and size of PtNPs on the surface influence electrochemical detection

The density and changes in PtNP morphology directly impacted electrochemical detection of both AD and DA with FSCV. The concentration of K₂PtCl₆ during deposition directly impacted the measured change in oxidative peak current for AD and DA (Figure 3A and B). Oxidative peak current for AD and DA were analyzed prior to and after deposition and the ratio (Ipost/Ipre) was plotted as a function of deposition variable (Figure 3). The change in oxidative peak current for AD using 1.0 mg/mL was significantly different than 0.1 mg/mL and 0.5 mg/mL (Figure 3A, one-way ANOVA with Bonferroni post-tests, p < 0.05, n = 8). The largest change was observed at 2.0 mg/mL K₂PtCl₆, which was significantly different than 0.1 mg/mL and 0.5 mg/mL (Figure 3A, one-way ANOVA with Bonferroni post-tests, p < 0.001, n = 8); however, no significant differences were observed between 2.0 mg/mL and 1.0 mg/mL (p > 0.05). There was no statistical difference between concentration treatments on peak dopamine oxidative current (Figure 3B, one-way ANOVA with Bonferroni post-tests, p > 0.05, n = 7–8). The nonfaradaic background current was analyzed before and after modification for each deposition procedure (Figure S-4). No statistically significant differences in background current were observed for all deposition procedures (Figure S-4, one-way ANOVA, p > 0.05, n = 5); however, we did consistently observe larger changes in background current at higher deposition concentrations and deposition times. While the average largest change in oxidative current for AD was observed with 2.0 mg/mL, the optimized concentration was determined to be 1.0 mg/mL. Electrodes modified with 2.0 mg/mL K₂PtCl₆ often resulted in larger background currents, although not significantly different than 1.0 mg/mL, (Figure S-4). leading to overloading and irreproducibility between batches of electrodes. This resulted in a larger SEM associated with this treatment (Figure S-4). Overall, these results indicate as we increase the density of Pt on the surface by manipulating the deposition concentration, we improve AD interaction at the surface.

Figure 3. Varying platinum nanoparticle electrodeposition on carbon-fiber microelectrodes impacts detection of adenosine and dopamine.

The ratio of oxidative peak current post-deposition over pre-deposition was plotted as a function of deposition variable. A ratio of 1.0 indicates no change, whereas a ratio above 1.0 indicates an improvement in oxidative current. (A) Increasing the PtNP concentration resulted in increases in oxidative peak current for AD. (B) Conversely, increasing the concentration of PtNP deposition resulted in decreases in DA oxidative peak current. (C) 1.0 mg/ml K₂PtCl₆ was deposited at varying deposition times. Increasing the deposition time resulted in increases in peak oxidative AD current. (D) Conversely, increasing the deposition time resulted in decreases in DA oxidative current. (n = 6–8)

Increasing the deposition time results in changes in the surface coverage and NP morphology which impacts electrochemical detection of AD and DA (Figure 3C and D). Deposition times were varied from 15 s to 60 s, in 15-s increments. No significant differences in oxidation current were observed when comparing deposition treatment times for AD or DA; however, a trend in increasing oxidation current for AD as a function of increasing deposition time was observed empirically (Figure 3C and D, one-way ANOVA, p > 0.05, n = 6–8.) By comparing statistically the post to pretested oxidative current within each treatment time, we observed a statistical difference above a theoretical mean of 1.0 for each time treatment (One Sample t-test, p > 0.05, n = 6–8.) The statistical significance for each concentration and time treatment is shown in Table 1. The longest deposition time did not statistical reduce DA current (One Sample t test, p = 0.0866, n = 6.); however, we observed empirically that longer deposition times negatively impacted DA detection. Overall, these results indicate that increasing the NP size on the surface does alter the interaction of AD and DA at the surface.

Table 1. PtNP modification procedure influences analyte detection (n = 6–8).

Table 1 shows average ratio of post-modification current to pre-modification current. Values above 1.0 indicate an increase in oxidative current.

	Concentration of K2PtCl6a
	0.1 mg/mL	0.5 mg/mL	1.0 mg/mL	2.0 mg/mL
Adenosine	1.34 ± 0.20	1.42 ± 0.12**	2.83 ± 0.49**	4.06 ± 0.52 ***
Dopamine	1.13 ± 0.17	0.81 ± 0.23	1.26 ± 0.14	0.61 ± 0.16
	Deposition Timeb
	15 s	30 s	45 s	60 s
Adenosine	2.39 ± 0.38 *	2.83 ± 0.49**	3.51 ± 0.32 **	3.39 ± 0.51 **
Dopamine	1.23 ± 0.10	0.98 ± 0.15	1.05 ± 0.07	0.83 ± 0.08*

^aDeposition time was held constant at 30 s

^bDeposition concentration was held constant at 1.0 mg/mL

^{*, **, ***}One Sample t-test comparing to a theoretical mean of 1.0 (* p < 0.05, ** p < 0.01, *** p < 0.001)

The experimental electrodeposition conditions tested above were used to inform the optimized procedure for detecting AD at PtNP-modified CFME’s. From the data, electrodepositing 1 mg/mL K₂PtCl₆ for 45 s yielded large increases in AD oxidative current and was reproducible and stable (Table 1). Longer deposition times and larger concentrations were not chosen as optimal due to instability from larger background currents and reproducibility issues at these conditions. On average, a 3.5 ± 0.3-fold increase in AD oxidative current is observed at the optimized procedure. Conversely, this procedure did not improve DA detection (1.05 ± 0.07-fold change). Representative example CVs for AD and DA before and after modification at the optimized procedure demonstrate increases in oxidative current for AD, with minimal changes in DA (Figure 4). The sensitivity was determined at the optimized deposition procedure by testing the extent to which oxidative current changes as a function of concentration (Figure 4C). Concentrations tested ranged from 100 nM to 10 μM (n = 10–12). The sensitivity improved 3.0-fold for AD after depositing PtNPs on the surface; however, for DA, the sensitivity did not change (slope changed from 41.67 ± 1.66 nA/μM to 44.61 nA/μM ± 3.67 when modified). In addition to changes in sensitivity, we observed improvements in the catalytic conversion of adenosine oxidation products at the surface of PtNP-modified electrodes. This is evident by increases in the secondary and even tertiary oxidation peaks. We previously observed this behavior for ATP at PtNP-modified electrodes; however, an extensive analysis of how the density and morphology of the NPs impact this behavior was not done.²⁰ The following section describes how the PtNP size and shape impacts the catalytic behavior at the surface.

Figure 4. Optimized PtNP deposition procedure significantly increases sensitivity to AD, with minimal changes to DA.

The deposition procedure that resulted in the highest increase in oxidative current for AD and stability in background current was: 1 mg/mL K₂PtCl₆ and a 45 s deposition time. (A) CV for 5 μM AD and (B) for 1 μM DA before and after PtNP deposition. (C) Current increases linearly with concentration ranging from 300 nM to 3 μM. The slope (sensitivity) increased significantly for AD (from 3.31 nA/μM to 10.06 nA/μM) but not for DA (from 41.67 nA/μM to 44.61 nA/μM) at the optimized procedure for PtNP-modification (n = 10–12). The r² are reported on the plot.

PtNP size on the surface impacts the electrocatalytic conversion of adenosine oxidation products

Adenosine is often identified in tissue by observing the secondary oxidation peak (Figure S-3). At low concentrations, the secondary peak is difficult to identify because it is almost always at least 3-fold less in current than the primary oxidation peak.^8,9 At PtNP-modified surfaces, we have observed previously and we observe here that adenosine’s oxidation reaction is further catalyzed at PtNP-modified surfaces. To test the extent to which the size and morphology of the NP’s impacts this conversion, we quantified the conversion from the primary to the secondary oxidation product on both bare and modified electrodes by calculating the ratio of the secondary oxidation peak to the primary oxidation peak. We divided these two ratios (Ratio_post/Ratio_pre, Figure 5, n = 6) and a number above 1.0 indicates that there was an increase in the conversion. A ratio below 1.0 indicates less conversion after PtNP modification. This was done for each of the Pt concentrations and deposition times to correlate how changes in the size and density on the surface impact this electrochemical property. Overall, we show that the concentration of the deposition solution did not significantly affect the primary to secondary peak conversion (One-way ANOVA, p > 0.05, n = 5). This indicates that changing the density of the particles on the surface does not impact the conversion rate. Manipulating both the density and the size of the NP’s by changing the deposition time did not have an overall effect on the conversion of adenosine primary to secondary oxidation between treatments (One-way ANOVA, p > 0.05, n = 5); however, there was significant improvement when assessing them individually. There was a significant improvement for the 45 s and 60 s deposition time in the conversion of AD oxidation products compared to the bare electrode (One Sample t test, p = 0.0067 (45 s) and p = 0.021 (60 s), n = 5–6). These results indicate that larger clusters of Pt on the surface help facilitate the conversion of adenosine oxidation at the surface. This could be due to momentary trapping of AD between the larger particles on the surface which enable facilitation of AD conversion.

Figure 5. PtNP density on the surface impacts the electrocatalytic conversion of adenosine’s primary oxidation product to the secondary oxidation product.

The ratio of the secondary oxidation peak current to the primary oxidation peak current for the bare electrode and the modified electrode was first calculated. Then, we divided these two ratios (Ratio_post/Ratio_pre). A number above 1.0 indicates an increase in conversion between the primary and secondary oxidation peak. (A) No consistent trend in conversion was observed with increasing Pt concentration during deposition (n = 5). (B) Significant increases in conversion were not observed with increasing deposition time between treatments (One-way ANOVA, p > 0.05, n = 5–6).

PtNP density and size impacts the interaction of adenosine at the surface

The density and size of PtNPs on the surface impacts the interaction of AD, but not DA, on the electrode surface. To test the extent to which the rate-limiting step at the electrode changes as a function of PtNP density and size, we tested the change in oxidative current as a function of scan rate. Scan rates tested ranged from 50 to 800 V/s. Data was plotted as the log of current vs the log of scan rate. Log-log plots can be used to analyze shifts in the behavior at the surface by observing the slope. Slopes closer to 0.5 indicate diffusion-limited processes, whereas slopes closer to 1.0 indicate adsorption-limited processes. Contrary to prior reports for AD with FSCV, we observed that AD was primarily diffusion-limited at bare carbon-fibers (Figure 6A, pink). Prior reports have suggested adsorption-control; however, comparing r² values for scan rate experiments at rapid scan rates can be somewhat inconclusive (r² values are often very close to one another). A log-log plot may be a better metric for analyzing FSCV data. Bare CFME’s were compared to electrodes modified with PtNPs for 15 s, 30 s and 45 s. The 60 s deposition time (Figure 6B, C, and D, respectively) was not tested because significant differences in the morphology and density compared to 45 s was not observed between these deposition times. An increase in the slope of the log-log plot was observed for AD as a function of increasing deposition time, this indicates that increasing the density and size of NPs leads to more adsorption-limited AD interaction at the electrode surface. Conversely, no significant change in dopamine behavior at the surface was observed. To further assess the extent to which AD and DA interacts with PtNPs, we conducted surface-enhanced Raman scattering (SERS) to estimate the apparent Kd on the surface.

Figure 6. The density and size of PtNPs on the surface impacts the interaction of adenosine on the surface more than dopamine.

The scan rate was varied from 50 to 800 V/s. The log of current was plotted as a function of the log of scan rate. A slope (m) closer to 0.5 indicates diffusion-limited interaction and a slope closer to 1.0 indicates adsorption-limited interaction. The r² values are shown on the graph. Dopamine is shown in blue, adenosine is shown in pink. (A) Adenosine was primarily diffusion-limited on bare carbon-fiber electrodes (B).Increasing the deposition time from 15 s (B), 30 s (C), 45 s (D) resulted in an increase in the slope for adenosine (pink) demonstrating a switch to more adsorption-limited interaction. Dopamine remained relatively unchanged (blue). (n = 8–12).

Characterization by Surface-enhanced Raman spectroscopy

To further investigate the role of the Pt surface on the observed electrochemical changes for AD vs. DA, we detected the adsorption of AD and DA on PtNP using SERS. These SERS studies eliminate any possible role of charge and/or of the carbon surface, which could not be ruled out for the electrochemical signal changes. In SERS, the signal from AD and DA can only be observed when the molecule is adsorbed on the nanostructured metallic surface (i.e., PtNP). This effect is due to the weak spontaneous Raman signal and the strong distance dependence of the SERS enhancement. To verify this assumption, we measured the Raman signal from the solutions used in the experiments in the absence of PtNP and, as expected, we did not observe any significant signal for the peaks under analysis (Figure S5). Thereby, the observed signal in the SERS measurements is proportional to the number of molecules on the metallic surface. To further prove that we can use SERS to measure surface coverage on PtNP, we performed a surface coverage experiment using BPE, a molecule commonly used in SERS experiments due to its strong adsorption to metals and cross section. We show a significant enhancement from our PtNP surface (Fig S6) and, as expected, the SERS signal follows a saturation behavior as a function of BPE concentration (Fig S7). By exposing SERS substrates composed of PtNP, we demonstrated how AD has a higher affinity than DA to adsorb/bind on the surface of PtNP.

To determine that AD has a higher affinity than DA to adsorb/bind on the surface of PtNP, we immersed PtNP substrates in solutions containing AD and DA at increasing concentration in 0.1 M HCl. Figure 7A and 7B show the SERS peaks for different concentration of AD and DA, respectively. For completeness, the full Raman and SERS spectra for these molecules are available in the supporting info with labels for the peaks used in the analysis (Figure S8 and S9). As it can be observed, the signal for AD is detected at lower solution concentrations, 10 mM, in contrast with the 50 mM for DA. To prove that the observed result is not intrinsic to these molecules but rather dependent on the number of molecules adsorbed on the surface, we also tested the signal from AD and DA solutions (100 mM) dried on an aluminum surface, where we do not expect any SERS. The spontaneous Raman signal from the dried solutions was observed to be similar (less than 18% difference), which demonstrate that the Raman cross section (intrinsic Raman signal intensity) for these molecules are similar. This result also agrees with what is reported in the literature.³⁴ Thereby, the signal observed at lower concentrations for AD is indicative of a higher number of molecules adsorbed on the surface (i.e., higher affinity). Note that the high concentrations used in these experiments to observe SERS signal (i.e., mM range) are due to the fact that Pt is not an ideal material to perform SERS and has a low enhancement; however, even with Pt we observed significant SERS enhancement that permitted to complete these studies.

Figure 7. SERS shows the analytes bind to PtNP with different binding affinities.

Low to high concentrations of AD and DA were added to PtNPs that are fixed to a glass substrate. The PtNP were about 50 nm in diameter. For AD, (A) a signal can be detected as low as 1 mM AD and up to 100 mM AD were adenosine is no longer soluble. DA (B) signal is not detected until 10 mM DA and in general has lower signal intensity than AD.

To try to quantify the difference in affinity between AD and DA on PtNPs, we measured the SERS signal for the different solution concentrations. Because the SERS signal is proportional to the surface coverage, we plotted the SERS signal normalized for the max observed signal (i.e., 100% coverage for our model) as a function of analyte concentration. We fit this plot with a modified Langmuir isotherm (Figure 8). The fit used to describe the surface saturation was $y=\frac{[X]}{[X]\cdot k_d}$, where [X] represents the analyte concentration and Kd the dissociation constant. This fit enables calculating the apparent Kd for the two molecules. We observed a Kd of 22 ± 8 mM for AD and 48 ± 12 mM for DA (n = 3). We used more data points for DA than for the AD curve because AD is not soluble at concentrations higher than 100 mM. The increase in Kd for DA is in agreement with our discussion above and our central hypothesis that AD has a stronger affinity for Pt, causing an increased electrochemical signal when electrodes are electrodeposited with PtNP’s.

Figure 8. Modified Langmuir isotherm to demonstrate binding affinity of analyte to PtNPs.

(A) Shows AD from low to high concentrations as it binds to PtNPs and (B) DA. (n = 3).

Conclusions

Platinum nanoparticles electrodeposited to a carbon fiber electrode surface is a relatively easy and quick procedure, and can be very useful for biological sensing.^21,22 In this paper, we carefully control the nanoparticle size and morphology and correlate these changes to electrochemical detection of AD and DA. We demonstrate that the larger the nanoparticles and the more densely they are packed on the surface, the larger the increase in AD oxidative current. Opposite results were observed for DA detection indicating the differences in how these two analytes interact at varying surfaces. To further quantitate this, we used SERs to demonstrate that AD has a higher affinity for Pt than DA. Overall, this work provides a fundamental study of how electrode surface material, topology, and morphology can have differential impacts on neurochemical detection depending on the target analyte.