Electrochemical Evaluation of 4-(Dimethylamino) pyridine Adsorption

Annie Dorris, Simona Rucareanu, Linda Reven, Christopher J. Barrett, and R. Bruce Lennox. Langmuir 2008 24 (6), 2532-2538. Abstract | Full Text HTML |...
1 downloads 0 Views 211KB Size
Download PDF
Langmuir 2007, 23, 1555-1563

1555

Electrochemical Evaluation of 4-(Dimethylamino)pyridine Adsorption on Polycrystalline Gold Burke C. Barlow and Ian J. Burgess* Department of Chemistry, UniVersity of Saskatchewan, Saskatoon, Saskatchewan, S7N 5C9 Canada ReceiVed September 13, 2006. In Final Form: NoVember 3, 2006 Using differential capacity and chronocoulometry, we have studied the electrosorption of 4-(dimethylamino)pyridine (DMAP) on polycrystalline gold electrode surfaces. Our results indicate that the orientation of DMAP is highly dependent on the electrode potential and electrolyte pH. At pH values at or above the primary pKa, the adsorbed species is DMAP and orients vertically on the electrode surface via the lone pair of electrons on the pyridine ring’s nitrogen atom. At very low pH values (18.2 MΩ cm-1) water. 4-(dimethylamino)pyridinium perchlorate ((DMAP)HClO4) was obtained by precipitating a saturated, aqueous solution of DMAP with the addition of HClO4. The large, needle-like (DMAP)HClO4 crystals were collected on a fine frit and first washed with copious amounts of acidic (0.1 M HClO4) water and finally with ice cold neutral water. Caution: although we are unsure of the reactiVity of (DMAP)HClO4, organic perchlorate salts in general are often potentially explosiVe and should be treated with great care! All salts were dried and stored in a vacuum desiccator. All aqueous solutions were made using Milli-Q water. The gold polycrystal electrode was prepared as follows. Approximately 2 g of 99.99% gold (Alfa Aesar) was cut into small pieces and briefly etched in aqua regia (3:1 HCl/HNO3) to remove surface impurities. The gold was then placed in a graphite mold and heated well above its melting temperature with use of an inductive furnace. After the melting stage, the crucible was allowed to rapidly cool and the formed crystal was removed from its graphite mold in the shape of a bead attached to a rodlike stem. Etching in aqua regia revealed obvious fault lines indicative of polycrystallinity. Next, the crystal was set in a polypropylene holder with epoxy, cut, and polished using successive grades of sandpaper. A mirror finish on the electrode’s surface was achieved by serial polishing the crystal with decreasing (6, 3, 1, and 0.25 µm) sized diamond pastes (cloths and diamond sprays obtained from LECO). The electrode was removed from the epoxy by soaking in chloroform overnight and then electropolished in 0.1 M HClO4. This was achieved by repeated (30) Andreasen, G.; Vela, M. E.; Salvarezza, R. C.; Arvia, A. J. Langmuir 1997, 13, 6814. (31) Kunze, J.; Burgess, I.; Nichols, R.; Buess-Herman, C.; Lipkowski, J. J. Electroanal. Chem., in press.

DMAP Adsorption on Polycrystalline Gold oxidization of the electrode with a current density of ∼1 mA cm-2 and subsequent dissolution of the gold oxide in 10% HCl. Further electropolishing was achieved by 3 h of cycling the potential applied to the electrode in 50 mM KClO4 (-0.75 V < E (vs Ag/AgCl) < 1.2 V). Finally, the electrode was repeatedly flame annealed in a natural gas flame until a CV characteristic of polycrystalline gold was obtained.32 Electrochemical Measurements and Instrumentation. Prior to each electrochemical experiment the working electrode was flameannealed, cooled first in air and then in water, dried in the flame, and directly transferred to the electrochemical cell to cool in an argon atmosphere. A flame-annealed gold coil was used as the counter electrode, while the reference electrode was an external, saturated Ag/AgCl electrode (+197 mV vs SHE). All glassware used for the electrochemical measurements was cleaned in hot acid (1:3 mixture of HNO3 and H2SO4) and thoroughly rinsed with Milli-Q water. The electrochemical cell was soaked in Milli-Q water overnight and rinsed again prior to the experiment. All electrochemical measurements were carried out in an all-glass cell using the working electrode in the hanging meniscus configuration.33 Electrolyte solutions were 50 mM KClO4 and were adjusted to the desired pH value with either KOH or HClO4 solutions prior to electrochemical characterization. The electrolyte solution was deaerated by purging with Argon (Praxair, Saskatoon, SK, Canada) for at least 30 min prior to the experiments, and an argon blanket was maintained over the solution throughout the experiment. Measurements were carried out at room temperature (20 ( 2 °C). A computer-controlled system, consisting of a HEKA potentiostat PG590 (HEKA, Mahone Bay, NS, Canada) and an SRS 830 lock-in amplifier (Stanford Research Systems), was used for the differential capacity and chronocoulometry experiments. Data were collected using a multifunction DAQ card (PCI 6251 M Series, National Instruments, Austin, TX) and in-house software written in the LabVIEW environment. To record the differential capacity curves, a 25 Hz, 5 mV (rms), ac perturbation was superimposed on a 5 mV/s dc voltage sweep. In the chronocoulometry method, the Au electrode was held at a base potential, Ebase, for 30 s. Then the potential was stepped to a variable value of interest, Ec, and held for a sufficient time to achieve adsorption equilibrium (between 180 and 30 s depending on the DMAP(H+) concentration). To desorb DMAP(H+) species from the electrode surface, the potential was stepped to a sufficiently negative potential, Edes, for 200 ms and the current transient corresponding to the desorption was measured. The potential was then stepped back to the base value, and the cycle was repeated using a new value of Ec. The values of Ebase and Edes were -600 and -950 mV for pH 9.7 and -500 and -900 mV for pH 4.5. To enhance the mass transport of DMAP(H+) from the solution to the electrode surface, the electrolyte solution was stirred when the electrode potential was held at Ebase and Ec. To ensure a quiescent solution, the stirring was interrupted 10 s before the potential was stepped to Edes and the current transient was measured. Integration of the current transients gives the difference between charge densities, ∆σM, at potentials Ec and Edes. The absolute values of the charge density were then determined using the potential of zero charge (pzc) of the polycrystalline electrode and the fact that the charge density in solutions containing adsorbing species is equal to the charge density for the pure supporting electrolyte at the potential of total desorption. In a separate experiment, the pzc was determined to be -75 mV vs Ag/AgCl from the diffuse layer capacitance minimum observed in 5 mM KClO4. Data processing such as integration and differentiation was performed with Origin 7.5 (Microcal, Northampton, MA) using custom-written macros. For the experiments performed at pH 9.7, an equimolar solution of DMAP and DMAPH+ was used as the spiking solution to maintain a constant electrolyte pH. In the case of experiments performed at pH e 4.5 the spiking solution only contained (DMAP)HClO4. (32) Clavilier, J. J. Electroanal. Chem. 1977, 80, 101. (33) Dickertmann, D.; Schultze, J. W.; Koppitz, F. D. Electrochim. Acta 1976, 21, 967.

Langmuir, Vol. 23, No. 3, 2007 1557

Figure 2. Differential capacity curves for polycrystalline gold in 50 mM KClO4 supporting electrolyte (‚‚‚, red) and in the presence of 0.1 mM formal concentration DMAP (s) as a function of pH: (a) pH 11; (b) pH 4.5; (c) pH 2.0. The pH of the supporting electrolyte was adjusted to the desired value with the addition of either dilute KOH or HClO4.

Results and Discussion Differential Capacity. As an initial survey of the adsorption of DMAP, we performed differential capacity (DC) measurements as a function of the electrolyte pH. Figure 2 shows DC curves at three different pH values for solutions of constant formal DMAP(H+) concentration (0.1 mM). The formal concentration is the sum of [DMAP] and [DMAPH+]. The window of potentials corresponding to the double-layer region is pH dependent, and consequently, the DC curves in Figure 2 shift to more positive potentials with decreased pH. At pH 11, and the most negative of potentials (Figure 2a), the capacity curve for the solution containing DMAP merges with the electrolyte curve, indicating that DMAP is completely desorbed from the electrode’s surface. As the potential is scanned in the positive direction, a somewhat broad, but featureless, pseudocapacitive peak (denoted R) centered at E ≈ -0.7 V is observed. Following this peak, the capacity drops to a limiting value of ∼7.5 µF cm-2 and remains

1558 Langmuir, Vol. 23, No. 3, 2007

Barlow and Burgess

significantly lower than the capacity for the bare gold surface between -0.4 V < E < 0.3 V, indicating a region of DMAP adsorption that we term state II. For completeness we note that the capacity does not remain constant in this region but rather slowly increases as the potential is moved increasingly positive. This may be due to either partial desorption of the organic film or competitive adsorption of hydroxide ions or a combination of both. In slightly less basic solutions (see the DC curves for pH 9.7 shown below) the capacity remains constant in this potential range, indicating that competitive adsorption is the most likely explanation. The DC curve for an intermediate electrolyte acidity (pH 4.5) is shown in Figure 2b. Again, at the most negative potentials the curves for the surfactant-containing and surfactantfree solutions essentially merge at the most negative potentials. However, the positive-going trace is appreciably different compared to the equivalent curve shown in Figure 2a. The R peak which is attributed to DMAP(H+) adsorption is slightly shifted to E ≈ -0.6 V but is the same magnitude as in Figure 2a. In fact, the R peak is largely unaffected by pH and occurs at roughly the same potential at constant DMAP(H+) concentration for all accessible electrolyte acidities studied. After passing through the R peak, the capacity remains quite high (albeit significantly lower than the capacity for the supporting electrolyte) at ∼21 µF cm-2. This implies that a lower coverage film (termed state I) is adsorbed at these potentials and at this pH. As the potential is scanned further positive, a second, larger, pseudocapacitive peak (denoted β) appears at E ≈ 0 V, and the capacity then drops rapidly into a capacitive pit region (0.2 V < E < 0.7 V) having a minimum value of ∼10 µF cm-2. The similarities of the low capacity regions at positive potentials in Figure 2a,b indicate a similar state (state II) of DMAP(H+) adsorption is occurring at both pH values at these positive electrode polarizations. Peak β can be observed in electrolytes where ∼3 < pH < ∼9 and shifts in the anodic direction with increasing acidity by ∼60 mV/decade. We therefore interpret the β peak as corresponding to a change in the adsorbed species from DMAPH+ (state I) to DMAP (state II). Below we provide further evidence of this assessment based on our chronocoluometry studies. Finally, Figure 2c presents DC curves obtained at pH 2.0. Due to hydrogen evolution, it was impossible to scan the electrode sufficiently negative to observe the R peak. The capacity at ∼-0.4 V is very close to the capacity observed at similar potentials at pH 4.5, indicating the occurrence of adsorption state I. A broad pseudocapacity peak (centered at ∼0.1 V) is again observed as the potential is scanned positive, but significantly, the capacity for the solution in the presence of 0.1 mM DMAPH+ merges with the electrolyte-only curve at potentials greater than 0.2 V. The similarity of this peak to the peak observed for the electrolyte in the absence of DMAPH+ precludes assigning this feature to the process labeled β above. Furthermore, the fact that the DC curves merge indicates that, in contrast to higher pH values, the surfactant is totally desorbed at positive potentials and state II of adsorption does not exist in very acidic electrolytes. If one neglects the contribution of the diffuse part of the double layer, the experimentally measured capacity can be used to determine the extent of organic molecule adsorption as the inner layer, or Helmholtz, capacity, CH, is given as

CH ) /x

(1)

In eq 1,  is the permitivitty of the adsorbed organic layer and x is the thickness of the adsorbed organic molecules. One can approximate the fractional coverage, θ, using Damaskin’s theory

Figure 3. Fractional coverage of state II, as determined using eq 2, of DMAP(H+) as a function of the electrolyte pH.

of parallel capacitors34

C ) Corg(θ) + C0(1 - θ)

(2)

where C is the measured capacity at a given pH, Corg is the capacity of a complete monolayer, and C0 is the capacity of the bare electrode. The value of θ for potentials corresponding to state II can be approximated as a function of pH from eq 2, using the assumption that the value of Corg ) 7.0 µF cm-2 (i.e., the lowest capacity measured at the highest pH). Figure 3 plots θ versus pH and clearly displays a sigmoidal dependence reminiscent of a titration curve. This implies that desorption of DMAP molecules from the gold surface is driven by the protonation of the endocyclic nitrogen. It should be noted that the midpoint of the curve occurs at pH ≈ 4, which is nearly 6 units shifted from the pKa of DMAPH+’s endocyclic nitrogen (pKa ) 9.7). However, the acid equilibrium constants of many acidic and basic compounds are known to shift by comparable amounts when the molecules are assembled into monolayers on surfaces,35 and this may explain the shift observed in Figure 3. We emphasize that the θ dependence on pH described in Figure 3 should be considered semiquantitative due to the assumptions described above. Nevertheless, the results are very consistent with reports of decreased stability of DMAP-AuNPs at low pH values.16,20 Below we show that a rigorous thermodynamic analysis of DMAP(H+) adsorption derived from chronocoloumetric data strongly supports the pH-dependent coverage obtained from the DC data. To further characterize the potential-dependent and concentration-dependent adsorption behavior of DMAP(H+) on gold, differential capacity measurements were performed as a function of DMAP(H+) concentration at two different pH values. Figure 4a shows the capacity curves obtained for polycrystalline gold in the absence (dotted lines) and in the presence (solid lines) of DMAP/DMAPH+. In Figure 4a, the base electrolyte’s pH was adjusted to 9.7 and the reported concentrations are formal concentrations. The curve for 0.02 mM DMAP/DMAPH+ is similar to the curve shown in Figure 2a. At potentials more negative of the pseudocapacitive peak, and at low DMAP/ DMAPH+ concentrations, the capacity merges with that of the electrolyte-only curve, indicating complete desorption of DMAP(H+). As the potential is swept in the positive direction, the onset of adsorption is clearly indicated by the broad pseudocapacitive peak R followed by a plateau in the measured capacity in the (34) Damaskin, B. B.; Petrii, O. A.; Batrakov, V. V. Adsorption of Organic Compounds at Electrodes; Plenum Press: New York, 1971. (35) Burgess, I.; Seivewright, B.; Lennox, R. B. L. Langmuir 2006, 22, 4420 and refs 21-37 therein.

DMAP Adsorption on Polycrystalline Gold

Figure 4. Differential capacity curves for polycrystalline gold in 50 mM KClO4 supporting electrolyte (‚‚‚, red), 0.02 mM formal concentration DMAP (---), and 0.2 mM formal concentration DMAP (s): (a) pH 9.7; (b) pH 4.5.

potential region ∼-0.3 V < E < 0.4 V. In this potential window the capacity is between 7.5 and 8.5 µF cm-2 and remains largely invariant with potential. The return voltage sweep shows considerable hysteresis, which is expected for such low concentrations of surfactant. At increased concentration (formal [DMAP] ) 0.2 mM) peak R becomes sharper and shifts in the negative direction. At even higher DMAP/DMAPH+ concentrations the R peak continues to shift to more negative potentials until at ∼1 mM it overlaps with the onset of hydrogen evolution. The capacity in the positive potential plateau also decreases somewhat to a limiting value of 7.0 µF cm-2. Additionally, the broad features observed at low surfactant concentration are replaced by a noticeable shoulder on the anodic side of the R peak. This shoulder evolves into the β peak and shifts increasingly to positive potentials with decreasing pH. The DC curves shown in Figure 4b were performed in 50 mM KClO4, the pH of which was adjusted to 4.5 with dilute HClO4. At this pH, the surfactant exists almost exclusively in the bulk of the solution as the 4-(dimethylamino)pyridinium ion. Similar to Figure 4a, the capacity of the interface in pure electrolyte and in the presence of surfactant essentially merges when the potential is negative of E ) -0.8 V. At these lower pH values, the desorption potential overlaps with hydrogen evolution even at low bulk surfactant concentration, but due to the phase-sensitive method used to measure the capacity, the curves are largely unaffected by the contribution of the Faradaic signal. As the potential is scanned in the positive direction for [DMAPH+] ) 0.02 mM, the R peak is observed, which corresponds to the onset of the initial state I of adsorption. At potentials close to E ) 0 V the β pseudocapacitance peak is observed, following which the capacity decreases further to a limiting value of ∼11 µF cm-2 (state II). An appreciable hysteresis is again observed due to slow kinetics of adsorption and reorganization at these low

Langmuir, Vol. 23, No. 3, 2007 1559

surfactant concentrations. At higher concentrations, i.e., [DMAPH+] ) 0.2 mM, the DC curve exhibits the same features, but both R and β peaks are shifted negatively, and the capacities in the regions corresponding to states I and II of adsorption are lowered to ∼20 and 8.5 µF cm-2, respectively. At this point it is insightful to compare the differential capacity reported herein to that of the well-studied system of pyridine adsorption on gold. On polycrystalline gold36 and Au(111)37 Lipkowski and co-workers reported capacitive behavior very similar to our results shown in Figures 2b and 4b. After adsorption from slightly basic electrolytes, pyridine initially lowers the capacitance of the interface to ∼15 µF cm-2. As the potential is scanned positive, a second pseudocapacity peak is observed near the potential of zero charge. Lipkowski and co-workers attributed this peak to the phase transition from horizontally adsorbed pyridine at negatively charged gold surfaces to a vertical orientation when the electrode surface is positively charged. The similarity with the current system strongly suggests a horizontal to vertical transition for the DMAP(H+) system at intermediate electrolyte pH. However, it is interesting to note that in alkaline solutions (pH g pKa) there is no evidence of state I of adsorption and in very acidic electrolytes (pH , pKa) state II does not appear to exist. On the basis of only the capacity measurements, it is impossible to determine the speciation of the adsorbed species in states I and II. It is also not possible to tell if the speciation in a given state remains invariant with pH (e.g., does the vertical orientation correspond to protonated or deprotonated DMAP at all pH values?). Clearly, the electrosorption of DMAP(H+) is significantly dependent on the electrolyte pH, but a more quantitative methodology is required to provide these important details. Chronocoulometry. We performed two series of chronocoulometry experiments at different DMAP/DMAPH+ and DMAPH+ concentrations, one at pH 9.7 and the other at pH 4.5. In these experiments, the relative charge density (∆σm) on the metal at any given potential, Ec, was determined from the current transient resulting from stepping the potential from Ec to the desorption potential. On the basis of the differential capacity measurements, all forms of DMAP desorb at E e -0.9 V when its bulk concentration is less than 1 mM. Consequently, the absolute charge density, σm, can be evaluated knowing the potential of zero charge, Epzc, for the surfactant-free electrolyte. Epzc was determined to be -0.075 V (vs Ag/AgCl) as determined from the potential of the diffuse layer capacity minimum observed in the DC of the electrode in 5 mM KClO4. The results of the chronocoulometry experiments performed at pH 9.7 are shown in Figure 5a. The curves can be separated into three sections. Region 1a corresponds to the most negative electrode polarizations where all the curves merge with the electrolyte curve, indicating the absence of DMAP(H+) adsorption at these potentials. Region 2a (∼-0.6 V < E < ∼0 V) is characterized by its quasi-linear appearance where the charge density varies weakly with potential and is consistent with state II of adsorption described in the Differential Capacity section. The curves are characterized by a pronounced inflection point at approximately -0.6 V (its exact position shifts with bulk surfactant concentration) that separates regions 1a and 2a. The potentials of the inflection points correspond well with the adsorption/desorption peaks observed in the differential capacity measurements. All the σm-E curves in the presence of DMAP/ DMAPH+ intersect the curve for the supporting electrolyte close (36) Stolberg, L.; Richer, J.; Lipkowski, J.; Irish, D. E. J. Electroanal. Chem. 1986, 207, 213. (37) Stolberg, L.; Morin, S.; Lipkowski, J.; Irish, D. E. J. Electroanal. Chem. 1991, 307, 241.

1560 Langmuir, Vol. 23, No. 3, 2007

Barlow and Burgess

resents the potential of maximum adsorption, Emax. In Figure 5a, a single point of intersection occurs at 0.0 V and is independent of the surfactant concentration. This value of Emax is close to the value of Epzc for our bare electrode. The nearly numeric equivalence of Emax and Epzc is expected for the one-state adsorption of neutral, organic molecules.34 In Figure 5b, there are two, concentration-independent, intersection potentials, the first occurring at Emax,I ) -0.225 V and the second at Emax,II ) 0.4 V. The appearance of two potentials of maximum adsorption indicates that the two states of adsorption are clearly delineated. For comparison, Stolberg et al.36 reported that pyridine did not have a potential of maximum adsorption on polycrystalline gold but rather exhibited a slow transition from the horizontal state of electrosorption to the vertical orientation with increasing positive potential. It is also apparent that the adsorption of DMAP species shifts the potential of zero charge. This shift can be expressed as

EN ) (Epzc)surfactant - (Epzc)0

Figure 5. Charge density versus electrode potential curves for polycrystalline gold in 50 mM KClO4 supporting electrolyte (s, red) and the following formal DMAP concentrations: (a) pH 9.7, (9) 0.01 mM, (O) 0.02 mM, (b) 0.04 mM, (4) 0.07 mM, (2) 0.14 mM, (3) 0.27 mM, (1) 0.75 mM; (b) pH 4.5, (9) 0.005 mM, (O) 0.010 mM, (b) 0.020 mM, (4) 0.05 mM, (2) 0.10 mM, (3) 0.20 mM, (1) 0.50 mM, (sideways triangle) 1.0 mM (for clarity not all concentrations studied are shown in the figure).

to the Epzc, indicating that there is a potential of maximum adsorption, Emax. The slopes of the charge density curves slightly increase at potentials positive of Emax, and this delimits region 2a from region 3a. The charge density curves for pH 4.5 are presented in Figure 5b. Although these curves can also be divided into three discrete potential regions, they exhibit significant differences compared to those in Figure 5a. In region 1b, the charge density curves all merge with the electrolyte’s curve, indicating total desorption, and an inflection in the σm-E plots is also observed at ∼-0.6 V, indicating the onset of state I of adsorption. However, in region 2b, -0.6 V < E < -0.2 V, the charge densities obtained for chronocoulometry experiments performed in the presence of DMAPH+ seem significantly closer to the electrolyte-only data as compared to those of the equivalent experiments at pH 9.7. The dependence of the metal’s charge on surfactant concentration is very weak in this region. The most striking difference between parts a and b of Figure 5 occurs near zero charge density (E ≈ 0 V) where a second inflection is evident at pH 4.5 but is not observed at all in the more basic electrolyte system. This second inflection is far more pronounced than the first inflection, exhibits a strong dependence on the bulk concentration of DMAPH+, indicates the onset of the second state of adsorption (state II), and is qualitatively similar to the single inflection observed at negative potentials in Figure 5a. As discussed in the Differential Capacity section, we interpret this second inflection as a phase transition in the adsorbed DMAP(H+) film. The potential where the σm-E curves for the pure electrolyte solution and the surfactant-containing solutions intersect rep-

(3)

where (Epzc)surfactant and (Epzc)0 represent the potentials of zero charge in the presence and absence of DMAP species, respectively. At low concentrations, the value of (Epzc)surfactant for the gold surface in the presence of DMAP(H+) adsorption can be read directly from the charge plots in Figure 5a. At higher bulk surfactant concentrations the potential of zero charge appears in the inflection of the curve and (Epzc)surfactant needs to be determined by extrapolating the linear portions of the σm-E curves (region 2a) to the zero charge intercept. Doing this for the highest surfactant concentration studied yields a value of (Epzc)surfactant ) -0.800 V and a corresponding shift of the pzc equal to -0.725 V. An equivalent analysis can be performed for the data shown in Figure 5b. It should be noted that, at this lower pH, the charge curves need to be extrapolated from region 2b for all surfactant concentrations studied. Doing so for [DMAPH+] ) 10-3 M gives values of (Epzc,I)surfactant and EN,I of 0.00 and +0.075 V, respectively, where the subscript I denotes state I of adsorption. In region 3b (corresponding to state II of adsorption) extrapolation of the σm-E curve back to zero charge density yields (Epzc,II)surfactant and EN,II of -0.725 and -0.650 V. Below, we use a simple electrostatic model to equate these shifts in the potential of zero charge to the dipole moment of the adsorbed species from which we can further infer the identity of the adsorbed species. Surface Pressures. Thus far, our description of DMAP(H+) adsorption has been based on a somewhat qualitative analysis. The chronocoulometry results provide data that can be used for more quantitative information. From the electrocapillary equation, integration of the σm-E curves at any given concentration of DMAP species (FDMAP) yields the surface pressure

π)(

∫EE σm dE)F c

des

DMAP

∫EE σm dE)F

-(

c

des

DMAP)0

(4)

Figure S1 (Supporting Information) shows the calculated film pressures for the two pH systems studied. In pH 9.7, the curves exhibit a bell shape as expected for a single state of adsorption with one potential of maximum adsorption. The film pressure curves reach a maximum at 0.025 V, which corresponds to the potential of maximum adsorption, Emax. The value of Emax remains invariant with increasing formal concentration of DMAP. The values of the film pressure offer insight into the energetics of DMAP(H+) adsorption as large values of π imply large absolute values for the Gibbs energy of adsorption. The maximum measured film pressure of ∼125 mN m-1 for FDMAP ≈ 0.75 mM

DMAP Adsorption on Polycrystalline Gold

Langmuir, Vol. 23, No. 3, 2007 1561

is quite high and indicates that the adsorbing species is strongly bound to the gold surface. In the case of pH 4.5, the surface pressure curves appear as the superposition of two parabolas, each one corresponding to a different state of adsorption. The parabolas have maximal surface pressures at potentials corresponding to Emax,I (-0.225 V) and Emax,II (-0.400 V) as predicted from Figure 5b and eq 4. The first parabola reaches a maximum surface pressure of 13 mN m-1 at Emax,I, and the second parabola reaches a maximum value of 52 mN m-1 at Emax,II. The surface pressures are appreciably smaller in magnitude for the more acidic series of measurements. For comparable surfactant concentration in the electrolyte solution, the surface pressure is less than half at the lower pH. Below we use our surface pressure data and the Henry law isotherm to provide numeric values of ∆Gads° for the two systems studied. Both citrate- and DMAP-stabilized particles are appealing because their ligands can be exchanged, postsynthesis, with a more strongly adsorbing ligand. When comparing different adsorbates, one can potentially use both surface pressure and ∆Gads° values as a yardstick of adsorption energies. In principle, the latter should be a better choice as it describes a defined system in terms of the variables that affect the adsorption. However, ∆Gads° values determined for any given system depend on the choice of standard state and a judicious selection of isotherm, which are often not consistent from one adsorbate to another. On the other hand, surface pressures can be calculated without assumption of the standard state or choice of isotherm. Values as high as π ) 240 mN m-1 have been recently reported for citrate ion adsorption on gold.31 This value is nearly 5 times larger than surface pressures measured herein for similar solution concentrations of adsorbate. ∆Gads° values were also reported for this system, but because they were determined using a different isotherm, we prefer to make a somewhat qualitative comparison of the adsorption energies based upon the surface pressure data and the assumption that the adsorption is driven by the decrease in the interfacial tension. Our results indicate that DMAP may be more preferable for post synthesis ligand exchange reactions as its energy of adsorption is apparently significantly lower than that of the citrate species.38 Gibbs Surface Excesses. The Gibbs surface excess, Γ, can be calculated from the film pressure data by differentiation with respect to the natural log of the formal DMAP concentration at constant electrode potential:

Γ)

(

∂π 1 RT ∂ ln FDMAP

)

(5)

T,P,E

For the pH 9.7 system, we found that, at potentials less than ∼-0.3 V, Γ varied appreciably with the bulk DMAP species concentration, but at higher potentials this dependence was much weaker. Consequently, for potentials less than -0.3 V plots of π vs ln FDMAP were fit to a third-order polynomial, whereas at higher potentials a linear fit was used to avoid aberrant data scatter. The Gibbs excess was then calculated by differentiation of the fitted curves, the results of which are shown in Figure 6a. As expected from the differential capacity data, the Γ vs E plots exhibit a pronounced increase at potentials corresponding to the adsorption/desorption peaks and shift cathodically with increasing DMAP/DMAPH+ concentration. At values in the range -0.3 V < E < 0 V the amount of adsorbed species plateaus at a value of 7.4 × 10-10 mol cm-2. Consistent with the potential of maximum adsorption, as the potential increases above 0 V, the (38) Rucareanu, S.; Gandubert, V. J.; Lennox, R. B. Chem. Mater. 2006, 18, 4674.

Figure 6. Gibbs surface excesses as a function of electrode potential for the following formal DMAP concentrations: (a) pH 9.7, (9) 0.01 mM, (0) 0.02 mM, (b) 0.04 mM, (O) 0.07 mM, (2) 0.14 mM, (4) 0.27 mM, (1) 0.75 mM, (3) average result for all concentrations; (b) pH 4.5, (9) 0.020 mM, (0) 0.05 mM, (b) 0.075 mM, (O) 0.10 mM, (2) 0.20 mM, (4) 0.50 mM, (1) 0.75 mM, (3) 1.0 mM.

Gibbs excess slowly starts to decrease from its maximum value. At E ) 0.4 V, Γ has dropped by about 20% to 5.9 × 10-10 mol cm-2. Within the 10% error associated with the back-integration method,25 the former value is in excellent agreement with the maximum Gibbs surface excesses reported by Stolberg et al. for the vertical orientation of pyridine on polycrystalline gold (7.0 × 10-10 mol cm-2) and Au(111) (6.7 × 10-10 mol cm-2) and the predicted packing density of pyridine based on van der waal radii (7.6 × 10-10 mol cm-2). In the case of adsorption from the pH 4.5 electrolyte (Figure 6b), the surface coverage of DMAP(H+) depends on the bulk surfactant concentration for all potentials studied. The resulting isotherm reveals two limiting coverages, the first appearing at potentials corresponding to state I of adsorption and the second at state II potentials. In state I, the largest coverage is 1.5 × 10-10 mol cm-2, whereas the limiting coverage in state II more than triples to 5.0 × 10-10 mol cm-2. These limiting coverages are observed at Emax,I and Emax,II. The much lower coverage in state I is consistent with a flat-lying configuration of DMAP(H+) molecules. For comparison, Lipkowski and co-workers reported thecoveragesofhorizontallyadsorbedpyridine37 and4-cyanopyridine27a on Au(111) to be 1.4 × 10-10 and 1.2 × 10-10 mol cm-2, respectively. The similarity of the results is strong evidence that, in state I, the DMAP(H+) species adsorbs on the gold surface via the delocalized electrons of the pyridine ring. In the transition from state I to state II of adsorption, the surface coverage significantly increases but is nearly 30% less than the maximum surface coverage observed at pH 9.7. The coverage data indicate that the phase transition occurring near zero charge density is

1562 Langmuir, Vol. 23, No. 3, 2007

Barlow and Burgess

Figure 7. ∆Gads° versus potential obtained using the Henry law isotherm: (9) pH 9.7 and (4) pH 4.5.

consistent with the transformation of a horizontally adsorbed layer to a partial monolayer of vertically aligned DMAP(H+) molecules. Gibbs Energy of Adsorption. As discussed above, when the standard Gibbs energy of adsorption, ∆Gads°, is determined, the results are dependent on the choice of isotherm and definition of the standard state. The most appropriate isotherm is often dictated by the experimental data. For example, Figure 6a indicates that the coverage of DMAP(H+) is only very weakly dependent on the bulk surfactant concentration at high electrolyte pH. For this reason, the Frumkin and Langmuir isotherms cannot be used to extract ∆Gads°. However, in the limit of low coverages all isotherms simplify to the Henry’s law isotherm25

π ) RTΓmax

( ) (

)

-∆Gads° c exp 55.5 RT

(6)

where 55.5 is the molar concentration of water, c is the formal DMAP concentration, and Γmax is the limiting surface concentration. We used the Henry law isotherm to evaluate ∆Gads° at each potential by plotting our measured surface pressures versus mole fraction, c/55, and determining the tangent of these plots as π f 0 (see Figure S2 in the Supporting Information for a representative example). We defined the standard state to be the unit mole fraction of DMAP/DMAP(H+) in the bulk and the limiting surface coverage to be the maximum coverage obtained for DMAP(H+) in the vertical configuration (7.4 × 10-10 mol cm-2). The advantage of the Henry law isotherm is its simplicity and applicability to all experimental data. However, like other isotherms, the values of ∆Gads° obtained using this isotherm are dependent on the choice of Γmax, which makes it difficult to compare adsorption energies for different organic compounds or even different adsorption states of the same compound. To alleviate this problem, we used Γmax ) 7.4 × 10-10 mol cm-2 for both pH values and only calculated the values of ∆Gads° for potentials corresponding to the vertical adsorbate orientation. In Figure 7, we present plots of ∆Gads° versus a normalized (E Emax) potential scale for both pH 9.7 and pH 4.5 systems. Both curves exhibit a quasi-parabolic appearance with maximum absolute values of the standard adsorption energies at the potential of maximum adsorption (E - Emax ) 0). At all potentials, the magnitude of the energy of adsorption is larger at the higher pH by approximately 5 kJ mol-1. The most favorable energies of adsorption at pH 9.7 and 4.5 are -40 and -34 kJ mol-1, respectively, and straddle the largest adsorption energy of -38 kJ mol-1 reported by Stolberg et al. for pyridine adsorption on Au(111). On polycrystalline gold, the same group reported a

Figure 8. Models summarizing the states of DMAP(H+) adsorption on polycrystalline gold: (a) full monolayer of N-bonded DMAP at high pH and positively charged electrode densities; (b) π-bonded configuration of DMAPH+ ions observed at negatively charged surfaces and in acidic electrolyte solutions.

maximum ∆Gads° of -42 kJ mol-1.36 As these authors also used the Henry law isotherm, the values of ∆Gads° can be readily compared, although it should be noted that in the case of pyridine adsorption the value chosen for Γmax was 4.7 × 10-10 mol cm-2 for Au(111) and 6.1 × 10-10 mol cm-2 for polycrystalline gold. However, the values of Γmax for these systems do not differ by more than a factor of 1.6, which corresponds to a maximum difference in ∆Gads° of only slightly more than 1 kJ mol-1. It is apparent that the standard adsorption energy of DMAP species from high-pH electrolytes is, within reasonable estimation of the error of the experiments, nearly equivalent to that of pyridine adsorption on Au(111) and polycrystalline gold. However, at pH 4.5, the adsorption of DMAP(H+) species is significantly (∼15%) less energetically favorable. Orientation and Speciation of Adsorbed DMAP. The shift in the potential of zero charge data provides insight into the nature of the adsorbed film of DMAP species. In a simple electrostatic model, the change in the potential of zero charge (EN) arises from the displacement of m water molecules on the film-free electrode by each adsorbing molecule (DMAP or DMAPH+ in the present case):39,40

EN ) Γmax

(

)

µorg - mµw 

(7)

where Γmax is the maximum surface concentration of organic molecules,  is the dielectric of the adsorbed molecules, and µorg and µw are the dipole moments (per mole) of DMAP(H+) and water in the direction normal to the electrode’s surface. Recent reports41 have indicated that, at the pzc, water molecules are more preferentially oriented on polycrystalline gold compared to mercury or Au(111) surfaces. However, in an effort to compare our DMAP(H+) results with pyridine adsorption, we will follow the example of Stolberg et al. and assume the dipole moment is negligible, i.e., no preferential orientation of water on the gold surface. The square of the refractive index of 2-(dimethylamino)pyridine (1.57) was used as an estimate of the inner layer permittivity. In the case of pH 9.7, the calculated dipole moment per molecule, µorg, is -1.1 D (1 D ) 3.34 × 10-30 C m). Following the same procedure, the dipole moment for state II of adsorption at pH 4.5 was determined to be -1.4 D. These values are very (39) Payne, R. J. Electroanal. Chem. 1973, 41, 277. (40) Damaskin, B.; Frumkin, A.; Chizhov, A. J. Electroanal. Chem. 1970, 28, 93. (41) Becucci, L.; Moncelli, M. R.; Guidelli, R. Langmuir 2003, 19, 3386.

DMAP Adsorption on Polycrystalline Gold

close to the dipole moments reported for the vertical adsorption of pyridine for Au(111) (-0.9 D) and Au(100) (-1.3 D). The higher dipole moments reported in this work are consistent with the larger gas-phase dipole moment of DMAP (-4.31 D)42 compared to the equivalent value of pyridine (-3.0 D).37 Furthermore, the negative sign of the dipole moment indicates that DMAP is vertically bound to the gold through the endocyclic nitrogen in its state II of adsorption. As a reminder, state II is the observed adsorption state at all potentials for pH 9.7 but only appears at more positive potentials when the electrolyte pH is 4.5. In the latter case, this implies that the adsorbed species is the basic form (i.e., DMAP) even though the pH of the electrolyte is more than 5 units more acidic of the pKa. At this pH, the dipole moment for the horizontal state of adsorption (state I) can also be calculated and is found to be +0.5 D. We can once again compare this value to pyridine electrosorption where Lipkowski and co-workers reported dipole moments of -0.1 D (Au(111))37 and +0.04 D (Au(100))43 for pyridine’s flat orientation. Our value for DMAP(H+) is significantly more positive and strongly indicates that in state I the adsorbed species is DMAPH+ lying flat on the surface with perhaps a small angle of inclination.

Summary and Conclusions Using electrochemical measurements, we have provided a detailed study of the pH- and potential-dependent electrosorption of 4-(dimethylamino)pyridine on polycrystalline gold. Our results indicate that at pH values at or above the primary pKa, the adsorbed species is DMAP and adsorbs vertically on the electrode surface via the free electrons on the pyridine ring’s nitrogen atom. At very low pH values (