Electrochemical Impedance Spectroscopy study of the adsorption of adenine on Au(111) electrodes as a function of the pH

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in: JOURNAL OF ELECTROANALYTICAL CHEMISTRY on 2017, available at: https://doi.org/10.1016/j.jelechem.2017.03.021” 1 Electrochemical Impedance Spectroscopy study of the adsorption of adenine on Au(111) electrodes as a function of the pH. F. Prieto*, J. Alvarez-Malmagro, M. Rueda Department of Physical Chemistry.University of Seville.C/ Profesor García González nº 1, 41012 Seville (Spain). phone: +34 954557174, e-mail: [email protected] ABSTRACT The adsorption of the three adenine forms involved in two acid-base equilibria on Au(111) electrodes is studied by Electrochemical Impedance Spectroscopy. The experiments are performed in solutions of pH values 1, 7.5 and 12, at which the cationic, the neutral and the anionic adenine forms are present, respectively. Two adsorption models are adopted, both based on the theory of Frumkin and Melik- Gaykazyan for the adsorption process, but one of them takes also into account a deprotonation step preceding the adsorption step. The same frequency dependence of the impedance data is observed in the three pH media. The Nyquist admittance plots suggest a mixed kinetic control, by adsorption and by diffusion. The analysis of the electrode impedance or admittance as a function of the frequency according to the adsorption models provides the values of adsorption resistance, 𝑅𝑎𝑑, adsorption Warburg coefficient, 𝜎𝑎𝑑 and adsorption capacitance, 𝐶𝑎𝑑 as a function of potential. From the 𝜎𝑎𝑑vs E plot obtained at pH 1 an equilibrium constant for adenine deprotonation at the vicinity of the electrode surface of c.a. 0.03 is calculated, which is three orders of magnitude higher than the equilibrium constant in the bulk solution. The different potential dependency of the adsorption rates at the three pH values is discussed on the basis of a Frumkin isotherm for the adsorption of adenine and a Butler-Volmer type potential dependence for the adsorption rate constant. 2 INTRODUCTION The interaction of biomolecules with well defined solid electrode surfaces is an interesting task because this subject has potential applications in the development of electrochemical biosensors or the design of new nano-platforms for drug delivery. The availability of metallic single crystal electrodes with smooth well oriented surfaces [1], and the biological relevance of DNA derivatives, have prompted us to study the adsorption of DNA bases and derivatives on gold single crystal electrodes. The adsorption of adenine on low index surfaces of gold single crystal electrodes from neutral solutions was characterised by means of cyclic voltammetry and pseudocapacitance measurements [2]. It was found that the adsorption of adenine strongly depends on the surface orientation and on the reconstructed or unreconstructed state of the electrode. It was concluded that the adsorption of adenine induces the lifting of the reconstruction of the Au(111), as was later confirmed by STM experiments [3]. The experimental capacitance data obtained at potentials within the adsorption/desorption signals showed a clear dispersion with the frequency of the ac perturbation that was ascribed to the kinetics of the adsorption process [4].In fact, the smoothness of the single crystal electrode surface, the severe cleaning procedure that assures the absence of any faradaic processes and the specifically designed measurement procedure to avoid interferences by the reconstruction/unreconstruction phenomena permitted the assignment of the frequency dispersion of the capacitance exclusively to kinetics of adenine adsorption and the suitable analysis of the impedance spectra allows us to obtain the characteristic adsorption parameters [4]. In the analysis of the impedance data an adsorption kinetic model was adopted, based on the Frumkin and Melik–Gaykazyan theory [5] that takes into account the mixed kinetic control by transport and by activation of the adsorption [8]. Initially the model assumed a Butler- 3 Volmer potential dependence of the adsorption rate constant and a Langmuir adsorption isotherm, however the impedance equations that were applied were deduced for a model that avoids any ‘a priori’ assumption about the dependence of the rate constants of adsorption/desorption with the potential or about the kind of adsorption isotherm [6–9]. The analysis about the kinetics of adenine adsorption by this impedance model was previously limited to solutions of pH values around 7 in which the neutral form of adenine (AdH3) existing in solution can get adsorbed. However, two other adenine forms, the cationic form (AdH4+) and the anionic form (AdH2-) can be in solution at other pH values, according to the two protonation/deprotonation equilibria [10,11]: HN NN N NH2 H N NN N NH2 H N NN N NH2 pK1 = 4.2 H+ pK2 = 9.8 H+ AdH4+AdH3AdH2- 10 98 7 65 4 3 2 1 Scheme 1.-Acid-base equilibria of adenine in aqueous solutions. The adsorption of adenine as a function of the pH was studied by cyclic voltammetry and ‘in-situ’ Fourier transform infrared reflection absorption spectroscopy (FT-IRRAS) [12,13]. It was concluded that from neutral and acid pH solutions, only the neutral adenine is chemically adsorbed, so in very acid solutions where adenine is protonated in solution, the chemical adsorption process of the molecule must include a deprotonation step. On the contrary, at pH values around the second pKa of adenine, both species involved in the equilibrium, the neutral and the anionic adenine forms can be adsorbed, and the pKa value in the interface is very similar to the value in the bulk solution [13]. 4 Recently, we have deduced the impedance equations for an adsorption model based also on the Frumkin and Melik–Gaykazyan model without any "a priori" assumption about the potential dependence of the rate constant for the adsorption or the type of isotherm, but including a deprotonation step preceding the adsorption process [14]. The impedance equations include the equilibrium and the rate constants for the preceding chemical step and predict a frequency dependence different than the one implicit in the previous adsorption kinetic model [6–9,15] The aim of this article is to take advantage of the possibilities of the impedance analysis to provide kinetic information of adsorption processes on single crystal electrodes in order to explore the differences in the adsorption of the three adenine forms existing in solution at different pH values and to compare their behaviours with the conclusions reached in previous FT-IRRAS studies. With this purpose impedance measurements have been performed at three pH values (pH 1, pH 7.5 and pH 12). EXPERIMENTAL Supporting electrolyte solutions were prepared with highly purified water, freshly obtained from a MilliQ system, and suprapure grade reagents HClO4, NaF and NaOH. Adenine (Sigma-Aldhrich) working solutions of concentration 1 mM were obtained by the addition of the adequate amount of 10 mM stock solution to the electrochemical cell. All the solutions were degassed by bubbling argon during 20 minutes before running each experiment and then a flow of argon was kept over the solution surface during the measurements. 5 A three electrode set-up was used for the electrochemical measurements. The working electrode was a Au(111) single crystal, prepared according to the Clavilier method [1]. Before each experiment it was flame annealed with a butane micro-torch and transferred to the electrochemical cell with a water droplet protecting the Au(111) surface from any eventual contamination of the laboratory atmosphere. The contact with the cell solution was made by the meniscus method at a controlled potential low enough as to maintain the reconstructed Au(111) surface obtained with the flame annealing. The counter electrode used was a gold wire also cleaned by flame annealing. The reference electrode was a Hg/Hg2SO4(s)/K2SO4(sat) electrode connected to the cell via a salt bridge containing the same supporting electrolyte as in the electrochemical cell. All the electrochemical measurements were performed with an IviumStat multipurpose electrochemical system, from Ivium Technologies, equipped with a frequency response analyser. A single sine ac perturbation with amplitude of 10 mV was applied at frequencies ranging from 20 to 7500 Hz. The adsorption of adenine induces the lifting of the reconstruction of the Au(111) surface obtained during the flame annealing process [2]. Previously to any impedance measurement the Au(111) reconstructed surface was scanned at 10 mV s-1 up to a potential higher than the onset of adenine adsorption, Econt , and it was held at that potential during several minutes to guarantee the lifting of the reconstruction of the Au(111) surface. Then, two measurement procedures were assayed: in the accumulative procedure the impedance data at a single frequency are obtained as a function of the potential in two consecutive scans in opposite directions, first towards lower potentials and secondly towards higher potential values. This accumulative procedure of potential scan permits to check if the electrochemically induced reconstruction of the Au(111) surface at low potentials affects the impedance measurements. In cases where this influence manifests a step by 6 step measurement procedure was developed to avoid the interference of the potential induced reconstruction. The procedure was described in [4]. Briefly, it consists in the application of the contact potential (high potential at which adenine is adsorbed) during 60 s before the application of every sample potential, Esample, that is applied during 2 seconds before the impedance measurement. The cycle of potential steps Econt / Esample is repeated for every frequency until the full impedance spectrum has been measured at every sample potential. Following the step by step procedure the impedance spectra are collected at potentials along the adsorption/desorption wave in two series of intercalated potentials, each series with step increment of 40 mV. 7 RESULTS AND DISCUSSION Pseudocapacitance vs potential plots In figure 1 the pseudocapacitance, 𝐶𝑠=(𝜔 𝑍´´)−1, versus potential plots at 30 Hz, obtained for the adsorption of 1 mM adenine in basic media using the accumulative and step by step procedures are shown. E vs SCE/V -0.6 -0.4 -0.2 0.0 Cs/F cm-2 0 50 100 150 200 250 Figure 1.- Cs vs E plots measured at 30 Hz for a Au(111) electrode in contact with 1mM adenine solution in 0.3 M NaF/NaOH of pH 12 with different measurement procedures: step by step procedure from -0.1 V to -0.54 V, with Econt = 0V (filled circles) followed by the step by step scan from -0.52 to -0.32 V Econ t= 0V (hollow circles); accumulative scan from 0 to -0.6 V (filled squares) and subsequent accumulative scan from -0.58 to -0.02 V (hollow squares). The different Cs vs E plots in figure 1 are coincident at potentials higher than c.a. -0.4 V, however clear differences can be observed at lower potentials depending on the measurement procedure. The step by step procedure provides an excellent agreement between the two consecutive scans in opposite directions, confirming that the pretreatment at Econt guarantees the unreconstructed electrode surface before every impedance measurement. However, the Cs values obtained with the accumulative procedure clearly deviate from the plot obtained with the step by step procedure at 8 potentials around the maximum of the plot and show clear hysteresis between the negative and the positive scans. It suggests that the electrochemical reconstruction taking place while scanning the low potential region is affecting the measurement. Therefore, all the results shown below have been measured by the step by step procedure. The influence of pH in the pseudocapacitance vs potential plots at two frequencies values (30 and 938 Hz) for adenine adsorption is shown in figure 2 for 1mM adenine concentration in solution. It can be observed that the maximum of the capacitance plots shifts from c.a. -0.02 V to -0.38 V when the pH changes from 1 to 7.5, but with similar values around 120-150 F cm-2 at 30 Hz. At pH 12 the Cs maximum shifts only 0.06 V, up to -0.44 V, however the value is nearly twofold the values obtained at pH 1 and pH 7.5. E vs SCE / V -0.5 -0.3 -0.1 0.1 Cs / F cm-2 0 50 100 150 200 250 pH 1 pH 7.5 pH 12 Figure 2.- Cs vs E plots obtained at 30 Hz (hollow symbols) and at 998 Hz (filled symbols) with the step by step procedure in two series of alternating potentials for the adsorption of adenine 1 mM on Au(111) electrodes from solutions of pH=1 (triangles), pH=7.5 (squares) and pH=12 (circles). Fig 2 also shows the influence of the frequency on the pseudocapacitance. 15 instead of the electrode impedance is that the real component of the electrode admittance does not depend on the 𝐶𝑑𝑙,Γ parameters, so the experimental data can be fitted to the theoretical equation using only three adjustable parameters. Recently we have shown that the frequency dependence of the real component function 𝑌𝑒𝑙 ′𝜔 ⁄ is to be preferred for the fitting as the 𝑌𝑒𝑙 ′𝜔 ⁄ vs ω1/2 plots can show characteristic peaks which in the case of an adsorption system with a preceding chemical step are very sensitive to the equilibrium and rate constants of the chemical step, [14]. Moreover, the corresponding imaginary function 𝑌𝑒𝑙 ′′ 𝜔 ⁄ provides the value of 𝐶𝑑𝑙,Γ by extrapolation at high frequencies and the value of 𝐶𝑑𝑙,Γ+ 𝐶𝑎𝑑 at the lowest frequencies. Therefore, the combination of both frequency limits provides a good initial value for 𝐶𝑎𝑑 to be used in the fitting of the real component function, so that the three adsorption parameters 𝑅𝑎𝑑, 𝜎𝑎𝑑 and 𝐶𝑎𝑑 can be usually obtained with precision from the fitting of the 𝑌𝑒𝑙 ′𝜔 ⁄ vs 𝜔−1/2 plots to the theoretical corresponding equation. The values of these three parameters so obtained are used as starting values for the fitting of the imaginary function 𝑌𝑒𝑙 ′′ 𝜔 ⁄ vs 𝜔−1/2 to the corresponding theoretical equation in order to obtain also the 𝐶𝑑𝑙,Γ parameter in addition to the other three parameters. The differences between the values of the three parameters that are obtained from the fitting to the real and to the imaginary functions are within the standard deviations with which each parameter is obtained. Moreover, the simultaneous four adjustable parameters analysis to both components of the electrode admittance using as initial estimates the values previously obtained in the separate analysis to each component provides an excellent agreement between the experimental and generated curves and consistent parameters 16 values, which have been plotted in figure 6. The qualities of the fittings are shown in figure 5 for the adsorption at the three pH values studied at potentials close to their respective pseudocapacitance maximums. An excellent agreement is found between the experimental values and the generated curves (represented by continues lines in Fig. 5) using the theoretical equations and the parameters obtained from the analysis. 20 40 Y'el / F cm-2 50 100 150 0.04 0.08 40 80 pH 1 0.0 V pH 7.5 -0.34 V pH 12 -0.44 V  / s1/2 a) 50 100 150 50 100 150 0.04 0.08 100 200 b) Y''el / F cm-2 Figure 5.- Plots of (𝑌𝑒𝑙 ′ 𝜔−1) 𝑣𝑠 𝜔−1/2 (a) and (𝑌𝑒𝑙 ′′ 𝜔−1) 𝑣𝑠 𝜔−1/2(b) at the indicated pH and potential values. The symbols represent the experimental plots and the lines the theoretical ones generated with the parameters obtained from the analysis with the frequency. It is specially remarkable that the admittance data at pH 1 fit the equations corresponding to the circuit in scheme 2, although the general impedance or admittance 17 equations for a model including a deprotonation reaction preceding the adsorption step do not provide the same frequency dispersion [14]. However, the pronounced peak observed in the 𝑌𝑒𝑙 ′𝜔 ⁄ vs 𝜔−1/2 plot for the results at pH 1 suggests the existence of the chemical coupled reaction, [14]. Effectively, in the case of the model including a chemical step preceding the adsorption step the impedance should account also for the rate of the deprotonation step, previous to the adsorption one, that can be expressed by: 𝑣𝑐ℎ𝑒𝑚 =𝑑 𝑐𝐴𝑑𝐻3 𝑑𝑡 =𝑘1𝑐𝐴𝑑𝐻4+ − 𝑘−1𝑐𝐴𝑑𝐻3𝑐𝐻+ (5) 𝑘1and 𝑘−1 are the forward and backward rate constants of the chemical step, respectively. For the second step the same definition of net rate in equation (1), with 𝑐 =𝑐𝐴𝑑𝐻3 is still valid. The equation for the adsorption impedance according to this kinetic model with coupled chemistry has been derived [14], assuming that the mass transport under the ac perturbation is exerted by semi-infinite linear diffusion (with equal diffusion coefficient for the cationic and the neutral adenine molecules) and by the preceding chemical reaction: 𝑍𝑎𝑑 =𝑅𝑎𝑑+ 𝜎𝑎𝑑[𝐾1′ 𝐾1′+ 1(𝑖 𝜔)−1/2+ 1 𝐾1′+ 1(𝑘 + 𝑖 𝜔)−1/2] + 1 𝐶𝑎𝑑 𝑖 𝜔 (6) Where 𝐾1′=𝑘1𝑘−1 ′ ⁄ is the apparent equilibrium constant, with 𝑘−1 ′=𝑘−1𝑐𝐻+and 𝑘=𝑘1+ 𝑘−1 ′ . It must be noted here that the backwards chemical step has been considered a pseudo-first order process, as the concentration of H+ ions is much higher than the adenine concentration, at the experimental conditions at which the model is 18 planned to be applied (pH 1). The frequency dependence of the impedance cell according to equation (6) is different from the dependence expressed in equation (3) for an adsorption process without coupled chemistry. However, two limiting cases can be distinguished (depending on the relative values of the chemical step rate constant and the range of frequencies used in the measurements) for which the frequency dependence is the same as implicit in equation (3). In the low frequency limit, that applies if 𝑘≫ 𝜔, the general equation (6) can be simplified to: (𝑍𝑎𝑑)𝐿𝐹 =𝑅𝑎𝑑+𝜎𝑎𝑑 𝑘−1/2 𝐾1′+1+𝜎𝑎𝑑 𝐾1′ 𝐾1′+1 (𝑖 𝜔)−1/2+(𝐶𝑎𝑑 𝑖 𝜔)−1 (7) Equation (7) involves the same frequency dependence of the impedance cell expressed in equation (3) but with a different meaning for the apparent Warburg coefficient and the apparent adsorption resistance, which in equation (7) includes the equilibrium and the rate constants of the chemical reaction: 𝜎𝑎𝑑 𝑎𝑝 = 𝜎𝑎𝑑 𝐾1′ 𝐾1′+ 1 (8a) And ..... 𝑅𝑎𝑑 𝑎𝑝𝑝 = 𝑅𝑎𝑑+ 𝜎𝑎𝑑 𝑘−1/2 𝐾1′+1 (8b) On the other hand, in the high frequency limit, that applies if 𝑘≪ 𝜔, the general adsorption impedance equation (6) is reduced to an expression identical to equation (2), with the same meaning for the frequency independent parameters. Therefore the frequency dependence of the cell impedance obeys equation (3). 19 We hypothesise that the protonation/deprotonation step is fast, with a rate constant value in the order of the molecular vibrations frequencies (1012-1014 s-1), so in the frequency range used in our measurements the low frequency limit of the impedance equation can be applied to the adsorption of adenine from acid solutions, and the adsorption Warburg coefficient values obtained from the analysis of the data at pH 1 must be considered "apparent σ values", because apart from the true Warburg coefficient defined in equation (11) it includes the equilibrium constant of the chemical step, according to equation (8). Discussion The four frequency independent parameters in eq. (4) include in their definition the kinetic and thermodynamic information of the adsorption process, as expressed in the equations (9-12): 𝐶𝑑𝑙,Γ =(𝜕 𝜎𝑀𝜕 𝐸 ⁄)Γ (9) 𝑅𝑎𝑑 =(𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸 −1(𝜕 𝑣𝜕 E ⁄ )c,Γ−1 (10) 𝜎𝑎𝑑 =1 𝐷1 2(𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸 −1(𝜕 𝑣𝜕 E ⁄ )c,Γ−1(𝜕 𝑣𝜕 𝑐 ⁄ )E,Γ= =− 1 𝐷1/2(𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸 −1(𝜕 Γ𝜕 E ⁄ )c−1(𝜕 Γ𝜕 𝑐 ⁄ )E (11) 𝐶𝑎𝑑 =−(𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸(𝜕 𝑣𝜕 E ⁄ )c, Γ(𝜕 𝑣𝜕 Γ ⁄ )𝑐,𝐸−1 = (12) 20 =−(𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸(𝜕 Γ𝜕 E ⁄ )c Where 𝜎𝑀 is the charge density on the metal surface and D is the diffusion coefficient of the adsorbate, which it is assumed to be the same for the two diffusing species in the case of the preceding chemical step. The values of the four frequency independent parameters obtained from the frequency analysis at every potential along the adsorption wave ( 𝑅𝑎𝑑, 𝜎𝑎𝑑 𝑜𝑟 𝜎𝑎𝑑 𝑎𝑝, 𝐶𝑎𝑑 and 𝐶𝑑𝑙,Γ) are plotted as a function of potential in figures 6a-6c. It can be observed that 𝑅𝑎𝑑 and 𝜎𝑎𝑑 have a minimum, and 𝐶𝑎𝑑 exhibits a maximum at potentials around the onset of adenine adsorption. 21 E vs SCE / V -0.4 -0.2 0.0 Cad or Cdl, / Fcm-2 0 100 200 300 ad /  cm2 s-1/2 0 2000 4000 Rad /  cm2 0 5 10 15 20 25 30 a) b) c) Figure 6.- Plots of 𝑅𝑎𝑑𝑣𝑠 𝐸 (a),𝜎𝑎𝑑𝑣𝑠 𝐸 (b) and 𝐶𝑎𝑑(hollow symbols) and 𝐶𝑑𝑙(filled symbols) 𝑣𝑠 𝐸 (c) obtained from the analysis with the frequency according to equation (4) for the adsorption of adenine on Au(111) from 1mM solutions at pH 1 (triangles), pH 7.5 (squares) and pH 12 (circles). In plot c) the pseudo-capacitance obtained for the supporting electrolyte is included (lines) for comparison. In fig 6b, the values plotted at pH 1 correspond to 𝜎𝑎𝑑 𝑎𝑝. The error bars have been estimated from the standard deviations obtained from fitting procedure. 22 The 𝐶𝑑𝑙,Γ 𝑣𝑠 𝐸 plots in neutral and basic media are nearly coincident with those of the supporting electrolyte in the same conditions at the lower potentials of the adsorption capacitance peaks, and become slightly lower at higher potentials. In acid media the differences with the values in the absence of adenine is somewhat higher, but not significant in view of the lower precision of the frequency analysis (the frequency dispersion of the impedance data is lower) so the 𝐶𝑑𝑙,Γ values calculated can be considered to be close to the pseudo-capacitance of the supporting electrolyte. The comparison of 𝑅𝑎𝑑, 𝜎𝑎𝑑 and 𝐶𝑎𝑑 𝑣𝑠 𝐸 plots obtained at pH 12 and at pH 7.5 shows that at pH 12 the plots are slightly shifted towards lower potentials, about 60 mV for 4.5 pH units. This potential shift with the pH could in principle be ascribed to a different nature of the adsorbate species at each pH values, being the predominant forms of adenine in the bulk solutions the adsorbed species at each of both pHs values, the neutral form at pH 7.5 and the anionic deprotonated form at pH 12. Similar minimum values of 𝑅𝑎𝑑 are obtained for the adsorption at both conditions, suggesting similar kinetic contributions. On the contrary, significant differences at pH 7.5 and 12 can be observed in the minimum values of 𝜎𝑎𝑑 and maximum values of 𝐶𝑎𝑑. At pH 12 these values of 𝜎𝑎𝑑 and 𝐶𝑎𝑑 are lower and higher, respectively than the corresponding values at pH 7.5.According to their definitions in equations (11) and (12), both parameters include the partial derivatives (𝜕 𝜎𝑀𝜕 Γ ⁄)𝐸 and (𝜕 Γ𝜕 E ⁄ )c or their inverses in the case of 𝜎𝑎𝑑. Both factors can be expected to depend on the nature of the adsorbate molecule, especially if the differences between adsorbates at pH 12 and 7.5 involve or not a negative net charge on the molecule. Therefore these differences in the parameters obtained at pH 12 and pH 7.5 are in agreement with the previous spectro- 23 electrochemical studies [13] that suggested that adsorbed species at pHs 7.5 and 12 are the neutral adenine and the anionic deprotonated adenine, respectively. Regarding the data at pH 1, the plots of 𝑅𝑎𝑑, 𝜎𝑎𝑑 and 𝐶𝑎𝑑 𝑣𝑠 𝐸 are shifted c.a. 350 mV towards higher potentials as compared with the plots at pH 7.5. This shift is in agreement with an adsorption process that involves a deprotonation of the predominant cationic species, AdH4+ in solution. On the other hand, the values of 𝑅𝑎𝑑 𝑎𝑝𝑝and 𝜎𝑎𝑑 𝑎𝑝 obtained at pH 1 are lower than those of 𝑅𝑎𝑑 and 𝜎𝑎𝑑 at pH 7.5. Considering that at both pHs the adsorbed species are the neutral form (as shown by the IRRAS experiments [12]) it can be expected that the adsorption/desorption thermodynamic parameters should be equivalent. Then the minimum values of the true Warburg coefficient defined in equation (11) can be considered the same at both pH values. Therefore, from the minimum values of 𝜎𝑎𝑑 𝑎𝑝 and 𝜎𝑎𝑑 obtained at pH 1 and 7.5, respectively, it is possible to estimate the value of the apparent equilibrium constant 𝐾1′ using equation (8a). A value of 0.3±0.1 is obtained in this estimation for 𝐾1′ at pH 1, which corresponds to a dissociation constant of the protonated adenine at the interphase of 0.03±0.01. This value of equilibrium constant cannot explain, however, the lower values for 𝑅𝑎𝑑 𝑎𝑝𝑝 obtained at pH 1 as compared to the 𝑅𝑎𝑑 values obtained at pH 7.5, because according to equation (8b) and the high value of the rate constant for the deprotonation / protonation reactions the adsorption resistance values obtained at pH 1 must be the true adsorption resistances and should coincide with the values obtained at pH 7.5. The differences will be discussed bellow based on the potential dependence of the rate of adsorption and the shifting of the adsorption potential due to the deprotonation step. 24 The value for the apparent dissociation constant in the electrode/electrolyte interface is three orders of magnitude higher than the dissociation constant of AdH4+ in bulk aqueous solutions. The study of the orientation of adsorbed adenine on Au(111) electrode by a quantitative FT-IRRAS method [16] concluded that the neutral adenine form is adsorbed with nitrogen N10 from the amine group and nitrogen N1 from the pirimidinic ring directed towards the electrode. Thus, the interaction of the electric field at the electrode vicinity over the molecular fragment involved in the dissociation equilibrium, N1 and with the protons can be high enough as to explain the value obtained for the dissociation constant of AdH4+ at the vicinity of the electrode surface. Delgado et al. [17] showed that the second pKa value of dicarboxylic acids adsorbed on gold electrodes decreases from the bulk solution value as the uncoordinated carboxylic group becomes closer to the electrode surface, and this behaviour is explained as a consequence of the electrostatic interaction of the acid moiety and the positively charged electrode surface. The combination of the parameters 𝑅𝑎𝑑, 𝜎𝑎𝑑 and 𝐶𝑎𝑑provides the values of the relaxation times of diffusion, 𝜏𝐷, and of adsorption, 𝜏𝐻 [15]. 𝜏𝐷 1/2 = −(𝜕 Γ𝜕 𝑐 ⁄ )E (2 𝐷)1/2 =2−1/2𝜎𝑎𝑑𝐶𝑎𝑑 (13a) 𝜏𝐻= (𝜕 𝛤𝜕 𝑣 ⁄ )𝐸,𝑐 =𝑅𝑎𝑑𝐶𝑎𝑑 (13b) The estimated values of 𝜏𝐷 and 𝜏𝐻at different potentials are summarised in Table I for the respective adenine adsorptions at the three pH values. pH 1 pH 7.5 pH12 31 Electrochem. 107 (1980) 205–209. doi:10.1016/S0022-0728(79)80022-4. [2] C. Prado, F. Prieto, M. Rueda, J. Feliu, A. Aldaz, Adenine adsorption on Au(111) and Au(100) electrodes: Characterisation, surface reconstruction effects and thermodynamic study, Electrochim. Acta. 52 (2007) 3168–3180. doi:10.1016/j.electacta.2006.09.062. [3] C. Vaz-Domínguez, M. Escudero-Escribano, A. Cuesta, F. Prieto-Dapena, C. Cerrillos, M. Rueda, Electrochemical STM study of the adsorption of adenine on Au(111) electrodes, Electrochem. Commun. 35 (2013) 61–64. doi:10.1016/j.elecom.2013.07.045. [4] F. Prieto, M. Rueda, C. Prado, J.M. Feliu, A. Aldaz, Kinetics of adenine adsorption on Au(1 1 1) electrodes: An impedance study, Electrochim. Acta. 55 (2010) 3301–3306. doi:10.1016/j.electacta.2009.12.105. [5] A. Frumkin, V. Melik-Gaykazyan, Determination of the Kinetics of Adsorption of Organic Substances by a.-c. Measurements of the Capacity and the Conductivity at the Boundary: Electrode-Solution, Dokl. Acad. Nauk. 77 (1951) 855–858. [6] T. Pajkossy, T. Wandlowski, D.M. Kolb, Impedance aspects of anion adsorption on gold single crystal electrodes, J. Electroanal. Chem. 414 (1996) 209–220. doi:10.1016/0022-0728(96)04700-6. [7] T. Pajkossy, Capacitance dispersion on solid electrodes: anion adsorption studies on gold single crystal electrodes, Solid State Ionics. 94 (1997) 123–129. doi:10.1016/S0167-2738(96)00507-3. [8] Z. Kerner, T. Pajkossy, Measurement of adsorption rates of anions on Au(111) 32 electrodes by impedance spectroscopy, Electrochim. Acta. 47 (2002) 2055–2063. doi:10.1016/S0013-4686(02)00073-7. [9] D. Eberhardt, E. Santos, W. Schmickler, Impedance studies of reconstructed and non-reconstructed gold single crystal surfaces, J. Electroanal. Chem. 419 (1996) 23–31. doi:10.1016/S0022-0728(96)04872-3. [10] J.M. Broomhead, The Structures of Pyrimidines and Purines. II. A Determination of the Structure of Adenine Hydrochloride by X-ray Methods, Acta Crystallogr. 1 (1948) 324–329. [11] J.M. Broomhead, The structures of pyrimidines and purines. IV. The crystal structure of guanine hydrochloride and its relation to that of adenine hydrochloride, Acta Crystallogr. 4 (1951) 92–100. doi:10.1107/S0365110X51000349. [12] M. Rueda, F. Prieto, A. Rodes, J.M. Delgado, In situ infrared study of adenine adsorption on gold electrodes in acid media, Electrochim. Acta. 82 (2012) 534– 542. doi:10.1016/j.electacta.2012.03.070. [13] J. Álvarez-Malmagro, F. Prieto, M. Rueda, A. Rodes, In situ Fourier transform infrared reflection absorption spectroscopy study of adenine adsorption on gold electrodes in basic media, Electrochim. Acta. 140 (2014) 476–481. doi:10.1016/j.electacta.2014.03.074. [14] F. Prieto, M. Rueda, J. Álvarez-Malmagro, Electrochecmical Impedance Spectroscopy analysis of an adsorption process with coupled preceding chemical step., Electrochim. Acta. In press (2017). doi:http://dx.doi.org/10.1016/j.electacta.2017.02.106. 33 [15] M. Sluyters_Rehbach, Impedance of electrochecmical systems: Terminology, nomenclature and representation. Part I: Cells with metal electrodes and liquid solutions, Pure Appl. Chem. 66 (1994) 1831–1891. [16] F. Prieto, Z. Su, J.J. Leitch, M. Rueda, J. Lipkowski, Quantitative Subtractively Normalized Interfacial Fourier Transform Infrared Reflection Spectroscopy Study of the Adsorption of Adenine on Au(111) Electrodes, Langmuir. 32 (2016) 3827–3835. doi:10.1021/acs.langmuir.6b00635. [17] J.M. Delgado, A. Berná, J.M. Orts, A. Rodes, J.M. Feliu, In Situ Infrared Study of the Adsorption and Surface Acid−Base Properties of the Anions of Dicarboxylic Acids at Gold Single Crystal and Thin-Film Electrodes, J. Phys. Chem. C. 111 (2007) 9943–9952. doi:10.1021/JP071489M. [18] A. Szulborska, A. Baranski, Numerical simulation of kinetically controlled electrosorption processes under cyclic voltammetric conditions, J. Electroanal. Chem. 377 (1994) 23–31. doi:10.1016/0022-0728(94)03437-0. [19] A. Szulborska, A. Baranski, Kinetics and thermodynamics of thioglycol adsorption on mercury ultramicroelectrodes, J. Electroanal. Chem. 377 (1994) 269–281. doi:10.1016/0022-0728(94)03438-9.