Insights into electrode-electrolyte adsorption dynamics via double potential step chronocoulometry and cyclic voltammetry

The crucial role of analyte adsorption in electrode reaction kinetics underscores the need for advanced methodologies to study the extent and nature of adsorbed molecules on electrode surfaces. In this study, we employed the classical yet underutilized technique of double potential step chronocoulometry (DPSC) to investigate the nature of adsorbed molecules on electrodes. We assessed the extent of adsorption of dopamine (DA) and the products of ascorbic acid (AA) and uric acid (UA) at increasing concentrations on carbon nanofiber (CNF) electrodes via DPSC measurements and compared the results with those obtained from the commonly used cyclic voltammetry (CV). DPSC findings indicated significant adsorption of component ions, specifically phosphate and chloride of phosphate buffer saline (PBS) electrolyte, on the electrode surface. The results presented in the manuscript demonstrate that DPSC can be a reliable technique for studying the specific adsorption in DA sensors, thus providing valuable insights for modifying materials to enhance DA sensor performance.


Introduction
The theory of adsorption from the solution phase to the electrode constitutes a complex area within the field of electrochemistry.The interface between the electrode and electrolyte often displays properties distinct from those observed in the solution away from the interface or within the bulk of the electrode.One of the important characteristics of the interface is its ability to attract and adsorb analyte molecules [1].The crucial role of adsorption in the electrochemistry of organic molecules on carbon electrodes has been acknowledged for a significant period [2].Carbon electrodes possess different chemical functionalities and structures at their interfaces, which can retain and adsorb species involved in electrochemical reactions thereby profoundly influencing the heterogeneous kinetics of redox reactions.Adsorption of numerous molecules can also decrease the overall reaction rate [3].Hence, it is critical to evaluate the adsorption trends of the different analytes and their reaction products on the electrode surface to understand and optimize the electrochemical sensor designs for enhanced performance.
CV has been commonly used to estimate the amount of adsorbed electrochemically active species on the electrode surface by studying the linear correlation of the oxidation peak current as a function of scan rate.
The surface concentration is directly proportional to the bulk concentration at very low analyte concentrations.Yet, in cases of high analyte concentrations, CV is often governed by semi-infinite diffusion control, making it challenging to accurately evaluate the extent of adsorption on the electrode surface.Moreover, CV provides limited information about the nature of adsorbed species and their interactions with the electrode surface.A classical technique, DPSC, however, can provide a reliable and straightforward way to quantify adsorbed species on the electrode surface by incrementally stepping, rather than continuously scanning, the electrode potential while measuring the electric charge that results from the applied potential step.This technique allows us to distinguish between the charge associated with double-layer charging (a non-Faradaic process) in the presence of adsorbed species and the charge due to the reaction of adsorbed redox species (Faradaic process), as well as the charge resulting from Faradaic reactions due to diffusion [4,5].In our previous report, we illustrated that an increase in the adsorption of DA and dehydroascorbic acid leads to the improvement of selectivity and overall performance of the DA sensor using CV [6].However, limitations inherent to this method, such as semi-infinite diffusion at higher analyte concentrations, the challenge in distinguishing the nature of adsorbed species, the lack of direct quantification of adsorbed species, and the inability to discriminate between charge produced by Faradaic and non-Faradaic processes, prompted us to conduct further investigations on this matter.
In an attempt to measure quantities of adsorbed reactants or products in the case of DA, AA and UA redox reactions on CNF electrodes, we utilized DPSC by stepping and reversing the electrode potential from − 0.4 to 0.6 V for the duration of 250 ms while measuring the electric charge induced as a result of potential step.CV was also performed at different analyte concentrations and results were compared.The chronocoulograms were recorded also at varying analyte concentrations and in blank electrolyte solution.We correlated CV and DPSC results to evaluate the extent and nature of adsorbed species on electrode surfaces.Anson plots were generated by plotting Q vs t 1/2 and Q r vs θ and their slopes and intercepts were calculated.Measured Q ads values from Anson plots indicated that reactant adsorption takes place on CNF electrode surface in the case of DA redox reaction.Strong adsorption of ions present in PBS electrolyte is also observed through DPSC.Some observations from CV results also supported these findings.While, the case is opposite for UA and AA electrochemical reactions, where the oxidation product of UA and AA redox reaction adsorbs on the electrode surface partially passivating the electrode surface for further reaction.These insights highlight the potential of DPSC to offer detailed insights into the electrochemical species adsorbed, thereby facilitating the understanding of electrode-electrolyte interface for electroanalytical applications.

Experimental section
CNFs were grown in a plasma-enhanced chemical vapor deposition (PECVD) system with the procedure described elsewhere [7].The preparation of CNF electrodes involved placing a piece of sample onto a conductive copper clap, which served as a double-sided FR4 glass fiber substrate with a thickness of 0.3 mm.The sample was enclosed in inert PTFE tape (Saint-Gobain Performance Plastics CHR 2255-2) with a 3 mm hole (radius = 1.5 mm) to define the working area of the electrode and to insulate copper from the electrolyte.To enhance contact between the substrate and the carbon sample, the back side of the carbon sample was scraped with a piece of copper.The uncompensated resistance (Ru) of each electrode was assessed in phosphate-buffered saline (PBS).The PBS solution was prepared by blending 8 g of NaCl, 1.44 g of NaHPO 4 , 0.2 g of KCl, and 0.24 g of KH 2 PO 4 in 1 L of distilled water.The pH of the solution was adjusted to 7.4 using either 2 M NaOH or HCl.Dopamine hydrochloride (Sigma-Aldrich), AA (Merck), and UA (Sigma) were utilized to study the adsorption dynamics.A few drops of 2 M NaOH were added to the stock solution of UA to dissolve it in PBS.The morphology of CNF samples was studied using a scanning electron microscope (SEM) (Zeiss Supra 40 and Zeiss Sigma VP).
The electrochemical measurements were conducted using a Gamry reference potentiostat with a three-electrode setup, including a Ag/AgCl reference electrode and a platinum wire counter electrode.Prior to each measurement, the solutions were subjected to a 10-minute purge with N 2 gas.Freshly prepared electrodes with a radius of 1.5 mm were used for each measurement.Log I pa vs log v plots were drawn using a range of scan rates i.e., 2.5 V/s-10 mV/s.For DPSC measurements, a potential step of − 0.4 to 0.6 V of 250 ms duration was used.Anson plots were drawn according to the information provided in the theory section.The average values and standard deviations of 3 − 4 samples are provided.

Theory
DPSC is a powerful technique that can be used to determine the adsorbed electroactive substances on the electrode surfaces [4,8].This technique involves the measurement of the charge that flows across the electrode-electrolyte interface when the potential of the electrode is changed from a value when no current is flowing to a value when faradaic current flows in the system.The technique helps to distinguish between the charge used to electrolyze the adsorbed species from the charge used to electrolyze the components of the solution.The relationship between charge and the square root of time is given as follows: Where Q and Q r are the charges corresponding to forward and reverse reaction respectively, τ is the time of potential step, t is the time of double layer charging, n is the number of electrons transferred, F is the Faraday's constant, A is the geometric surface area, C is the bulk concentration of reactants, D is the diffusion coefficient, Γ is the electrode surface coverage and Q dl is the double-layer charge.DPSC predicts that the adsorption of the species to be zero when Q vs 2)] plots intersect each other at Q = 0 and their intercepts and slopes are equal.However, in the case of the adsorption of reactants, the intercept of Q vs t 1/2 should be larger than the Q r vs θ intercept.The difference between the two Q-axis intercepts demonstrates the amount of adsorbed reactant because the Q r vs θ plot depicts the corrected Q-axis intercept equal to Q dl .

Results and discussion
CNF electrodes, featuring a forest-like geometry and fiber lengths in the range of a few micrometers (Fig. 1), were employed to investigate the adsorption of electroactive species involved in the redox reactions of DA, AA, and UA.DPSC was conducted in a blank PBS solution to investigate whether the electrolyte components undergo adsorption on the electrode surface.In a system where there is no electroactive adsorption of the product or reactant, the intercepts, and slopes of Q vs t 1/2 and Q r vs θ should be identical.However, as illustrated by the results shown in the table below, this was not observed with CNF electrodes in PBS.The intercepts and slopes of both Q-axes were dissimilar, resulting in Q ads of 0.045 ± 0.002 mC.PBS is a water-based solution enriched with chloride and phosphate ions.The adsorption of ions from PBS onto the CNF electrodes can be attributed to their extensive surface area, which is a consequence of the high population density and length of carbon nanofibers (Fig. 1).Our prior research demonstrated that the abundant surface chemistry associated with carbon nanofibers leads to greater pseudofaradaic contributions to the double-layer charge [9].It is worth mentioning that a partial charge transfer occurs between the electrode and the adsorbing ions instead of the electrode reactions associated with PBS component ions.Note that the zero-charge potential (pztc) of CNFs is approximately 0 V vs Ag/AgCl at pH 7 [10].Consequently, after the forward step, the ions in the electrolyte are exposed to a positively charged surface, and vice versa for the backward step.This leads to the adsorption of the component ions of the PBS on the electrode surface.There is abundant evidence about the adsorption of species present in PBS on carbon surface [11] and about partial charge transfer in case of specific adsorption [12][13][14].The presence of sp 2 carbon on the electrode surface chemisorbs oxygen, forming various functional groups, predominantly carbonyl and phenolic groups.This increase in differential pseudocapacitance consequently influences the rate of electrode processes [15].

Dopamine
DA adsorption is a critical phenomenon that impacts the reaction kinetics, sensitivity, and selectivity of DA sensors.Our previous study demonstrated DA adsorption on CNF electrodes through CV, evidenced by log I pa vs log v plots approaching a slope of 1 at low concentrations.The adsorption of DA was further qualitatively confirmed through a series of washout experiments [6].With the increase in the concentration of DA from 0.7 to 1.6 mM, the oxidation potential of DA shows a shift in the anodic direction (Fig. 2A).Moreover, the effect of semiinfinite diffusion becomes more dominating, leading to a decrease in the slope values of log I pa vs log v with a further increase in concentration, as depicted in Table 1.
To measure the time dependence of charge flow during the DA reaction, double-step potential chronocoulometry was employed (Fig. 2B).To obtain Q ads using chronocoulometric plot of Q vs t 1/2 , knowledge of Q dl is necessary.In cases where the adsorption of the reactant minimally affects the interfacial capacitance, Q dl can be determined from a blank solution.However, in most cases, the adsorption of the reactant affects the interfacial capacitance, making it impractical to assess Q dl values from blank electrolyte solutions in the absence of the analyte.DPSC resolves this challenge by stepping back the electrode potential to its initial value before the experiment concludes.Consequently, the charge at the double layer remains consistent with its initial state at the start of the experiment.
Fig. 3 demonstrates non-linearity in the Q r versus θ plot, with both slopes and intercepts exhibiting a proportional increase as DA concentration rises.Similarly, the slope values of Q vs t 1/2 show a continuous increase with increasing DA concentration.According to the theory of DPSC, when reactant adsorption occurs, the intercepts of Q vs t 1/2 should surpass those of Q r vs θ.Additionally, deviations from linearity in the Q r vs θ plot may arise due to changes in the concentration profile of the product resulting from reactant adsorption, compared to situations where no reactant is adsorbed.Moreover, as demonstrated by Yap et al. [16], in cases of reactant adsorption, the slope of Q r vs θ should exceed the slopes of Q vs t 1/2 .Taken together, these observations suggest that in the electrochemical redox reaction of DA on CNF electrodes, DA, rather than dopamine-o-quinone (DAQ), undergoes adsorption consistent with the anodic shift in the oxidation peak observed with CV measurements.The deviation from linearity observed at the beginning of the Q vs t 1/2 and Q r vs θ plots is due to the delay in attaining the potential step, caused by the finite rise time of the potentiostat.This delay typically lasts a few milliseconds before reaching the desired final potential value.However, this non-ideal region remains constant irrespective of changes in concentrations.
Contrary to expectation, the values of Q ads exhibit a decrease rather than an increase with rising concentration (Fig. 3, Table 1).In blank PBS, we observed that ions comprising the PBS electrolyte adsorb onto the electrode surface, as indicated by the unequal slopes and intercepts of the Q vs t 1/2 and Q r vs θ plots.Upon introduction of DA into the electrolyte solution, it promptly binds to available sites on the electrode surface, enhancing Q ads .However, with increasing DA concentration, the adsorption of DA begins to compete with that of PBS constituents, particularly phosphate ions.This competition leads to the displacement of some electrolyte ions by adsorbed DA, thereby inducing a gradual decline in Q ads .Hence, it is plausible that the reduction in Q ads is attributable to the removal of PBS components from the electrode surface.

Ascorbic acid
The electrochemical oxidation of AA leads to the reversible formation of unstable dehydroascorbic acid via two-electron transfer process, which subsequently undergoes hydrolysis to yield the irreversible end product, 2,3-diketogluconic acid.Above pH 5 only hydrated dehydroascorbic acid monomer and its degradation product 2,3-diketogulonic acid are expected to be present in the solution [17].CV results show a negligible reduction peak indicating that the dehydroascorbic acid is not undergoing reduction to AA due to its rapid hydration reaction (Fig. 4A) [18].The hydrated form of dehydroascorbic acid is reported to adsorb on electrode surfaces causing the partial blocking of the electrode surfaces [19].In our prior investigation, we demonstrated with CV and washout experiments that dehydroascorbic acid adsorbs onto CNF electrodes [6].Results from CV indicate that the current is lower than expected for the semi-infinite diffusion control reaction regime indicated by the slope of log I pa vs log v < 0.5 (Table 2).This discrepancy may arise from the partial obstruction of the electrode surface by electrochemically inactive products of AA oxidation, even at concentrations as low as 0.7 mM, resulting in an electrode fouling effect.

Table 1
Parameters for Anson plots derived from DPSC measurements and slopes of log I pa vs log v calculated from CV measurements at various concentrations of DA.
Concentration DA (mM)  The shapes of the double potential step chronocoulograms suggest that the variations in charge relative to changes in concentration are less pronounced compared to the DA reaction (Fig. 4B).Indeed, with the increase in AA concentration, changes in Q ads are negligible in comparison to Q ads observed in the blank solution (Fig. 5, Table 2).These results are in agreement with the CV measurements, where an increase in the concentration of AA does not cause the oxidation peak current spike as expected in the reactions where reactant adsorption takes place.There is a very slight increase in the average values of the slope of Q vs t 1/2 indicating a slight increase in the mass transfer of AA with the increase in concentration.However, the constant Q ads indicate that AA does not adsorb on the electrode surface.Similarly, the intercepts and slopes of Q r vs θ remain nearly constant irrespective of changes in AA concentration.This further supports the notion of adsorption of electrochemically inactive hydrated dehydroascorbic acid causing subsequent partial fouling of the electrode surface.

Uric acid
UA oxidation involves a quasi-reversible, two-electron transfer process to a short-lived diimine specie (pH 7-9), which subsequently undergoes irreversible chemical reactions to form allantoin, the electrochemically inactive end-product.The diimine, the primary product, yields a reduction peak at pH 8 observed in CV results with a half-life of 21 ms [20].The short half-life of diimine intermediate indicates that it is extremely short-lived and rapidly undergoes further reaction.On CNF electrodes, UA oxidation appears nearly irreversible at 100 mV/s indicated by a negligible reduction peak due to the highly reactive nature of diimine intermediate.However, a more pronounced reduction peak is noted at 0.7 mM concentration, particularly at 1 V/s compared to slower scan rates indicating that the EC step outruns the subsequent chemical step at higher scan rates (Fig. S1).It is worth noting that the reduction peak does not appear even at 1 V/s at higher concentrations, pointing towards the complete blockage of the electrode surface causing the reduction peak to not appear even at high scan rates.
The observation of steeper slope of log I pa versus log v at 0.7 mM (Table 3), followed by lower values at higher concentrations, further suggests the adsorption of UA oxidation products obstructing the electrode surface with increasing concentration.Previous studies have highlighted strong adsorption of unstable diimine at the electrode surface, with the adsorbed state being notably more stable than in solution [21,22].The absence of a reduction peak at high UA concentrations suggests a shortage of adsorption sites on the electrode surface, likely due to blockage by follow-up reaction products.
Furthermore, an increase in UA concentration shifts the oxidation potential cathodically, likely due to the aforementioned adsorption of reaction products on the electrode surface (Fig. S3).However, owing to the complex reaction scheme involving chemical steps and short-lived intermediate species the effect is not as clear as in the case of AA.Finally, pH changes to dissolve UA into PBS may also contribute to this cathodic shift [23].
DPSC results show that intercepts of Q vs t 1/2 and Q r vs θ are nearly constant (Fig. S2, Table 3).Except for the slight change in Q ads at 0.7 mM UA concentration, no notable change is occurring in Q ads as the concentration increases.According to a charge-time relation for double potential-step large-amplitude chronocoulometry model given by Yap et al., [16] if the slope of Q r vs θ is smaller than that of Q vs t 1/2 and constant it indicates adsorption of a product that is deposited on top of the electrode.However, this model assumes that the concentration of the product near the electrode cannot be larger than the solubility of the product.While, in the case of UA, the solubility of UA in the PBS electrolyte is significantly lower for instance, in comparison to AA solubility.This can make the concentration of the UA and its product near the electrode surface exceed its solubility in the PBS solution and result in surface segregation.This could be the reason for the constant, but greater slope of Q r vs θ in comparison to the slope of Q vs t 1/2 plot (Table 3).At increasing concentrations, the negligible changes in Q ads could be pointing towards the passivation of the electrode surface by the follow-up reaction products of unstable diimine.
It is important to note that in our study (both for AA and UA), the substance fouling the electrode surface is electrochemically inactive, unlike the scenario described by Yap et al [16].Nevertheless, considering the combined findings from CV and DPSC, along with an understanding of the reaction mechanisms involved in both processes, we contend that product adsorption and subsequent fouling are probable processes in our case as well.

Conclusions
DPSC was employed to assess the adsorption behavior of DA, AA and UA on CNF electrodes as a function of their increasing concentrations.Comparative analysis with CV was conducted to deepen our A. Kousar and T. Laurila comprehension of the adsorption susceptibility of CNF electrodes.Deviations in the slopes and intercepts of Anson plots in a blank PBS solution suggested significant adsorption of phosphate and chloride ions from the electrolyte onto the CNF electrodes.In the case of DA, changes in Q ads with increasing DA concentration indicated its adsorption onto the electrode surface.Conversely, for AA and UA, constant Q ads and chronocoulogram shapes with increasing concentrations suggested adsorption of their oxidation products onto the electrodes, hindering further reaction.These findings, in conjunction with our prior investigations, elucidate the observed peak shifts during simultaneous measurement of AA, DA, and UA, and provide insights into modifying CNF electrodes to induce selective electrochemical measurements.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table 3
Parameters for Anson plots derived from DPSC measurements and slopes of log I pa vs log v calculated from CV measurements at various concentrations of UA.

Fig. 1 .
Fig. 1.SEM micrographs of CNF electrodes, showing the morphology of the sample (A) tilted to 25 • and (B) a cross-sectional view.