3162
Langmuir 2007, 23, 3162-3171
Interactive Adsorption Behavior of NAD+ at a Gold Electrode Surface Alexis Damian† and Sasha Omanovic* Department of Chemical Engineering, McGill UniVersity, Montreal, Quebec, H3A 2B2, Canada ReceiVed August 11, 2006. In Final Form: December 18, 2006 The adsorption of an oxidized form of nicotinamide adenine dinucleotide, NAD+, on a polycrystalline gold electrode surface and the subsequent surface conformation of the molecule were investigated over a wide temperature and potential range, using electrochemical differential capacitance and PM-IRRAS techniques. The adsorption process was described by the Langmuir adsorption isotherm. The corresponding thermodynamic parameters were determined: the Gibbs energy, enthalpy, and entropy of adsorption. The large negative Gibbs energy of adsorption (-43 ( 4 kJ mol-1 and -39 ( 2 kJ mol-1 on a positively and negatively charged surface, respectively) confirms that the NAD+ adsorption process is highly spontaneous, while the large entropy gain (285 J K-1 mol-1 and 127 J K-1 mol-1 on a positively and negatively charged surface, respectively) was found to represent the adsorption driving force. It was demonstrated that the energetics of the adsorption process is surface-charge controlled, while its kinetics is both mass-transport and surface-charge controlled. A surface-charge dependent conformation model for the adsorbed NAD+ molecule is proposed. These findings suggest that the origin of the NAD+ reduction overpotential is related to the surface conformation of the adsorbed NAD+ molecule, rather than to the electrode Fermi level position.
1. Introduction Nicotinamide adenine dinucleotide NAD(H) (Scheme 1) is an enzymatic cofactor of a significant industrial and biomedical importance. A large number of biochemical reactions catalyzed by redox enzymes (dehydrogenases or oxidoreductases) are based on the use of NAD(H). In its reduced and enzymatically active form, 1,4-NADH, the molecule transfers two electrons and a proton to a substrate in the presence of a suitable enzyme, resulting in the oxidation of NADH to NAD+. Due to the high price of NADH, the electrochemical reduction of NAD+ to NADH (i.e., NADH regeneration) has attracted considerable scientific attention over the years. Unfortunately, the direct reduction of NAD+ on nonmodified electrodes results in a low yield of enzymatically active 1,4-NADH,1-3 which is predominantly due to the dimerization of the NAD free radical formed in the first reaction step (see Scheme 1). Therefore, in order to develop efficient electrochemically based NADH regeneration methods/electrodes, comprehensive knowledge of the interactive behavior of NAD+ at electrode surfaces is required. This assumes extensive research on the NAD+ redox mechanisms and kinetics, adsorption thermodynamics and kinetics, and the conformation of the molecule at electrode surfaces/interfaces. In our previous papers, we discussed the mechanisms and kinetics of NAD+ reduction on a gold4 and ruthenium-modified glassy carbon (RuGC)5,6 electrode. In both cases, the reaction was found to be irreversible and to occur at high overpotential and under mass-transport control. The amount of active NADH produced on pure gold was found to be highly overpotential dependent, ranging from 78% at low overpotentials down to * Corresponding author. Prof. Sasha Omanovic, Department of Chemical Engineering, McGill University, 3610 University Street, Montreal, QC, H3A 2B2, Canada. Phone: (514) 398-4273. Fax: (514) 398-6678. E-mail:
[email protected]. † Present address: Physique de la Matie ` re Condense´e, Ecole Polytechnique, CNRS, 91128 Palaiseau, France. (2) Schmakel, C. O.; Santhanam, K. S. V.; Elving, P. J. J. Am. Chem. Soc. 1975, 97, 5083. (3) Jaegfeldt, H. Bioelectrochem. Bioenerg. 1981, 8, 355. (4) Takamura, K.; Mori, A.; Kusu, F. Bioelectrochem. Bioenerg. 1981, 8, 229. (5) Damian, A.; Omanovic, S. J. Mol. Catal. A: Chem. 2006, 253, 222. (6) Man, F.; Omanovic, S. J. Electroanal. Chem. 2004, 568, 301. (7) Azem, A.; Omanovic, S. J. Mol. Catal. A: Chem. 2004, 219, 283.
Scheme 1. Nicotinamide Adenine Dinucleotide in Its Oxidized Form (NAD+) and Its Reduction to Enzymatically Active 1,4-NADH and Enzymatically Inactive Dimer NAD2a
a
R stands for adenosine diphosphoribose.
28% at high overpotentials. On the other hand, by modifying a GC surface with Ru nanoislands, the NAD+ reduction mechanism was found to change, resulting in a significant increase in the amount of active NADH, up to 96%. Electrochemical systems usingchemicallymodified7-13 andenzyme-mediatedelectrodes14-19 (8) Baik, S. H.; Kang, C.; Jeon, I. C.; Yun, S. E. Biotechnol. Tech. 1999, 13, 1. (9) Long, Y.-T.; Chen, H.-Y. J. Electroanal. Chem. 1997, 440, 239. (10) Shimizu, Y.; Kitani, A.; Ito, S.; Sasaki, K. Denki Kagaku 1993, 61, 872. (11) Beley, M.; Collin, J.-P. J. Mol.. Catal. 1993, 79, 133. (12) Warriner, K.; Higson, S.; Vadgama, P. Mater. Sci. Eng., C 1997, 5, 91. (13) Karyakin, A. A.; Bobrova, O. A.; Karyakina, E. E. J. Electroanal. Chem. 1995, 399, 179. (14) Vuorilehto, K.; Lu¨tz, S.; Wandrey, C. Bioelectrochemistry 2004, 65, 1. (15) Fry, A. J.; Sobolov, S. B.; Leonida, M. D.; Voivodov, K. I. Tetrahedron Lett. 1994, 35, 5607.
10.1021/la062385q CCC: $37.00 © 2007 American Chemical Society Published on Web 02/08/2007
Adsorption BehaVior of NAD+ at Au Electrode
have also been proposed, and the yield in the active 1,4-NADH produced ranged from ca. 75% on a cholesterol-modified goldamalgam electrode7 to ca. 82% on a silver electrode modified with covalently adsorbed L-histidine.8 Since the NAD+ electroreduction reaction is a heterogeneous reaction, the adsorption of NAD+ on an electrode surface might play a significant role in the overall reduction reaction mechanism and kinetics. Therefore, to completely understand the interaction of NAD+ with charged electrode surfaces, it has been of great interest to investigate the thermodynamics and kinetics of NAD+ adsorption on charged electrode surfaces, and also the chargedependent surface orientation of the molecule. Although the NAD+/NADH redox process is well-established to occur at the nicotinamide ring (Scheme 1), the adsorption of NAD+ on electrode surfaces could take place at several possible NAD+ sites due to the complexity of the molecule. Moreover, the orientation and conformation of the molecule at a charged electrode/electrolyte interface could change with the electrode potential and thus influence not only the NAD+ adsorption thermodynamics and kinetics, but also the NAD+ reduction mechanisms and kinetics. Adsorption of NAD+ on mercury,1,20-25 glassy carbon,26 gold,3,27-31 and silver31,32 electrode surfaces has been investigated. Elving et al. concluded that the adenine moiety is the site involved in the adsorption of NAD+ in the potential region near and positive of the potential of zero charge (pzc) on a mercury1,20-24 and glassy carbon electrode.26 They postulated that in an aqueous (i.e., polar) solution and in the absence of specific enzymes, the NAD+ molecule exists in a folded conformation, which brings the nicotinamide and adenine rings together, with the pyrophosphate group as a hinge. They also concluded that both rings are parallel with a mercury electrode surface, with the adenine ring oriented in a flat position to the surface. Thus, when the electrode is positively charged, this conformation model20 suggests that the adenine ring is involved in the electron-transfer process between the electrode and nicotinamide ring, i.e., it serves as a mediator for nicotinamide reduction. Moiroux et al.25 reported that the adsorbed NAD+ molecule rapidly reorients on the mercury electrode surface when the positive charge of the nicotinamide moiety gets neutralized at a negatively charged electrode surface. They suggested that, after the first reduction step, the adsorbed (15) Voivodov, K. I.; Sobolov, S. B.; Leonida, M. D.; Fry, A. J. Bioorg. Med. Chem. Lett. 1995, 5, 681. (16) Sobolov, S. B.; Leonida, M. D.; Bartoszko-Malik, A.; Voivodov, K. I.; McKinney, F.; Kim, J.; Fry, A. J. J. Org. Chem. 1996, 61, 2125. (17) Chen, X.; Fenton, J. M.; Fisher, R. J.; Peattie, R. A. J. Electrochem. Soc. 2004, 151, E56. (18) Kim, S.; Yun, S.-E.; Kang, C. Electrochem. Commun. 1999, 1, 151. (19) Kim, S.; Yun, S.-E.; Kang, C. J. Electroanal. Chem. 1999, 465, 153. (20) Bresnahan, W. T.; Elving, P. J. J. Am. Chem. Soc. 1981, 103, 2379. (21) Elving, P. J.; Bresnahan, W. T.; Moiroux, J.; Samec, Z. Bioelectrochem. Bioenerg. 1982, 9, 365. (22) Jensen, M. A.; Bresnahan, W. T.; Elving, P. J. Bioelectrochem. Bioenerg. 1983, 11, 299. (23) Schmakel, C. O.; Jensen, M. A.; Elving, P. J. Bioelectrochem. Bioenerg. 1978, 5, 625. (24) Bresnahan, W. T.; Moiroux, J.; Samec, Z.; Elving, P. J. Bioelectrochem. Bioenerg. 1980, 7, 125. (25) Moiroux, J.; Deycard, S.; Malinski, T. J. Electroanal. Chem. 1985, 194, 99. (26) Moiroux, J.; Elving, P. J. J. Electroanal. Chem. 1979, 102, 93. (27) Takamura, K.; Mori, A.; Kusu, F. Bioelectrochem. Bioenerg. 1982, 9, 499. (28) Xiao, Y.-J.; Wang, T.; Wang, X.-Q.; Gao, X.-X. J. Electroanal. Chem. 1997, 433, 49. (29) Xiao, Y.-J.; Markwell, J. P. Langmuir 1997, 13, 7068. (30) Xiao, Y.-J.; Chen, Y.-F.; Gao, X.-X. Spectrochim. Acta, Part A 1999, 55, 1209. (31) Taniguchi, I.; Umekita, K.; Yasukouchi, K. J. Electroanal. Chem. 1986, 202, 315. (32) Yang, H.; Yang, Y.; Liu, Z.; Zhang, Z.; Shen, G.; Yu, R. Surf. Sci. 2004, 551, 1.
Langmuir, Vol. 23, No. 6, 2007 3163
free radical undergoes a very fast flat-to-perpendicular reorientation and becomes electroinactive. The use of different spectroscopic techniques, providing molecular-level information, has enabled scientists to establish more evolved models of the adsorption of NAD+ on metal surfaces. Reflectivity measurements3,27 and surface-enhanced Raman spectroscopy (SERS)28-32 have been employed for the investigation of the potential dependence of the NAD+ conformation. Although the adsorption of NAD+ on mercury has been extensively studied, our survey of the NAD(H) literature confirms that there are a number of contradictory conclusions, particularly with respect to the orientation (conformation) of the molecule on the electrode surface. In addition, there is a lack of information on the adsorption thermodynamics and kinetics. To the best of our knowledge, no thermodynamic and kinetic data on the adsorption of NAD+ on gold is available in the literature. Therefore, the major objective of this research was to investigate the thermodynamics and kinetics of NAD+ adsorption on a polycrystalline gold surface and to elucidate the electrode surfacecharge dependent orientation and conformation of the molecule. This will ultimately contribute to the general knowledge of the interactive behavior of NAD+ on a gold electrode surface and provide information useful for the design of modified goldbased electrode surfaces for the regeneration of NADH in bioreactors and biosensors, with the ultimate goal to increase the regeneration yield of active NADH. 2. Experimental Section Reagents and Solutions. The adsorption of nicotinamide adenine dinucleotide (NAD+) was studied in 0.1 M NaClO4 (ACS certified from Fisher) of pH 5.8. Stock solutions of NAD+ (sodium salt, purity 95%, Sigma, N-0632) were prepared in a separate container by dissolving a proper amount of the reagent in the supporting electrolyte. In ATR-FTIR experiments, 98% pure crystalline NADH was used (sodium salt, Sigma, N-8129). Chemicals used for the research were utilized without further purification. All solutions were prepared using ultrapure deionized water (resistivity of 18.2 MΩ cm). Electrochemical Cell and Electrodes. A standard three-electrode, two-compartment electrochemical cell was used in all experiments. The counter electrode was a platinum wire of high purity (99.99%, Johnson Matthey), which was degreased by refluxing in acetone, sealed in soft glass, electrochemically cleaned by potential cycling in 0.5 M sulfuric acid, and stored in 98% sulfuric acid. The reference electrode was a commercially available mercury/mercurous sulfate electrode (MSE; +0.4 V vs SCE), but all potentials in this paper are referred to the saturated calomel electrode (SCE). The working electrode in electrochemical experiments was a polycrystalline gold rotating disc electrode (Autolab, Eco Chemie, 3 mm in diameter). Before each experiment, the gold electrode was polished with diamond paste down to 3 µm, followed by thorough rinsing with ethanol and cleaning in an ultrasonic bath for 5 min in order to remove polishing residues. The electrode was then electrochemically pretreated (cleaned) in 0.5 M perchloric acid by potentiodynamic cyclic polarization (40 cycles) between -0.3 and 1.5 V at a scan rate of 300 mV s-1. From scan-dependent cyclic voltammetry measurements made in the electrochemical double-layer region in 0.5 M perchloric acid, the electrochemically active surface area of the electrode was determined (0.33 cm2). Hence, current and capacitance values reported in the paper are referred (normalized) to this electrochemically active surface area of the electrode. In PM-IRRAS experiments, the working electrode used as a NAD+ adsorption substrate was a gold-sputtered glass slide (EMF Corp., glass + 5 nm Ti + 100 nm Au). Before each experiment, the substrate was first thoroughly cleaned with water and ethanol in an ultrasonic bath and then electrochemically pretreated (cleaned) in 0.5 M perchloric acid by potentiodynamic cycling (40 cycles) between -0.3 and 1.5 V at a scan rate of 300 mV s-1.
3164 Langmuir, Vol. 23, No. 6, 2007 Instrumentation. Electrochemical techniques of cyclic and ac voltammetry (CV and ACV) and electrochemical impedance spectroscopy (EIS) were employed using an Autolab potentiostat/ galvanostat/frequency response analyzer (Ecochemie), PGSTAT30/ FRA2, controlled by the GPES/FRA v.4.9.5 software. In ACV, the ac voltage amplitude was 5 mV (peak to peak), frequency 25 Hz, and phase angle -90°. For the adsorption kinetic experiments, EIS measurements were done at a frequency of 25 Hz and ac voltage amplitude of 10 mV (peak to peak). Polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS) and attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy was done using a Bruker Tensor 27 spectrometer equipped with an external Bruker PM module, liquid nitrogen-cooled mercury-cadmium-telluride (MCT) detector, a zinc selenide photoelastic modulator, and an internal Pike Tech ATR module. The PM-IRRAS data were recorded at an incident angle of the infrared (IR) beam of 85° with the photoelastic modulator set at 2900 or 1400 cm-1, depending on the frequency region of interest. The PM-IRRAS spectra were taken by collecting 3000 scans with a spectral resolution of 3 cm-1. A ZnSe multibounce crystal was used for the ATR measurements. Experimental Methodology. All the measurements were carried out in an oxygen-free solution, which was achieved by continuous purging of the electrochemical cell with argon gas (99.998% pure). All the measurements, except the adsorption kinetics, were made in a quiescent solution, and the inert atmosphere was maintained by saturating the cell space above the electrolyte with argon. The NAD+ stock solution (in 0.1 M NaClO4) was prepared in a separate container. Before measurements in a NAD+-containing electrolyte, the background response of the electrode was recorded in 0.1 M NaClO4. Aliquots of NAD+ were then added to the electrochemical cell, and the electrochemical measurements were repeated for each aliquot. To characterize the adsorption behavior of NAD+ by ACV, the electrode was first held at a potential at which the adsorption is studied until reaching equilibrium. Then, the capacitance response was recorded by sweeping the potential in the positive direction, (50 mV around the adsorption potential. Adsorption kinetic measurements were carried out by recording capacitance values over time at 295 K using EIS. For the adsorption kinetic experiments, the argon bubbler was left in the solution during the measurement in order to enhance rapid distribution (mixing) of the NAD+ aliquot injected into the cell. PM-IRRAS measurements were done using a gold slide sample covered by an adsorbed NAD+ monolayer, formed potentiostatically at -0.5 or 0.4 V, in 1 mM NAD+ + 0.1 M NaClO4. After the electrochemical adsorption of NAD+, the sample was thoroughly rinsed with nanopure water and ethanol to remove any nonspecifically adsorbed NAD+, then dried in an argon stream, immediately followed by PM-IRRAS measurements. We found that the NAD+ desorption kinetics is extremely slow, and thus no desorption of the specifically adsorbed NAD+ occurred during the rinsing.
Damian and OmanoVic
Figure 1. (a) Linear polarization curves of a Au electrode recorded in (1) 0.1 M NaClO4 and (2) 0.1 M NaClO4 + 5 mM NAD+. Scan rate, sr ) 300 mV s-1; temperature, T ) 295 K. (b) Cyclic voltammogram of a Au electrode recorded in 0.1 M NaClO4. Scan rate, sr ) 200 mV s-1; temperature, T ) 295 K.
Linear polarization experiments were first done in order to determine the potential region of NAD+ reduction. Figure 1a shows the linear polarization curve recorded on a polycrystalline gold electrode in the absence (curve 1) and presence (curve 2) of NAD+ in the supporting electrolyte. A well-defined cathodic peak was recorded at ca. -1.18 V in the NAD+-containing electrolyte (curve 2). The absence of this peak in the NAD+-free solution (curve 1) demonstrates that the reduction of NAD+ occurs in the peak potential region. If the formal potential of a NAD+/ NADH couple at pH 5.8, E′ ) -0.485 V,33 is taken into account, this result also shows that the NAD+ reduction reaction is highly irreversible, i.e., it occurs at some appreciable rate (high currents) only at high cathodic overpotentials. Nevertheless, with the
corresponding pH difference taken into account, the NAD+ reduction peak potential recorded on gold (Figure 1) is in agreement with the value obtained on a cholesterol-modified and pure gold-amalgam electrode,7 mercury electrode,25 glassy carbon6,26 and ruthenium-modified glassy carbon (RuGC) electrode,5,6 and basal pyrolytic graphite electrode.34 Our recent measurements have also shown that the NAD+ reduction on a copper electrode occurs in the same potential region as on gold. All these data indicate that the NAD+ reduction overpotential is not electrode-material dependent, i.e., it does not depend on the electrode Fermi level position. Therefore, some other factor(s) must control the kinetics of electron transfer between the electrode and NAD+. This could be the conformation of the molecule on the electrode surface, which in turn influences the electron tunneling distance, and thus the electron-transfer kinetics. Since NAD+ is a large organic molecule, an outer-sphere reduction mechanism could be dismissed as a possible reaction mechanism. This, in turn, implies that NAD+ adsorbs on the gold surface before being reduced. Therefore, to investigate the thermodynamics and kinetics of NAD+ adsorption on a charged gold electrode surface, and to elucidate the surface conformation of the molecule, differential capacitance and infrared spectroscopy studies were done. The adsorption measurements were made over a wide potential and temperature range. 3.1. Thermodynamics of NAD+ Adsorption. In order to avoid the influence of faraday reactions, the adsorption of NAD+ on a gold electrode surface was investigated only in the electrochemical double-layer region (between ca. -0.8 and 0.5 V), where the electrode is considered to be ideally polarizable (Figure 1b). At more negative potentials, the reduction of NAD+ occurs (Figure 1a), while at more positive potentials, the gold surface gets oxidized during the anodic sweep (peaks A1/A2), followed by the reduction of the oxides during the cathodic sweep (peaks C2/C1). Therefore, it is important to note that, if the electrode is polarized in the electrochemical double-layer region, adsorption measurements are not influenced by any faraday reactions. This allows for a precise control of the electrode surface state (charge) and its influence on the adsorption and surface conformation of NAD+. The thermodynamics and kinetics of NAD+ adsorption on gold were investigated using differential capacitance measure-
(33) Clark, W.M. Oxidation and Reduction Potentials of Organic Systems; Williams & Wilkins: Baltimore, 1960.
(34) Nakamura, Y.; Suye, S.-I.; Kira, J.-I.; Tera, H.; Tabata, I.; Senda, M. Biochem. Biophys. Acta 1996, 1289, 221.
3. Results and Discussion
Adsorption BehaVior of NAD+ at Au Electrode
Langmuir, Vol. 23, No. 6, 2007 3165
θi )
Figure 2. Differential capacitance curves of a Au electrode recorded in 0.1 M NaClO4 containing various concentrations of NAD+: (1) 0, (2) 7.8, (3) 15.6, and (4) 31.3 µM. Scan rate, sr ) 2.5 mV s-1; frequency f ) 25 Hz; and ac amplitude ) (5 mV. Temperature, T ) 295 K. Anodic polarization direction.
ments. Figure 2 displays a set of differential capacity curves recorded on a polycrystalline Au electrode between -0.5 and 0.4 V in both the absence (curve 1) and presence (curves 2 to 4) of NAD+ in the supporting electrolyte. The capacitance curve recorded in the supporting electrolyte (curve 1) shows a capacitance minimum around 0 V, which corresponds to the potential of zero charge (pzc). This value is in very good agreement with the literature.35 It also signifies that negative of the pzc the electrode surface is negatively charged, while positive of the pzc the surface is positively charged. The differential capacitance values recorded in the supporting electrolyte (curve 1) are typical for the response of a double-layer capacitance,36 i.e., they range between ca. 20 and 35 µF cm-2. If NAD+ adsorbed on the gold surface, one could expect to see a decrease in differential capacitance due to both the electrode surface blockage caused by the adsorbed NAD+ molecules and a decrease in relative permittivity of the dielectric (water vs NAD+). Indeed, the graph in Figure 2 clearly shows that the addition of even a small amount of NAD+ in the supporting electrolyte (7.8 µM) resulted in a significant decrease in differential capacitance (curve 2; note the difference between the two capacitance scales on the ordinate). With an increase in NAD+ bulk solution concentration (curves 3 and 4), the measured differential capacitance further decreases. This is direct evidence of the adsorption of NAD+ on the Au electrode surface. Further, the graph also shows that the observed decrease in capacitance occurred in the entire electrochemical double-layer region investigated, indicating that NAD+ remained adsorbed on Au between -0.5 and 0.4 V. These observations are consistent with adsorption studies of NAD+ done on Hg in the same potential domain.1,20 3.1.1. Surface Charge Influence. To investigate the influence of surface charge on the adsorption of NAD+, differential capacitance measurements were made at several selected potentials (by polarizing the electrode (50 mV around the selected potential) on both sides of the pzc and at various NAD+ bulk solution concentrations. Then, using the measured capacitance values, the NAD+ surface coverage was estimated at the applied potential employing the following equation: 37 (35) Clavilier, J.; Nguyen Van Huong, C. J. Electroanal. Chem. 1977, 80, 101. (36) Trasatti, S.; Petrii, O. A. Pure Appl. Chem. 1991, 63, 711. (37) Damaskin, B. B.; Petrii, O. A.; Batrakov, V. V. Adsorption of Organic Compounds on Electrodes; Plenum Press: New York, 1971.
Codl - Cidl Codl - Cmin dl
(1)
where Codl (F cm-2) is the electrochemical double-layer capacitance in a NAD+-free solution, Cidl is the electrochemical double-layer capacitance at a specific equilibrium concentration of NAD+ in the bulk solution, and Cmin dl is the electrochemical double-layer capacitance at a maximum (saturated) NAD+ surface coverage. Figure 3a shows the dependence of the calculated NAD+ surface coverage on the equilibrium NAD+ bulk solution concentration at -0.5 V and 295 K. The obtained curve displays a shape characteristic of type I adsorption isotherm.38 The surface coverage rapidly increases with an increase in bulk NAD+ concentration and then levels off to a plateau at a bulk NAD+ concentration of ca. 0.4 mM, indicating that saturated (monolayer) NAD+ coverage is reached. Similar behavior was also obtained at other potentials, as illustrated by the 3D graph in Figure 4. The graph shows that the “curvature” of the adsorption isotherms slightly changes with potential, which indicates that the NAD+ adsorption is surface-charge dependent. In order to better quantify this dependency, i.e., to calculate thermodynamic adsorption parameters at each potential, the adsorption process was modeled by a suitable adsorption isotherm. Although the adsorption curves in Figures 3a and 4 display a typical Langmurian-type shape, several other isotherms were also tested. However, the Langmuir isotherm gave the best agreement between the experimental and modeled values38
θ ) Badsc 1-θ
(2)
where c (mol cm-3) is the equilibrium concentration of the adsorbate (NAD+) in the bulk solution, θ is the adsorbate surface coverage, and the parameter Bads (cm3 mol-1) reflects the affinity of the adsorbate molecules toward the adsorbent sites at a constant temperature and is termed adsorption affinity constant. Equation 2 can be further rearranged into a linear form to give
c 1 +c ) θ Bads
(3)
Hence, if a system behaves according to the Langmuir isotherm, a plot of c/θ versus concentration c should yield a straight line having a slope equal to 1. Then, the parameter Bads can be derived from the corresponding intercept. The NAD+ adsorption data recorded at -0.5 V and 295 K are presented in Figure 3b as an example, and indeed, the plotted c/θ vs c dependence is very linear, with a slope equal to 0.99 and a correlation coefficient R2 ) 0.9999. The same analysis method was applied to the data obtained at all the potentials and temperatures studied. An excellent agreement between the adsorption model and experimental data was obtained in each case, giving an overall mean correlation coefficient of R2 ) 0.9991 ( 0.0008. Such good agreement demonstrates the applicability of the Langmuir isotherm in describing the adsorption of NAD+ onto a gold surface in a wide range of potentials and temperatures. Figure 5 shows the corresponding adsorption affinity constant values, Bads, calculated at various adsorption potentials for measurements done at 295 K. With an increase in adsorption potential in the positive direction (the electrode surface charge becomes more positive), the affinity of NAD+ toward the gold surface also increases. The origin of the deviation of the value calculated at 0.4 V is currently (38) Levine, I. N.; Physical Chemistry, 5th ed.; McGraw Hill: New York, 2002; p 399.
3166 Langmuir, Vol. 23, No. 6, 2007
Damian and OmanoVic
Figure 3. (a) Dependence of the NAD+ surface coverage on the equilibrium concentration of NAD+ in 0.1 M NaClO4. The data were obtained from differential capacitance measurements recorded at -0.5 V. Temperature, T ) 295 K. (b) Linearized Langmuir adsorption isotherm for the adsorption of NAD+ on Au obtained using the data presented in graph (a). The symbols represent the experimental data, and the line is the modeled isotherm.
Figure 5. (a) Dependence of the NAD+ adsorption affinity constant on the adsorption potential, calculated by fitting experimental adsorption data recorded at 295 K. (b) Dependence of the Gibbs energy of adsorption of NAD+ on the applied temperature. The values were calculated from the adsorption measurements done at (O) -0.5 and (∆) 0.4 V. Table 1. Entropy and Enthalpy of Adsorption of NAD+ on a Negatively Charge (-0.5 V) and Positively Charge (0.4 V) Gold Electrode Surface adsorption potential thermodynamic variable -1
∆Hads/kJmol ∆Sads/Jmol-1K-1 T∆Sads,average/Jmol-1
Figure 4. Dependence of the NAD+ surface coverage on applied electrode potential and equilibrium concentration of NAD+ in 0.1 M NaClO4. Temperature, T ) 295 K.
not known. The trend in Figure 5a indicates that electrostatic forces between the molecule and surface play a significant role in NAD+ adsorption on gold, which will be explained in more detail later in the paper. 3.1.2. Temperature Influence. To characterize the thermodynamics of NAD+ adsorption on a gold electrode surface using state variables (Gibbs energy, enthalpy and entropy of adsorption), adsorption experiments were made over a wide temperature range at the two potentials on the opposite sides of the pzc, -0.5 and 0.4 V. Then, using eq 3, the corresponding adsorption affinity constant, Bads, was determined, and the Gibbs energy of adsorption was subsequently calculated by employing the following equation:39
Bads )
(
)
-∆Gads 1 exp csolvent RT
(4)
where R (J mol-1 K-1) is the gas constant, T (K) the temperature, ∆Gads (J mol-1) the Gibbs energy of adsorption, and csolvent the molar concentration of the solvent, which is for this case water (cH2O ) 55.5 mol dm-3). Figure 5b shows the Gibbs energy of adsorption values calculated in the whole temperature range and at the two potentials investigated. The data on the graph shows that the Gibbs energy of NAD+ adsorption is highly temperature (39) Gomma, G. K.; Wahdan, M. H. Mater. Chem. Phys. 1994, 39, 142.
-0.5 V
0.4 V
1.2 127 40 ( 2
46.1 285 90 ( 4
and potential dependent, i.e., it increases with temperature (in the negative direction). It is also clear that at a constant temperature the NAD+ displays higher affinity toward the positively charged surface. Such high negative Gibbs energy values indicate that the equilibrium for the adsorption process lies well in favor of adsorption of NAD+ on the Au surface. A close value was reported by Takamura et al.,3 -31 kJ mol-1, for the adsorption of nicotinamide (NA) on Au at 295 K. To the best of our knowledge, no thermodynamic data on the adsorption of NAD+ on gold is available in the literature, which does not allow us to compare our results to results obtained in other laboratories. To more completely characterize the thermodynamics of the NAD+ adsorption process and to determine the corresponding adsorption driving force, the enthalpy and entropy of adsorption were also determined. The relation between Gibbs energy, enthalpy, and entropy of adsorption can be presented by a wellknown thermodynamic relation, ∆Gads ) ∆Hads - T∆Sads. Hence, from the intercept of the lines and the slope in Figure 5b, the values of enthalpy ∆Hads (J mol-1) and entropy ∆Sads (J mol-1 K-1) of NAD+ adsorption on gold were calculated at the two electrode potentials, and presented in Table 1. The enthalpy values demonstrate that the adsorption of NAD+ on the negatively charged surface (-0.5 V) is slightly endothermic, while its adsorption on the positively charged surface (0.4 V) is a highly enthalpically unfavorable process. Nevertheless, the high negative Gibbs energy values (Figure 5b) demonstrate that the overall adsorption process is highly spontaneous. Therefore, the major contribution to this spontaneity has to come from a positiVe gain in entropy. Indeed, the average value of the T∆Sads product
Adsorption BehaVior of NAD+ at Au Electrode
(Table 1) is 40 ( 2 kJ mol-1 on the negatively charged surface and 90 ( 4 kJ mol-1 on the positively charged surface. Therefore, the large positive gain in entropy seems to be the main contributor to the driving force for the adsorption of NAD+ on gold. Our opinion is that the major contribution to the entropy gain in the system comes from the loss of the order of water molecules adsorbed on the Au surface upon NAD+ adsorption. Under the influence of the electric field, the adsorbed water molecules are highly ordered on the electrode surface,40 but when displaced by the considerably larger NAD+ molecules, this high degree of order significantly decreases due to the random orientation of the displaced molecules in the bulk solution. This results in an overall increase in the positional degree of freedom of the system, i.e., an entropy gain. On the other hand, adsorbed NAD+ molecules lose positional degrees of freedom, but this effect is largely overcome by the positive entropy contribution coming from water. Some other entropically governed processes investigated in our laboratory include the adsorption of caffeine on a positively charged platinum surface41 and also the adsorption of several proteins, such as yeast alcohol dehydrogenase, on a Pt surface,42 β-lactoglobulin on stainless steel,43 and bovine serum albumin on Ti.44 A small portion of the observed increase in entropy could also come from NAD+ conformational changes occurring upon the adsorption of the molecule on the gold surface. 3.2. Kinetics of NAD+ Adsorption. The previous section of the paper demonstrated that NAD+ spontaneously adsorbs on the gold electrode surface. Hence, the NAD+ adsorption indeed represents one of the reaction steps in the overall NAD+ electroreduction mechanism, which was already indicated earlier in the paper. To complement the adsorption thermodynamic studies, the kinetics of NAD+ adsorption was also investigated by measuring the variation of differential capacitance with time at a fixed frequency and electrode potential, employing EIS. The influence of surface charge and mass-transport on the adsorption kinetics was investigated. 3.2.1. Surface Charge Influence. The NAD+ adsorption kinetics was studied at two different electrode potentials on each side of the pzc, where the electrode surface is either negatively (-0.5 V) or positively (0.4 V) charged. Figure 6a shows the change in surface coverage with time for the electrode polarized at two different potentials. There is a significant difference in the kinetics of NAD+ adsorption at the two potentials, which shows that the NAD+ adsorption kinetics is highly surface-charge dependent, the same as the NAD+ adsorption thermodynamics (Figure 5b). On the positively charged surface (0.4 V, solid line), the surface coverage increases sharply upon NAD+ addition (
