Towards the rational design of N-(1,3-dimethylbutyl)-N′-phenyl-1,4-benzenediamine (6PPD) electrochemical sensor

Abstract

N-(1,3-Dimethylbutyl)-N′-phenyl-1,4-benzenediamine (6PPD) is a common additive in tires. 6PPD protects rubber from oxidative damage by ozone. Leaching of 6PPD in the environment leads to the formation of harmful byproducts such as 6PPD quinone. In this work we provide the fundamental basis for the detection of 6PPD by electrochemical techniques. We use cyclic voltammetry to study the adsorption of 6PPD on glassy carbon. We show that adsorbed 6PPD can be reversibly oxidized and reduced without disturbing the adsorption process. This result enables repeated electrochemical titrations. We determine, in neutral condition at 22 °C, an adsorption constant of K_ads = 1.2 ± 0.5 μM⁻¹ and kinetics of adsorption k_ads∈[0.74–5.60] × 10⁴ L mol⁻¹ s⁻¹. Based on this knowledge we demonstrate the lowest concentration of 6PPD ever detected electrochemically, 10 nM. We also identify current challenges for electrochemical sensing of 6PPD. Multiple layers are formed at concentrations above 4.6 μM and the slow kinetics of adsorption requires long (hour) measurement time to reach maximum sensitivity.

---

Introduction

N-(1,3-Dimethylbutyl)-N′-phenyl-1,4-benzenediamine, referred as 6PPD, is a common antiozonant molecule found in tires. 6PPD reacts with ozone to form several oxidation products, one of them being a quinone derivative often referred as 6PPD quinone. 6PPD quinone was found to be toxic for Coho salmon with an LC₅₀ of only 0.8 μg L⁻¹ (2.7 nM).¹ Brooke trout and Rainbow trout also have notable LC₅₀ values of 0.59 μg L⁻¹ (7 nM) and 1.00 μg L⁻¹ (3 nM), respectively.² While toxicity of 6PPD has not been evidenced yet, it was shown that 6PPD can be oxidized in water to form 6PPD quinone.³ The presence of both 6PPD and 6PPD quinone was evidenced in runoff water, dust and soil.^4–9 6PPD and 6PPD quinone have also been detected in adult human urine samples in southern China at 0.018 ng mL⁻¹ (67 pM) and 0.4 ng mL⁻¹ (1.34 nM), respectively.¹⁰ The American Environmental Protection Agency (EPA) has a screening value set for 6PPD and 6PPD quinone at 8.9 μg L⁻¹ (33 nM) and 11 ng L⁻¹ (37 pM), respectively.^11,12 This environmental threat generated a significant interest for developing methodologies of detection.

A major difficulty in detecting 6PPD and 6PPD quinone is their low concentration in the environment. The solubility of 6PPD and 6PPD quinone in water at 25 °C are reported to be 7 μM and 0.3 μM, respectively.^9,13 6PPD is colourless and 6PPD quinone absorbs slightly in the visible giving a pale-yellow colour. While their spectral properties in the UV-Visible range enables mechanistic studies in relatively concentrated solutions,³ it is so far not amenable for the detection of nM concentrations.

The gold standard for detection of 6PPD in samples from our environment is mass spectrometry (MS). Huang et al. extracted 6PPD and other analogue molecules from dust samples using acetonitrile and a mixture of dichloromethane:hexane introduced in an HPLC-MS. LOQ as low as 0.11 ng of 6PPD per g of dust is reported.⁴ Seiwert et al. used a combination of LC-HRMS to directly detect 6PPD quinone in filtered samples of melted snow and runoff water. They report an approximate LOQ of 84 pM.⁸ One major interest of those techniques is the ability to sperate and analyse complex samples containing multiple transformation products.^4,8,13–15 The cost and size of the instrument as well as the need for qualified operators is nonetheless preventing its usage outside of the lab.

On the other hand, electrochemistry is well-suited for rapid and inexpensive detection in liquid samples.^16,17 Detection of organic pollutants like pesticides,¹⁸ drugs,^19,20 and hormones^21,22 has been achieved in aqueous samples. In addition, metals like Pb, Cd, As, U, Tl, Cr, Ag and Cu are readily quantified by anodic stripping voltammetry.^23–25 While the LOQ is highly dependent on the target molecule and strategy of detection, concentrations of nM – pM of organic molecules are commonly detected in electrochemistry, making it competitive with sensitive analytical techniques like fluorescence. The relative simplicity of the electronics, ability to perform multiplexed measurements and usage of disposable electrodes reduce considerably the cost of electrochemical detection. For example, electrodes can be mass-produced by screen-printing a carbon paste on flexible plastic supports or paper sheets.²⁶ For these reasons it appears desirable to develop an electrochemical sensor for 6PPD and 6PPD quinone.

In 2024 Ye and coworkers reported the electrochemical response of aqueous solutions of 6PPD based on a molybdenum, nitrogen-doped carbon paste electrode.⁵ Using differential pulsed voltammetry and a calibration curve they determined concentrations of 6PPD in different samples of soil with values ranging between 2 μM and 21 μM. Their data suggests adsorption of 6PPD onto the electrode. However, the calibration curve does not follow any known isotherm, and no information is provided regarding the kinetics. While development of electrode materials with high affinity for a target is critical to improve the performance of a sensor, we must also recognize that transduction between a chemical event (adsorption in our case) and electrical signal (faradaic current from oxidizing/reducing 6PPD) contributes to the overall performances of a sensor. Thus, design of sensors requires fundamental knowledge of this transduction mechanism. In this report we investigate the redox mechanisms and adsorption of 6PPD involved in the electrochemical response. Equipped with this knowledge, we demonstrate detection of only 10 nM of 6PPD with inexpensive glassy carbon.

Experimental section

Reagents

6PPD, acetonitrile (≥99.9%), ethyl alcohol (95%), potassium phosphate monobasic (≥99.0%), potassium phosphate dibasic (≥98.0%), potassium chloride (99.0–100.5%), agarose, potassium nitrate (≥99.0%), acetic acid (≥99.7%), sodium hydroxide pellets (98%), were purchased from Sigma Aldrich (USA). The chemicals were used without further purification. Deionized water (18.2 MΩ cm) was obtained from a Barnstead MicroPure water purification system with a 0.2 μM ultra-filtration membrane. Potassium phosphate buffer at 100 mM was prepared using potassium phosphate dibasic and monobasic forms. The pH was measured to be 6.9 using a Ohaus pH meter. The basic (pH 10.0) and acidic (pH 4.0) phosphate buffers were prepared by spiking the pH 6.9 phosphate buffer with 1 M sodium hydroxide and 1 M acetic acid, respectively.

The aqueous 6PPD solutions were prepared by diluting a 5 mM stock solution of 6PPD in acetonitrile into the appropriate volume of phosphate buffer. The stock solution is stored in the dark at 5 °C. No change was observed over several weeks. However, the aqueous 6PPD solution is unstable in presence of air and light (the colour turns from clear to yellow within a day when exposed to air and light) and thus, was always kept in the dark and prepared freshly every five days. For all measurements the content of acetonitrile in water is at most 1.6%.

Electrochemical measurements

A CH601E potentiostat from CH Instruments (Austin, TX) and a three-electrode glass cell placed inside a home-made copper Faraday's cage were used to record electrochemical measurements in the dark. The working, counter and reference electrodes were a 3 mm diameter glassy carbon disk, 0.5 mm diameter × 32 mm long platinum wire and Ag/AgCl 1 M KCl from CH Instruments (Austin, TX), respectively. The reference electrode was placed in a salt bridge (5% agarose 1 M KNO₃) to avoid contamination of the reference by 6PPD as well as leakage of chloride. The salt bridges were prepared by filling a glass Pasteur pipette with molten agarose gel (5%_m in 1 M KNO₃ aqueous solution) and letting the gel cool down inside the pipette. Bridges were stored in large volumes of 1 M KNO₃ and changed regularly to avoid cross-contamination.

Electrode cleaning

Glassy carbon electrodes were sequentially polished with 1.0 μm, 0.3 μm and 0.05 μm alumina slurry on a microcloth pad and rinsed with water. The alumina powder and microcloth pads were purchased from CH Instruments (Austin, TX). This procedure ensures a reproducible surface and removal of the adsorbed 6PPD.

Results & discussion

Voltammetry of 6PPD adsorbed on carbon

The adsorption of 6PPD on glassy carbon was studied as illustrated in Fig. 1A. First, the cyclic voltammogram of a 3 mm diameter glassy carbon electrode was recorded in 100 mM phosphate buffer pH 6.9. Then, the electrode was rinsed with deionized-water water and transferred into a vial containing a solution of 6PPD in 100 mM phosphate buffer pH 6.9. The concentration of 6PPD and time of incubation were varied between 0.1 μM–20 μM and 5 min–2 h, respectively. The temperature was kept between 20 °C and 22 °C (room temperature). Finally, the electrode was rinsed with deionized water and placed again in the electrochemical cell to record a second cyclic voltammogram. Fig. 1B shows cyclic voltammograms recorded prior to incubation (dashed line) and after incubation (continuous line) at a scan rate v = 0.1 V s⁻¹. The arrow indicates the beginning of the scan. The blank shows no peaks. However, a pair of gaussian-shaped peaks is observed after incubation. These peaks were observed with eight different glassy carbon electrodes over several hundred of independent experiments. They are attributed to the oxidation and reduction of 6PPD molecules adsorbed onto glassy carbon. The mid-position between the anodic and cathodic peaks, E_1/2, is 0.083 ± 0.004 V. This value is extremely reproducible (5% of variation) regardless the day and electrode and thus can be used as a reliable fingerprint for identifying 6PPD.

Fig. 1 (A) Illustration of the experimental protocol used to adsorb 6PPD onto the electrode. A blank is measured in 100 mM phosphate buffer pH 6.9 (blue colour) and then, the electrode is incubated in a solution of the buffer containing 6PPD (salmon colour). Finally, the electrode is rinsed with the buffer and introduced into the electrochemical cell containing only the buffer (i.e. no 6PPD). (B) Typical cyclic voltammograms recorded with (dashed line) a pristine 3 mm diameter glassy carbon electrode and (continuous line) the same electrode after adsorption of 6PPD. The conditions of adsorption are 20 min in 1 μM of 6PPD. v = 0.1 V s⁻¹. (C) Anodic (red points) and cathodic (black points) peak current versus scan rate. The continuous red and black lines are fits of the equation y = ax with slopes of 30.7 ± 0.7 μA s V⁻¹ and −30.2 ± 0.7 μA s V⁻¹, respectively. R² = 0.996 (for both fits). (D) Anodic and cathodic peak potential (red and black point, respectively) as a function of the logarithm of the scan rate. The continuous lines are simulated using a Butler–Volmer model to represent the electron transfer kinetics. The values of the standard rate constant of electron transfer and transfer coefficient are 50 s⁻¹ and 0.5, respectively. See ESI† for details about the simulation.

Cyclic voltammograms were recorded between 0.010 V s⁻¹ and 1.000 V s⁻¹. The anodic and cathodic peak currents (i_p) are plotted as a function of the scan rate in Fig. 1C. The continuous lines are linear fits with the intercept forced to zero (|average slope| = 30.5 ± 0.7 μA s V⁻¹, R² = 0.996). The proportionality between peak current and scan rate is consistent with the voltammetry of immobilized redox molecules.²⁷ The position of the anodic and cathodic peaks (E_p,a and E_p,c, respectively) is plotted as a function of the scan rate in Fig. 1D. For scan rates below 0.1 V s⁻¹ the peak splitting, ΔE_p = |E_p,a − E_p,c|, has a constant value of 15 ± 9 mV. Above 0.1 V s⁻¹ the peak splitting increases. Such increase reflects the kinetics of electron transfer.²⁸ We used the Butler–Volmer model to represent the kinetics of electron transfer with only two parameters, the standard rate constant of electron transfer, k° and the transfer coefficient, α. The continuous lines in Fig. 1D correspond to the simulated peak splitting for an adsorbed redox couple with a k° = 50 s⁻¹ and α = 0.5. Details about the simulation are provided in (ESI, Fig. S1†). Note that the peak splitting is expected to be zero at low scan rate.²⁷ However, non-null values are commonly encountered in disorganized layers.²⁹ Another explanation could be the sequential transfer of two electrons with two apparent standard potentials close form each other (∼30 mV).³⁰ This non-null peak splitting was added to the simulation of the electron transfer kinetics.

The voltammetry of the adsorbed 6PPD was studied as a function of pH. The red, black and blue cyclic voltammograms in Fig. 2A were recorded with an electrode incubated in 1 μM 6PPD (pH 6.9) and then immersed in 100 mM phosphate buffer at pH 4.0, 6.9 and 10.0. The width and more importantly position of the peaks vary with pH. The lower the pH the higher E_1/2. Fig. 2B shows the position of the pair of peaks (E_1/2) as a function of pH. The red line is a linear fit with a slope of −59 mV per pH and R² = 0.999. Such high correlation can be advantageously used to deduce the pH of a sample from the electrochemical measurements.

Fig. 2 (A) Cyclic voltammograms recorded with a glassy carbon electrode modified with 6PPD immersed successively in 100 mM phosphate buffer at pH 6.9, 4.0 and 10.0. v = 0.1 V s⁻¹ The dashed lines are the blanks. (B) Mid-potential of the pair of peaks as a function of pH. The red line is an adjustment of a linear function on the experimental point with y = −0.059 x + 0.498 and R² = 0.999.

For a Nernstian system the surface concentrations of oxidized and reduced forms of 6PPD (Γ^ox_6PPD and Γ^red_6PPD, respectively) obey the Nernst law:

where E_WE, , m, n, R, T and F are the potential of the working electrode, the standard potential of the adsorbed 6PPD redox couple, the number of protons exchanged, the number of electrons exchanged, the ideal gas constant, the temperature and Faraday's constant, respectively. The summation of the first and second terms on the right of eqn (1) corresponds to the apparent standard potential. The second term was introduced to reflect the shift of the peak position as a function of pH. A shift of the apparent standard potential by −59 mV per decade of pH corresponds to a stoichiometry of one proton per electron (m/n = 1). A similar observation was made by Ye and coworkers.⁵ The shift of the peak with pH and the gaussian-shape of the peaks are reminiscent of quinone/hydroquinone compounds adsorbed on carbon surfaces.^31–33 These observations are also similar to the electrochemical features of p-phenylenediamine compounds, well-known to undergo electrochemical oxidation at two electrons and two protons, forming a di-imine compound.^34–37 We propose that 6PPD is oxidized at two electrons (n = 2) and two protons (m = 2) to a di-imine product as shown in Scheme 1 below.

Scheme 1 Molecular structures of the oxidized and reduced forms of 6PPD adsorbed on the glassy carbon electrode. Two protons and two electrons are transferred.

For a Nernstian redox couple immobilized on an electrode, the FWHM is given by the relation FWHM = 3.53 RT/(nF).³⁸ For n = 2 the FWHM is expected to be 45 mV. The experimental FWHM (67 mV ± 8) is larger than the expected value. Broadening of the peak may be an indication of repulsive interactions between adsorbates. Peak broadening may also be caused by a dispersion of and k°.^39–43

After characterizing the electrochemical behaviour of adsorbed 6PPD we investigated the stability of 6PPD on the surface (i.e. loss of adsorbed 6PPD by desorption) as well as the reversibility of the redox reaction (i.e. ability to reduced back all the oxidized 6PPD). Five consecutive scans were recorded and the anodic and cathodic charge, Q, were calculated by integration of the anodic and cathodic peaks, respectively. Anodic and cathodic charges are plotted in red and blue in Fig. 3 as a function of the number of the cyclic voltammogram. The inset in Fig. 3 shows a cyclic voltammogram with the baseline (dashed lines) used for background subtraction before integration. For each cyclic voltammogram the anodic and cathodic charges are identical within the error bars. This indicates that all the 6PPD molecules oxidizing during the electrochemical measurement can be reduced back. Comparing the average charge (anodic and cathodic charges) of the first and fifth cyclic voltammograms evidence a relatively small loss of 6%. In other words, the process of sweeping the potential and oxidizing/reducing the adsorbed 6PPD does not significantly perturbate the reaction of adsorption. This result is important because it allows continuous monitoring of the amount of 6PPD. Long-term stability was assessed by recording cyclic voltammograms over an extended period of 1 h (see ESI, section 1†). A loss of 29% of the initially adsorbed 6PPD is evidenced. We attribute this loss to the slow desorption of 6PPD. This observation indicates that reading of the amount of adsorbed 6PPD, if performed in a solution not containing any 6PPD, must be carried out within minutes.

Fig. 3 Anodic (red colour) and cathodic (blue colour) charge integrated under the anodic and cathodic peaks of 5 consecutive cyclic voltammograms. The inset shows an example of cyclic voltammogram with the baselines (dashed lines) and integrated area (shades area). The error bars correspond to the error made during baseline subtraction.

Application of a potential well beyond the standard potential of 6PPD leads to complete oxidation/reduction of the adsorbed molecules. Under such condition integrating the faradaic current provides the amount of adsorbed 6PPD without need of calibration. The surface concentration of 6PPD, Γ, is given by the relation:

where A is the surface of the electrode (0.0707 cm²). In the next section we will use this relation to monitor the surface concentration of 6PPD as a function of the concentration of 6PPD in solution and time.

Adsorption isotherm

We will first focus on the adsorption reaction at equilibrium. A glassy carbon electrode was incubated in aqueous solutions of 6PPD of increasing concentration (0.1 up to 20 μM) for durations ranging from 5 min up to 4 h. A cyclic voltammogram was recorded in 100 mM phosphate buffer (pH 6.9) between each incubation and the surface concentration was determined as described previously. Fig. 4 shows the surface concentration as a function of the concentration of 6PPD in solution. The red curve is a least square adjustment of the following equation:⁴⁴where [6PPD], K_ads, Γ₀, q and β are the concentration of 6PPD in solution, the adsorption constant, surface concentration for a saturated monolayer, and two adjustable parameters describing the divergent part of the isotherm. This equation corresponds to a specific case of the Aranovich–Donohue isotherm developed by the two authors to describe adsorption of multiple layers in liquid phase. This model was designed for the adsorption of poorly soluble organic molecules on carbon surfaces.^45,46 The first term in parathesis corresponds to a Langmuir isotherm and represents the adsorption of the first monolayer of 6PPD. The second term in parenthesis represents the adsorption of multiple layers and reproduces the divergent part of the isotherm (above 4.6 μM). An excellent agreement is observed between the adjusted curve and the experimental points (R² = 0.999). From this adjustment we determine an adsorption constant (for the first monolayer) of K_ads = 1.2 ± 0.5 μM⁻¹ and a concentration at saturation of Γ₀ = 98 ± 20 pmol cm⁻². A Langmuir isotherm is also plotted using those two parameters (blue trace in Fig. 4).

Fig. 4 Adsorption isotherm of 6PPD on glassy carbon. The amount of adsorbed 6PPD is tracked by recording a cyclic voltammogram and integrating the anodic peak. The points are averaged between two or three independent experiments and different electrodes. The error bars correspond to the standard deviation between replicates. The red line is a least squares fit of the eqn (3). The adjusted parameters are K_ads = 1.2 ± 0.5 μM⁻¹, of Γ₀ = 98 ± 20 pmol cm⁻², β = 1.3 ± 0.3, q = 0.02 ± 0.01 μM⁻¹ with R² = 0.999. All measurements were performed at room temperature (T ∼ 22 °C) in the dark.

The value of K_ads is on the same order of magnitude as aptamers, specifically designed to target quinine.⁴⁷ In other words, 6PPD naturally has a very high affinity for glassy carbon surface. As a point of comparison, the adsorption of hexanoic acid on carbon (estimated from the data provided by Aranovich et al.) is only 0.1 mM⁻¹. The unusual strength of the adsorption of 6PPD could be caused by a coplanar configuration of the phenyl groups in 6PPD favouring π–π interactions with the graphite surface. Further increasing the affinity between the electrode surface and 6PPD would likely require a form of molecular recognition.

The saturation value for a 6PPD monolayer is 98 ± 20 pmol cm⁻². This value is similar to the concentration of alkanethiol chains in a compact self-assembled monolayer (about 80 pmol cm⁻²).⁴⁸ This means that 6PPD molecules are densely packed on the surface. The right axis in Fig. 4 corresponds to the fractional coverage of one monolayer (θ = Γ/Γ₀). Above 4.6 μM of 6PPD in solution, the value of θ is larger than one. Fractional coverage reaches about 4 compact monolayers at 20 μM of 6PPD in solution (in presence of 1.6%_v/v acetonitrile). Interestingly, the formation of multiple layers occurs when the concentration of 6PPD in solution is nearing the solubility limit in water, about 7 μM.^3,9 We hypothesize that formation of multiple layers is a manifestation of preferential precipitation localized onto the electrode surface. It is well-known that precipitation occurs at nucleation centres, such as geometrical defects or low energy surfaces.⁴⁹ The lack of control over the nature, and density of nucleation centres may lead to poor reproducibility when forming multiple layers (the error bars in Fig. 4 are increasing significantly above θ = 1). This issue could be mitigated by including geometrical defects. For example, roughening the electrode at the molecular scale could be performed with carbon nanotubes.⁵⁰ Sensors based on induced precipitation are excellent at increasing the concentration of a target locally.⁵¹ The poor solubility of 6PPD in water and its preferential precipitation at the electrode surface could be used to improve sensitivity.

Kinetics of adsorption

Adsorption kinetics were measured by incubating an electrode in a solution containing 0.1 μM of 6PPD and recording voltammograms directly in the incubation solution (i.e. in presence of 6PPD). This procedure takes advantage of the ability to continuously monitor the amount of 6PPD adsorbed without affecting the adsorption process itself. The concentration of 6PPD in solution was chosen to form about 10% of a monolayer (according to the isotherm in Fig. 4). This low concentration of 6PPD in solution allows determining adsorption kinetics within the theoretical framework of a Langmuir adsorption process while minimizing the diffusional component of the current. Numerical simulations of the electrochemical response (see ESI, Fig. S2†) indicate that diffusional current introduces, at most, an error of +4% on the charge attributed to adsorbed species. Fig. 5 shows the adsorption kinetics (black points) as well as adjustments of three models of adsorption kinetics. The red, blue and green lines are adjustments of mixed reaction-diffusion kinetics, reaction-limited kinetics and diffusion-limited kinetics, respectively. The mixed reaction-diffusion kinetics is obtained by solving numerically an analogue to the Ward–Tordai equation as explained in detail in ESI.†⁵² The reaction-limited kinetics obeys the formula:⁵³

Γ = Γ_eq(1−e^{−k_ads[6PPD]t}) (4)

where k_ads is the adsorption rate constant and Γ_eq is the surface concentration at equilibrium. The diffusion-limited kinetics follows the formula:^54,55where D is the diffusion coefficient of 6PPD. Since the value of D for 6PPD is unknown, we used a value of D = 5.24 × 10⁻⁶ cm² s⁻¹ corresponding to n,n′-diphenyl-p-phenylenediamine, a molecule with size and chemical structure similar to 6PPD.⁵⁶ The adjusted parameters for the three equations are gathered in Table 1 below.

Fig. 5 Adsorption kinetics measured with a glassy carbon electrode immersed in a [6PPD] = 0.1 μM solution in 100 mM phosphate buffer pH 6.9. A cyclic voltammogram was measured directly in the 6PPD solution at regular time interval. The points are averaged over two independent experiments and electrodes. The error bars correspond to the standard deviation between replicates. The red, blue and green lines are adjustment of a mixed reaction-diffusion kinetics (see ESI†), a reaction-limited kinetics (eqn (4)) and a diffusion-limited kinetics (eqn (5)), respectively. All the parameters of the adjustments are gathered in Table 1.

Table 1 Adjusted parameters used for modelling the adsorption kinetics shown in Fig. 5. The fixed parameters are given below the table

Model	Reaction	Reaction-diffusion	Diffusion
Fixed parameters: D = 5.24 × 10^−6 cm^2 s^−1, [6PPD] = 0.1 μM.
R ^2	0.991	0.986	0.899
k ads (L mol^−1 s^−1)	5.60 × 10^4	0.74 × 10^4	—
Γ eq (pmol cm^−2)	7.3	10.9	8.3

---

The diffusion-limited kinetics (green line in Fig. 5) gives a low value of R² and cannot reproduce the shape of the experimental curve, especially at short times. We thus ruled out diffusion-limited adsorption kinetics. The best adjustment (highest R² = 0.991) is obtained with the reaction-limited kinetics. We note that a mixed reaction-diffusion kinetics also provides a satisfactory adjustment (R² = 0.986). Both models have two adjustable parameters (k_ads and Γ_eq) that are highly correlated. Lowering the value of Γ_eq increases the value of k_ads. During the adjustment we constrained the value of Γ_eq between the minimum and maximum values previously determined by the adjustment of the Aranovich–Donohue isotherm (0.5 < 10.2 < 17.1 pmol cm⁻²). After propagation of the uncertainty, we can provide a range of value k_ads∈[0.74–5.60] × 10⁴ L mol⁻¹ s⁻¹. This range is close from rate of adsorption of n-alkanethiols and thiophene on gold substrate (10³ L mol⁻¹ s⁻¹). A slow kinetic of adsorption suggests that 6PPD molecules may slowly self-organize on the surface similar to alkanethiols. As the apparent adsorption rate constant is directly proportional to the concentration of 6PPD in solution, we expect that detection of concentrations equal or below 10 nM will require several hours of incubation to lead to a maximum signal.

Limit of detection

In this section we demonstrate the lower limits of detection of 6PPD. The detection was carried out with square wave voltammetry, a pulsed method that naturally removes most of the capacitive current and hence reduces background. Fig. 6 shows the surface concentration measured from square wave voltammograms recorded after 6 h and 23.5 h of incubation in a 10 nM solution of 6PPD. The surface concentration is determined by integration of the peak. Then, the area is divided by the frequency of the pulse, f = 50 Hz (frequency of pulse) as well as the potential increment between cycles of pulse, ΔE_s = 4 mV (the product fΔE_s is equivalent to the scan rate). A typical square wave voltammogram with its baseline are plotted in the inset of Fig. 6 with continuous and dashed lines, respectively. The error bars correspond to the standard deviation between two independent experiments with different electrodes. The surface concentration after 6 h of incubation is 0.2 ± 0.1 pmol cm⁻², which is about 0.2% of a compact monolayer. The first interesting result is that simple combination of glassy carbon electrodes and pulsed technique can lead to a LOD of 10 nM, the lowest reported to date with an electrochemical reading. As can be seen in the inset of Fig. 6, the signal is barely above the noise. This simply reflects the limitation of electrochemical techniques to detect less than 1–0.1% of a monolayer. The second interesting observation is the increase of the signal after almost one day of incubation. This result is consistent with the kinetics of adsorption reported in the previous section. Assuming a reaction-controlled kinetics the adsorption process is expected to take 10 times longer detecting 10 nM than detecting 100 nM. Thus, the adsorption reaction should take between 6 h (67% of equilibrium coverage) to 30 h (99% of equilibrium coverage) to reach equilibrium and give the maximum signal intensity. Note that at such long times the kinetics may fall in the mixed reaction-diffusion regime and natural convection may affect the system.

Fig. 6 Surface concentration of 6PPD on a glassy carbon after 6 h and 23.5 h of incubation in a 10 nM 6PPD solution in 100 mM phosphate buffer pH 6.9. The error bars correspond to the standard deviation of two independent electrodes (A = 0.0707 cm²). The inset shows a typical square wave voltammogram recorded with ΔE_s = 0.004 V, ΔE_p = 0.025 V and f = 50 Hz. The dashed line is the baseline used before integration of the peak.

Combining long incubation time, the high affinity of 6PPD for glassy carbon and, sensitive techniques like square wave voltammetry leads to the detection of only 10 nM of 6PPD. Improving the limit of detection would require adsorbing more 6PPD onto the surface. This could be achieved by increasing even further the binding constant between the electrode and 6PPD. We note that increasing the surface of the electrode does not increase Γ or θ and thus is not expected to be a rational way to improve the LOD.

Conclusions

This work evidences the strong adsorption of 6PPD on glassy carbon and importantly the ability to track the amount of adsorbed 6PPD by electrochemistry. Electrochemical readout is commonly employed for sensing and this work offers the basis for the development of rationally designed electrochemical sensors. Without specific optimization of the experimental protocol and electrode, the LOD for electrochemical detection of 6PPD could be reduced to 10 nM. We evidenced that strategies of detection relying on adsorption on glassy carbon will be limited by the slow adsorption kinetics. To mitigate that issue forced convection could be used. Interestingly, we show that adsorption of 6PPD displays a complex multilayer behaviour above 4.6 μM in solution as well as a dependency of −59 mV per pH unit. While the first phenomena may require particular care for establishing a reliable calibration (at least for sensing above this concentration) the second one could be advantageously used to simultaneously probe pH.

Author contributions

E. D. acquired the data and participated to the development of the methodology. C. R. supervised the project, analysed the data, and wrote the manuscript.

Data availability

Data will be made available on request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

E. D. and C. R. thank Loyola University Chicago for funding this research.

References

1. Z. Tian, H. Zhao, K. T. Peter, M. Gonzalez, J. Wetzel, C. Wu, X. Hu, J. Prat, E. Mudrock, R. Hettinger, A. E. Cortina, R. G. Biswas, F. V. C. Kock, R. Soong, A. Jenne, B. Du, F. Hou, H. He, R. Lundeen, A. Gilbreath, R. Sutton, N. L. Scholz, J. W. Davis, M. C. Dodd, A. Simpson, J. K. McIntyre and E. P. Kolodziej, Science, 2021, 371, 185–189.
2. M. Brinkmann, D. Montgomery, S. Selinger, J. G. P. Miller, E. Stock, A. J. Alcaraz, J. K. Challis, L. Weber, D. Janz, M. Hecker and S. Wiseman, Environ. Sci. Technol. Lett., 2022, 9, 333–338.
3. C. Li, Y. Zhang, S. Yin, Q. Wang, Y. Li, Q. Liu, L. Liu, X. Luo, L. Chen, H. Zheng and F. Li, J. Hazard. Mater., 2023, 459, 132127.
4. W. Huang, Y. Shi, J. Huang, C. Deng, S. Tang, X. Liu and D. Chen, Environ. Sci. Technol. Lett., 2021, 8, 381–385.
5. Z. Yang, B. Cao, X. An, Z. Yu, W. Zhao, F. Su, G. Guan, Y. Zhang, Z. Xie and B. Ye, Talanta, 2024, 266, 125072.
6. G. Cao, W. Wang, J. Zhang, P. Wu, X. Zhao, Z. Yang, D. Hu and Z. Cai, Environ. Sci. Technol., 2022, 56, 4142–4150.
7. J. K. Challis, H. Popick, S. Prajapati, P. Harder, J. P. Giesy, K. McPhedran and M. Brinkmann, Environ. Sci. Technol. Lett., 2021, 8, 961–967.
8. B. Seiwert, M. Nihemaiti, M. Troussier, S. Weyrauch and T. Reemtsma, Water Res., 2022, 212, 118122.
9. P. Klöckner, B. Seiwert, S. Wagner and T. Reemtsma, Environ. Sci. Technol., 2021, 55, 11723–11732.
10. B. Du, B. Liang, Y. Li, M. Shen, L.-Y. Liu and L. Zeng, Environ. Sci. Technol. Lett., 2022, 9, 1056–1062.
11. O. o. Water, Acute Aquatic Life Screening Value for 6PPD, Environmental Protection Agency, 2024.
12. O. o. Water, Acute Aquatic Life Screening Value for 6PPD-quinone, Environmental Protection Agency, 2024.
13. K. Hiki, K. Asahina, K. Kato, T. Yamagishi, R. Omagari, Y. Iwasaki, H. Watanabe and H. Yamamoto, Environ. Sci. Technol. Lett., 2021, 8, 779–784.
14. Y. Zhang, C. Xu, W. Zhang, Z. Qi, Y. Song, L. Zhu, C. Dong, J. Chen and Z. Cai, Environ. Sci. Technol., 2022, 56, 6914–6921.
15. J. Ji, C. Li, B. Zhang, W. Wu, J. Wang, J. Zhu, D. Liu, R. Gao, Y. Ma, S. Pang and X. Li, Food Chem., 2022, 396, 133640.
16. N. J. Ronkainen, H. B. Halsall and W. R. Heineman, Chem. Soc. Rev., 2010, 39, 1747–1763.
17. R. Sivaranjanee, P. Senthil Kumar, R. Saravanan and M. Govarthanan, Chemosphere, 2022, 294, 133779.
18. J. S. Noori, J. Mortensen and A. Geto, Sensors, 2020, 20(8), 2221.
19. B. R. Baker, R. Y. Lai, M. S. Wood, E. H. Doctor, A. J. Heeger and K. W. Plaxco, J. Am. Chem. Soc., 2006, 128, 3138–3139.
20. E. De Rycke, C. Stove, P. Dubruel, S. De Saeger and N. Beloglazova, Biosens. Bioelectron., 2020, 169, 112579.
21. F. J. Arévalo, P. G. Molina, M. A. Zón and H. Fernández, J. Electroanal. Chem., 2009, 629, 133–137.
22. N. Haroon and K. J. Stine, Coatings, 2023, 13(12), 2040.
23. J. H. Aldstadt and H. D. Dewald, Anal. Chem., 1992, 64, 3176–3179.
24. S. Illuminati, A. Annibaldi, C. Truzzi, G. Libani, C. Mantini and G. Scarponi, J. Electroanal. Chem., 2015, 755, 182–196.
25. Q. Ding, C. Li, H. Wang, C. Xu and H. Kuang, Chem. Commun., 2021, 57, 7215–7231.
26. K. Yamanaka, M. d. C. Vestergaard and E. Tamiya, Sensors, 2016, 16(10), 1761.
27. E. Laviron, J. Electroanal. Chem. Interfacial Electrochem., 1979, 100, 263–270.
28. K. Weber and S. E. Creager, Anal. Chem., 1994, 66, 3164–3172.
29. N. Nerngchamnong, D. Thompson, L. Cao, L. Yuan, L. Jiang, M. Roemer and C. A. Nijhuis, J. Phys. Chem. C, 2015, 119, 21978–21991.
30. M. López-Tenés, J. González, E. Laborda and A. Molina, Electrochim. Acta, 2023, 462, 142694.
31. V. Budavári, Á. Szűcs, A. Oszkó and M. Novák, Electrochim. Acta, 2003, 48, 3499–3508.
32. P. S. Guin, S. Das and P. C. Mandal, Int. J. Electrochem., 2011, 2011, 816202.
33. K. Shi and K.-K. Shiu, J. Electroanal. Chem., 2004, 574, 63–70.
34. Y.-H. Bai, J.-Y. Li, Y.-h. Zhu, J.-J. Xu and H.-Y. Chen, Electroanalysis, 2010, 22, 1239–1247.
35. H. Y. Lee and R. N. Adams, Anal. Chem., 1962, 34, 1587–1590.
36. A. Meyer and K. Fischer, Environ. Sci. Eur., 2015, 27, 11.
37. S. M. Sayyah, S. S. Abd El-Rehim, M. M. El-Deeb, S. M. Kamal and R. E. Azooz, J. Appl. Polym. Sci., 2010, 117, 943–952.
38. L. R. Faulkner, A. J. Bard and H. S. White, Electrochemical Methods: Fundamentals and Applications, Wiley, 3rd edn, 2022.
39. O. Alévêque, P.-Y. Blanchard, C. Gautier, M. Dias, T. Breton and E. Levillain, Electrochem. Commun., 2010, 12, 1462–1466.
40. O. Alévêque and E. Levillain, Electrochem. Commun., 2016, 67, 73–79.
41. H. Matsuda, K. Aoki and K. Tokuda, J. Electroanal. Chem. Interfacial Electrochem., 1987, 217, 1–13.
42. W. J. Albery, M. G. Boutelle, P. J. Colby and A. R. Hillman, J. Electroanal. Chem. Interfacial Electrochem., 1982, 133, 135–145.
43. T. M. Nahir and E. F. Bowden, J. Electroanal. Chem., 1996, 410, 9–13.
44. G. d. V. Brião, M. G. C. da Silva, M. G. A. Vieira and K. H. Chu, Colloid Interface Sci. Commun., 2022, 46, 100557.
45. G. L. Aranovich and M. D. Donohue, J. Colloid Interface Sci., 1995, 173, 515–520.
46. G. L. Aranovich and M. D. Donohue, J. Colloid Interface Sci., 1996, 178, 764–769.
47. E. Rahbarimehr, H. P. Chao, Z. R. Churcher, S. Slavkovic, Y. A. Kaiyum, P. E. Johnson and P. Dauphin-Ducharme, Anal. Chem., 2023, 95, 2229–2237.
48. H. Hinterwirth, S. Kappel, T. Waitz, T. Prohaska, W. Lindner and M. Lämmerhofer, ACS Nano, 2013, 7, 1129–1136.
49. R. P. Sear, J. Phys. Chem. B, 2006, 110, 4985–4989.
50. C. Gao, Z. Guo, J.-H. Liu and X.-J. Huang, Nanoscale, 2012, 4, 1948–1963.
51. K. J. Vannoy, C. Renault and J. E. Dick, Anal. Chem., 2023, 95, 7286–7293.
52. T. Miura and K. Seki, J. Phys. Chem. B, 2015, 119, 10954–10961.
53. C. H. Chang and E. I. Franses, Colloids Surf., 1992, 69, 189–201.
54. J. F. Douglas, H. E. Johnson and S. Granick, Science, 1993, 262, 2010–2012.
55. D. B. Hibbert, J. J. Gooding and P. Erokhin, Langmuir, 2002, 18, 1770–1776.
56. C. L. Yaws and C. Gabbula, Yaws” Handbook of thermodynamic and physical properties of chemical compounds, Knovel, 2003.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4an00973h

---