Nitrophenyl Groups in Diazonium-Generated Multilayered Films

pubs.acs.org/Langmuir © 2010 American Chemical Society

Nitrophenyl Groups in Diazonium-Generated Multilayered Films: Which are Electrochemically Responsive? Marcel Ceccato, Lasse Tholstrup Nielsen, Joseph Iruthayaraj, Mogens Hinge, Steen Uttrup Pedersen,* and Kim Daasbjerg* Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus, Denmark, and Interdisciplinary Nanoscience Center (iNANO), Institute of Physics and Astronomy, Ny Munkegade 120, DK-8000 Aarhus C, Denmark Received February 11, 2010. Revised Manuscript Received March 31, 2010 Various nitrophenyl-containing organic layers have been electrografted to glassy carbon surfaces using diazonium chemistry to elucidate the extent by which the layer structure influences the solvent (i.e., acetonitrile) accessibility, electroactivity, and chemical reactivity of the films. For most of these films, cyclic voltammetric and impedance spectroscopy measurements show that the electron-transfer process at the electrode is facile and independent of film thickness and structure. This is consistent with the occurrence of self-mediated electron transfers throughout the film with nitrophenyl groups serving as redox stations. Importantly, this behavior is seen only after the first potential sweep, the effect of which is to increase the porosity of the layer by inducing an irreversible desorption of nonchemisorbed material along with a reorganization of the film structure. From a kinetic point of view, the radical anions of surfaceattached nitrophenyl groups are reactive toward the residual water present in acetonitrile. Thin layers (thickness of 1 to 2 nm) containing redox-active groups only in the outer part of the layer are protonated two to three times as fast as groups located in a more hydrophobic but still solvent-accessible inner layer. Hence, kinetic measurements can detect small differences in the layer environment. Finally, a deconvolution of the cyclic voltammetric response of an electrode grafted from 4-nitrobenzenediazonium discloses that roughly 25% of the overall signal can be attributed to the presence of 4-azonitrophenyl moieties introduced during the electrografting process.

Introduction

*Corresponding authors. (S.U.P) Tel: þ45 8942 3908. Fax: þ45 8619 6199. E-mail: [email protected]. (K.D.) Tel: þ45 8942 3922. Fax: þ45 8619 6199. E-mail: [email protected].

induction by simply immersing the substrate into a solution of diazonium salt.10 Covalent bond formation finds its origin in the surface-directed attack of the aryl radicals that are generated close to the surface upon reduction of the aryl diazonium salts. In most cases, the reactivity of the aryl radicals is so high that they also readily attack already-grafted aryl groups in a rather uncontrolled process to form a multilayered aryl-based structure.11,12 The resulting nanometer-thick multilayer is seldom well-defined because the aryl groups are located in an inhomogeneous environment. Quantitative studies of electron transfer and chemical reactivity as well as the development of high-performance sensors or molecular devices are difficult to accomplish with such inhomogeneous films. However, the recent introduction of the “formation-degradation” approach13 and the use of sterically hindered diazonium salts14 or aryl hydrazines15 as grafting agents holds great promise for creating surfaces covered by what is essentially a monolayer. Careful adjustment of the electrolysis potential and time16,17 and the choice of the grafting medium18 can provide

(1) Pinson, J.; Podvorica, F. I. Chem. Soc. Rev. 2005, 34, 429–439. (2) Downard, A. J. Electroanalysis 2000, 12, 1085–1096. (3) Delamar, M.; Hitmi, R.; Pinson, J.; Sav
eant, J-.M. J. Am. Chem. Soc. 1992, 114, 5883–5889. (4) Vase, K. H.; Holm, A. H.; Pedersen, S. U.; Daasbjerg, K. Langmuir 2005, 21, 8085–8089. (5) Vase, K. H.; Holm, A. H.; Norrman, K.; Pedersen, S. U.; Daasbjerg, K. Langmuir 2008, 24, 182–188. (6) Boukerma, K.; Chehimi, M. M.; Pinson, J.; Blomfield, C. Langmuir 2003, 19, 6333–6335. (7) Bernard, M.-C.; Chauss
e, A.; Cabet-Deliry, E.; Chehimi, M. M.; Pinson, J.; Podvorica, F.; Vautrin-Ul, C. Chem. Mater. 2003, 15, 3450–3462. (8) Henry de Villeneuve, C.; Pinson, J.; Bernard, M. C.; Allongue, P. J. Phys. Chem. B 1997, 101, 2415–2420. (9) Mirkhalef, F.; Paprotny, J.; Schiffrin, D. J. J. Am. Chem. Soc. 2006, 128, 7400–7401. (10) Bahr, J. L.; Tour, J. M. Chem. Mater. 2001, 13, 3823–3824.

(11) Kariuki, J. K.; McDermott, M. T. Langmuir 1999, 15, 6534–6540. (12) Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. Langmuir 2005, 21, 280–286. (13) Nielsen, L. T.; Vase, K. H.; Dong, M.; Besenbacher, F.; Pedersen, S. U.; Daasbjerg, K. J. Am. Chem. Soc. 2007, 129, 1888–1889. (14) Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. J. Am. Chem. Soc. 2008, 130, 8576–8577. (15) Malmos, K.; Iruthayaraj, J.; Pedersen, S. U.; Daasbjerg, K. J. Am. Chem. Soc. 2009, 131, 13926–13927. (16) Stewart, M. P.; Maya, F.; Kosynkin, D. V.; Dirk, S. M.; Stapleton, J. J.; McGuiness, C. L.; Allara, D. L.; Tour, J. M. J. Am. Chem. Soc. 2004, 126, 370–378. (17) Ricci, A.; Bonazzola, C.; Calvo, E. J. Phys. Chem. Chem. Phys. 2006, 8, 4297–4299. (18) Malmos, K.; Dong, M.; Pillai, S.; Kingshott, P.; Besenbacher, F.; Pedersen, S. U.; Daasbjerg, K. J. Am. Chem. Soc. 2009, 131, 4928–4936.

It is well established that aromatic organic layers can be covalently tethered to carbon-based surfaces by the electrochemical reduction of aryl diazonium salts.1,2 Since the first study on the subject was published in the early 1990s,3 the procedure has attracted considerable interest and films grafted from numerous substituted aryl diazonium and similar salts4,5 have also been examined on conducting surfaces other than carbon such as metals and semiconductors.6-8 A recent paper has also shown that diazonium salts can serve as ligands in the formation of nanoparticles.9 One among several advantages of covalent attachment is that it results in robust surface grafting that is able to withstand elevated temperatures and harsh chemical conditions. Another advantage of the diazonium salt approach is that spontaneous grafting can take place with no electrochemical

10812 DOI: 10.1021/la1006428

Published on Web 04/22/2010

Langmuir 2010, 26(13), 10812–10821

Ceccato et al.

Article

additional control over the grafting process itself. Ultimately, the possibility of placing redox-active molecules in well-defined surroundings will enable the creation of modified films sharing features with those of self-assembled monolayers (SAMs), which would be of immense importance in the understanding of fundamental surface-confined electron-transfer processes.19 Also, a fundamental understanding of heterogeneous reactions between surface-confined molecules and solution-based reagents is more easily attainable with well-defined layers. In this work, we use covalently attached multilayered films containing redox-active 4-nitrophenyl (NP) units placed deliberately at fixed positions and in various environments on glassy carbon (GC) electrodes to characterize their electrochemical addressability. The nice feature of the NP unit is its characteristic cyclic voltammetric behavior that allows the extraction of electrochemical data of high relevance to this work. Furthermore, the extraction of charge-transfer characteristics from peak separations in cyclic voltammetry as well as charge-transfer resistance from impedance data is straightforward. Finally, the reactivity of NP 3 - (i.e., the reduced form of the multilayered NP film generated transiently in cyclic voltammetry) toward residual water in acetonitrile (MeCN) can be used to probe the role that a specific environment exerts on the kinetic properties. Outer groups being in the boundary region between the solid and liquid might be expected to be more reactive than inner groups located in a better-protected hydrophobic environment. The interpretation of all data is facilitated by the fact that XPS20 and TOF-SIMS21 analyses of layers grafted from 4nitrobenzenediazonium are already available in the literature. It is noteworthy that a substantial amount of azobenzenes has been revealed in the films, which were presumably formed during the grafting process.

Experimental Section Materials. Acetonitrile (MeCN), acetone, and ethanol were of analytical grade and used as received without further purification. Toluene was freshly distilled from sodium prior to use. 1Isocyanato-4-nitrobenzene was obtained from Aldrich and used without further purification. 4-(2-Ammonioethyl)benzenediazonium, 4-nitrobenzenediazonium, and benzenediazonium tetrafluoroborates were synthesized from the corresponding anilines according to procedures published elsewhere.22 The synthesized compounds were purified by dissolution in either MeCN or acetone, precipitation by the addition of diethyl ether, filtration, rinsing, and vacuum drying, finally to be stored at -18 °C. Tetrabutylammonium tetrafluoroborate (Bu4NBF4) was synthesized using standard procedures. Ferrocene (Aldrich) and Ru(NH3)6Cl3 (ABCR) were obtained from commercial sources. The electrolyte solution applied was 0.1 M Bu4NBF4 in MeCN, which was dried by letting it pass through a column containing Al2O3 (thermally activated) prior to use. After this treatment, the water content was 0.03 ( 0.01 wt % as determined by Karl Fischer titration (TitraLab 980 from Radiometer Analytical). Electrodes. Glassy carbon (GC) rods (Sigradur G, HTW, diameter=1 mm) were embedded in epoxy resin, sanded (with sandpaper grids P180, P500, P1000, and P2000), and carefully polished with diamond suspensions (Struers, grain sizes 9, 3, 1, and 0.25 μm). Afterwards, the electrodes were washed thoroughly (19) Chidsey, C. E. D. Science 1991, 251, 919–922. (20) Saby, C.; Ortiz, B.; Champagne, G. Y.; B
elanger, D. Langmuir 1997, 13, 6805–6818. (21) Doppelt, P.; Hallais, G.; Pinson, J.; Podvorica, F.; Verneyre, S. Chem. Mater. 2007, 19, 4570–4575. (22) Starkey, E. B. Organic Syntheses; Wiley & Sons: New York, 1943; Collect. Vol. 2, pp 225-227.

Langmuir 2010, 26(13), 10812–10821

with distilled water and ethanol (96%) and sonicated for 5 min in ethanol (96%). GC plates (Sigradur G, HTW, 10  10  1 mm) employed for the ellipsometry measurements were cleaned by sonication in acetone and hexane (1 h in each solvent). Electrochemical Setup. A standard three-electrode electrochemical setup (CH Instruments 660B or CH Instruments 601C) consisting of a GC rod or plate as the working electrode, a platinum wire as the auxiliary electrode, and a Ag/AgI pseudoreference electrode was used in all electrochemical experiments. At the end of each experiment, the standard potential of the ferro0 cenium/ferrocene couple, EFc þ, was measured and all potentials were referenced against SCE using a previous determination of 0 23 EFc þ =0.41 V versus SCE in MeCN. An important experimental condition for all measurements reported is that the electrodes were not immersed in the solution before the experiment was started. In electrografting procedures, this minimized the amount of spontaneous adsorption of the aryl diazonium salt, and for already-prepared films, the soaking time could be let out as a variable. In Scheme 1, the various modification protocols along with the appropriate abbreviations for the modified electrodes prepared are shown. Unless otherwise noted, electrografting is carried out by means of the potentiostatic electrolysis of 2 mM aryl diazonium solutions in MeCN (without a supporting electrolyte) for 300 s at a grafting potential of Egraft=Ep - 200 mV, where Ep is the peak potential of the diazonium group determined by an initial cyclic voltammetric sweep (sweep rate=0.2 V s-1).

Preparation of GC-NPfs Electrodes Using a Fast Cyclic Voltammetric Sweep on 4-Nitrobenzenediazonium Salt. Electrodes were prepared via a single voltammetric cycle using a high sweep rate of 2 V s-1 in a 1 mM solution of 4-nitrobenzenediazonium salt in MeCN. The switch potential was set at Ep 200 mV. Electrodes were flushed with solvent, followed by sonication in MeCN or acetone for 5 min. 0

Preparation of GC-NPY and GC-NP800 Electrodes with Varying Film Thickness from Electrografting with 4-Nitrobenzenediazonium Salt. GC electrodes were electro-

grafted with 4-nitrobenzenediazonium salt for 300 s at a given grafting potential of Egraft=Ep -Y mV (Y=200, 400, 600, or 800) followed by sonication in MeCN or acetone for 5 min. In each case, the grafting current at the end of electrolysis is close to zero, showing that all of the electrografting processes, independent of driving force and the resulting film thickness, are self-limiting because of the formation of an insulating multilayer. The GC0 NP800 electrode is produced at Egraft=Ep - 800 mV using a short electrolysis time of 1.9 s. 0

Preparation of GC-Ph -NP Electrodes from Electrografting with Benzenediazonium and 4-Nitrobenzenediazonium Salts. GC electrodes were first electrografted with benzenedia-

zonium salt. After sonication in MeCN, the now phenyl-modified 0 (or more correctly polyphenylene-modified) electrodes (GC-Ph ) were electrografted with 4-nitrobenzenediazonium0 salt at Egraft= -0.8 V versus SCE for 300 s to give GC-Ph -NP. Finally, sonication of the electrodes in MeCN or acetone was carried out for 5 min. 0 0

Preparation of GC-NPfs-Ph and GC-NP200-Ph Electrodes from Electrografting with 4-Nitrobenzenediazonium and Benzenediazonium Salts. GC electrodes were grafted first with

4-nitrobenzenediazonium salt using either a fast cyclic voltammetric sweep as described above (GC-NPfs) or electrolysis to produce GC-NP200. After sonication in MeCN, the now NP-modified electrodes were electrografted with benzenediazonium salt using 10 cyclic voltammetric cycles in MeCN (initial potential, 0.9 V vs SCE; switch 0potential, -0.8 V vs0 SCE; sweep rate, 0.2 V s-1) to give GC-NPfs-Ph and GC-NP200-Ph , respectively. Finally, sonication of the electrodes in either MeCN or acetone was carried out for 5 min. (23) Daasbjerg, K.; Pedersen, S. U.; Lund, H. In General Aspects of the Chemistry of Radicals; Alfassi, Z. B., Ed.; Wiley: Chichester, U.K., 1999; pp 385-427.

DOI: 10.1021/la1006428

10813

Article

Ceccato et al. Scheme 1. Protocols for Modifying GC Electrodes with Covalently Attached Filmsa

a (A) Films consisting exclusively of NP groups produced in various thicknesses by using fast cyclic voltammetry (i.e., GC-NPfs, sweep rate=2 V s-1), varying the grafting potential in the electrolysis (3000 s) of 4-nitrobenzenediazonium (i.e., GC-NPY, where Egraft=Ep - Y mV; Y=200, 400, 600, or 800), and using a short electrolysis time (i.e., GC-NP800 ; electrolysis time= 1.9 s). (B) Outer or inner layer of NP groups obtained in two-step grafting procedures involving benzenediazonium and 4-nitrobenzenediazonium salts (i.e., GC-Ph0 -NP, GC-NPfs-Ph0 , and GC-NP200-Ph0 ). (C) Outer layer of NP groups using electrografting of 4-(2-ammonioethyl)benzenediazonium followed by a nucleophilic addition reaction with 1-isocyanato-4-nitrobenzene (i.e., GC-10 -NP). 0

Preparation of GC-1 -NP Electrodes from Electrografting with a 4-(2-Ammonioethyl)benzenediazonium Salt Followed by a Chemical Reaction with 1-Isocyanato-4-nitrobenzene. GC electrodes were electrografted using 4-(2-ammonioethyl)benzenediazonium salt. The 2-phenylethylammonium-modified electrode (GC-1Hþ) was sonicated for 5 min in a Bu4NOH solution (consisting of a drop of 40% Bu4NOH(aq) in MeCN) to deprotonate the ammonium group. After deprotonation, the electrodes (i.e., GC-1) were rinsed in MeCN and immersed in 20 mM 1-isocyanato-4-nitrobenzene in toluene. The reaction was allowed to proceed for 60 min on a shaking board. The derivatized 0 electrodes (GC-1 -NP) were rinsed by sonication for 10 min in DMF and MeCN or acetone. Finally, they were dried in a stream of nitrogen. Modification of GC Plates. The electrochemical derivatization of GC plates employed for ellipsometrical measurements was carried out using procedures as described above for the electrodes, the exception being that the supporting electrolyte (0.1 M Bu4NBF4 in MeCN) was used to lower the electrical resistance of the solutions. It was verified that the GC plates [placed in a holder with an O-ring (diameter = 3 mm) to reduce the surface area exposed to the electrolyte solution] and the grafted GC electrodes (diameter = 1 mm) prepared with and without the presence of supporting electrolyte gave the same electrochemical response when normalized with respect to the surface area. Analysis of CV Data. The total charge (Q) used for the reduction of the surface-grafted electroactive groups was obtained by numerical integration of the background-subtracted electrochemical response recorded in cyclic voltammetry. First, the background was subtracted using a fourth-degree polynomial function (obtained from a fitting of the data on either side of the wave). Then Q was determined by numerical integration of the electrochemical signal (see Supporting Information for the spe0 cific method employed for GC-NPfs, GC-NP800 , and GC-NPY). It is important to note that for film-coated electrodes there is an uncertainty associated with establishing an appropriate baseline because the properties of the film change during sweeping. Yu et al. have discussed this issue using polynomial background 10814 DOI: 10.1021/la1006428

subtraction and have concluded that the uncertainty could be higher than 20%.24 The surface coverage, Γ, was calculated from the values of Q using Faraday’s law (i.e., Γ=Q/nFAgeo, where n is the number of electrons transferred, F is the Faraday constant, and Ageo represents the geometric area of the electrode). The n value for the reduction of the NP substituent in MeCN is assumed to be 1. The fact that the voltammograms exhibit a high degree of reversibility justifies such a presupposition. However, considering that a certain fraction of the radical anions once generated can be protonated and further reduced, n may be somewhat larger than 1 for the forward sweep. This would give an overestimation of Γ but, importantly, would have no effect on the subsequent extraction of rate constants (vide infra) as long as n is the same for each of the repetitive sweeps. Likewise, the exact value of Ageo will have no influence on the rate determination, despite Γ being inversely proportional to Ageo. The decay rate constants, kdecay, were obtained from logarithmic fits to the equation Γ=A exp(-kdecay  tave), where A is the pre-exponential factor and tave is the average time in which the redox probe is in its radical anion form. The latter parameter is calculated by the cumulative addition of the time spent on passing from the reductive to the oxidative peak potential for all cycles with the zero point for tave set at the peak of the first sweep. The error in tave calculated by this method is small because both the reductive and oxidative peaks are Gaussian bell-shaped with 0 similar peak widths. For the GC-NP800 and GC-NPY films, Γ determined from the first sweep is much lower than that determined from the second sweep and the first sweep was omitted from the fitting analysis. A subsequent extrapolation of the fit to tave =0 provides an accurate determination of the initial surface coverage. Each experiment is repeated three to five times. Electrochemical Impedance Spectroscopy (EIS). EIS was recorded on the freshly polished GC and GC-NPY (Y =200 or 800) electrodes before and after carrying out a single cyclic (24) Yu, S. S. C.; Tan, E. S. Q.; Jane, R. T.; Downard, A. J. Langmuir 2007, 23, 11074–11084.

Langmuir 2010, 26(13), 10812–10821

Ceccato et al.

Article

Figure 1. Sketch illustrating the layer morphology of the various modified electrodes. voltammogram in MeCN (initial potential, 0.4 V vs SCE; switch potential, -2.1 V vs SCE; sweep rate, 2 V s-1). The purpose of the sweep is to induce a higher permeability of the film toward ions and solvent; these electrodes are denoted GC-NPY open. The electroactivity of the electrodes was evaluated by EIS in the presence of a redox probe. The specific probes and solutions used were Ru(NH3)6Cl3 (5 mM; 0.1 M KCl, phosphate-buffered solution at pH 7) and ferrocene (5 mM; 0.1 M Bu4NBF4 in MeCN). A frequency range of 1 Hz-100 kHz was applied to the GC-NPY open and bare GC electrodes, and 0.05 Hz-100 kHz was used for the GC-NPY electrodes with a signal amplitude of 5 mV. The ac impedance fitting program supplied with the equipment was used to analyze the electrochemical impedance data and fit them to an appropriate equivalent electrical circuit (Supporting Information). Ellipsometry. The thickness of the electrografted films in the dry state was measured using a rotating analyzer ellipsometer (Dre, Germany). The GC plates were measured at a 65° angle of incidence. The ellipsometric parameters of the bare (Δs, Ψs) and grafted (Δf, Ψf) substrates were measured in air at ambient temperature, where Δ is the phase shift and tan Ψ is the amplitude ratio upon reflection. The complex refractive index of the bare substrate was calculated from the measured Δs and Ψs values. A three-layer optical model consisting of a substrate with a complex refractive index, the grafted layer with a refractive index nf and thickness df, and the surrounding medium (air) was used to calculate the overall reflection coefficients for in-plane (Rp) and out-of-plane (Rs) polarized light using the method described elsewhere.25,26 Ellipsometric measurements were performed on the same area of the plates before and after electrografting. From the bare GC plates, the real and imaginary parts of the refractive index could be obtained. Unfortunately, one specific value for all of the GC substrates cannot be used because of surface heterogenities. The range of refractive index values that we have obtained for various bare substrates was for the real and imaginary parts, 2.06-2.46 and from -1.8 to -0.93, respectively. Because the measurements are carried out under dry conditions on collapsed films, the refractive index of the layer is fixed at a constant value (i.e., 1.55), independent of the thickness. The average and standard deviation values reported correspond to data points obtained with three spot measurements on each plate.

Results and Discussion All modified electrodes were prepared by placing electroactive NP groups either close to or further away from the surface following the strategies outlined in Scheme 1 (Experimental Section). Figure 1 shows a sketch of the layer morphology of the modified electrodes with respect to their electrochemically active NP and inactive layers with the latter being made either of (25) Tiberg, F.; Landgren, M. Langmuir 1993, 9, 927–932. (26) Landgren, M.; Joensson, B. J. Phys. Chem. 1993, 97, 1656–1660.

Langmuir 2010, 26(13), 10812–10821

polyphenylene or phenylethanamine moieties. The film thicknesses of intermediate (or inner) layers, dinner, and the entire film, dtotal, were determined by ellipsometry in the dry state after ultrasonic rinsing of the grafted plates. The thickness of the outer layer, douter, was calculated as the difference between dtotal and dinner. These data are listed in Table 1 along with the decay rate constant for NP 3 -, kdecay, and the surface coverage of NP groups, Γ. One-Component NP Films. For GC-NPfs with a thin layer of NP groups grafted directly onto the electrode surface using a single fast sweep in cyclic voltammetry, the thickness is determined by ellipsometry to be 0.7 nm, which would correspond on average to a monolayer of NP. However, it is important to emphasize that this layer is not expected to be close-packed but instead to have a rather inhomogeneous structure because of the electrografting occurring via uncontrollable radical reactions on an already inhomogeneous GC surface. For the GC-NPY films prepared by electrolysis, the thickness of the GC-NPY electrodes is, as expected, substantially larger. It increases from 5.2 to 11.4 nm as the grafting potential is set at a progressively more negative value (i.e., as the driving force of the grafting process is 0 increased). The GC-NP800 electrode is produced using an electrolysis time of only 1.9 s in contrast to the usual 300 s. With a thickness of 5.3 nm, it is comparable to that of GC-NP200 and thus . approximately half the size of that of GC-NP800 0 0 Two-Component Films. The GC-NPfs-Ph or GC-NP200-Ph films are characterized by containing exclusively NP groups in inner layers with thicknesses of 0.7 and 5.5 nm, respectively, depending on if a fast cyclic voltammetric sweep or electrolysis is used for the initial deposition. In a second electrochemical step, both of these layers are covered by a 3- to 4-nm-thick outer polyphenylene0 layer. The GC-Ph -NP electrode represents the opposite situation in which a relatively thin layer of NP is on top (douter=1.7 nm) of a pregrafted polyphenylene layer (dinner = 8.7 nm). The fact that the GC-Ph0 film is thicker than the equivalent GC-NP200 (dtotal= 5.2 nm) film and, in particular, the GC-1 (i.e., a 2-phenylethanamine-modified electrode of 2.1 nm thickness) film can be attributed to the polyphenylene layer being more conducting.27 In particular, on some metal surfaces this property is so pronounced that it leads to the formation of micrometer-thick films.28 Also, the GC-10 -NP electrode has an outer layer of NP groups that are generated by means of an efficient nucleophilic (27) This suggests that the presence of a substituent on the phenyl ring through steric inhibitions prevents the optimization of the π-π interactions between individual phenyl rings during the grafting process. Also, the substitution pattern at the phenyl ring is expected to be greatly affected by the electronic effect of the substituent. Finally, electrostatic effects may be large in the case of GC-1, considering that the amino group is in its protonated form (i.e., GC-1Hþ). (28) Adenier, A.; Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. Chem. Mater. 2006, 18, 2021–2029.

DOI: 10.1021/la1006428

10815

Article

Ceccato et al.

Table 1. Film Thicknesses of Inner Layers, dinner, Outer Layers, douter, and the Entire Film, dtotal; the Decay Rate Constant of NP.-, kdecay; and Surface Coverage of NP Groups, Γ electrodes

dtotal (nm)a

dinner (nm)b

douter (nm)c

kdecay (s-1)d

1010 Γ (mol cm-2)e

GC-NPfs 0.7 ( 0.3 0.7 ( 0.3 f 0.23 3.6 g GC-NP200 5.2 ( 0.8 0.09 h 12 i,j GC-NP400 6.7 ( 0.8 0.10 h 15 i,j GC-NP600 7.5 ( 1.2 0.11 h 16 i,j GC-NP8000 11.4 ( 1.1 0.12 h 17 i,j GC-NP800 0 5.3 ( 0.5 0.09 h 14 i,j GC-NPfs-Ph 0 4.8 ( 1.0 0.7 ( 0.3 4.1 ( 1.3 NA ∼0 k 8.7 ( 0.1 5.5 ( 0.6 3.2 ( 0.7 11l GC-NP0 200-Ph GC-Ph -NP 10.4 ( 0.2 8.7 ( 0.5 1.7 ( 0.7 0.27 16g 0 GC-1 -NP 2.4 ( 0.2 2.1 ( 0.3 0.3 ( 0.5 0.23 3.2g a Thickness of the entire film, determined by ellipsometry. b Thickness of the inner layer, determined by ellipsometry. c Thickness of the outer layer, calculated as douter=dtotal - dinner. d Determined from the expression Γ=A exp(-kdecay  tave); the uncertainty is 20% e Obtained from an integration of the cyclic voltammetric response in 0.1 M Bu4NBF4/MeCN; the uncertainty is 20%. f Assumed to be the same as dtotal. g Determined from the first cycle. h Excluding the data from the first cyclic voltammogram in the exponential fit. i The contribution from nitroazobenzenes has been excluded (see the text). j Determined from an extrapolation of Γ to zero time from the fit of Γ vs tave (see the text). k Not measured. l Determined from the second cycle.

addition reaction between GC-1 and 1-isocyanato-4-nitrobenzene.29 Because the diffusion of the latter into the film is limited by its molecular size, the reaction is expected to take place predominantly in the outermost part of the layer forming nitrophenylurea groups. This is not to say that reactions along pin holes or pores may not occur. Indeed, in such cases the concept “outer groups”, despite being pictorially strong, should be understood in a broader sense as easily solvent-accessible groups. The increase in the thickness of the layer upon nucleophilic addition (douter=0.3 nm) is smaller than expected, considering that the length of a single 4nitrophenylurea unit is ∼0.8 nm. To which extent the conformational changes occurring in the inner layer upon being exposed to the specific reaction conditions involving toluene contribute to this difference is not known. Cyclic Voltammetry. Figure 2A-F shows voltammograms of the GC-NPfs, GC-NPfs-Ph0 , GC-NP200, GC-10 -NP, GC-NP200Ph0 , and GC-Ph0 -NP electrodes, respectively. For GC-NP200, GCNP200-Ph0 , and GC-Ph0 -NP, the first cycle is not included (vide infra). Also, it may be noted that the voltammograms for GCNPfs, GC-NP200, and GC-NP200-Ph0 in Figure 2A,C,E, respectively, exhibit a reversible prewave system at around -0.9 V versus SCE. In contrast, the voltammograms of GC-10 -NP (Figure 2D) and GC-Ph0 -NP (Figure 2F) show no such features. We assign the prewave to the presence of 4-nitroazobenzenederived moieties formed during the electrografting process, and as discussed in detail in the Supporting Information (Figures 1S and 2S), they can amount to as much as 25% of all electroactive groups. Concerning the NP wave itself, we may use the average of the reduction and oxidation peak potentials, Emid, as a good approximation of the standard potential. Essentially, Emid (= -1.2 V vs SCE) is equal to that determined previously for the surfaceattached NP/NP 3 - redox pair in MeCN.30 The only exception is GC-10 -NP (Emid =-1.4 V vs SCE), in which we attribute the ∼200 mV more negative value of Emid to the electron-donating effect of the urea group formed in the nucleophilic addition reaction (Scheme 1C). Another noticeable feature in Figure 2 is that the peak potential separation, ΔEp, increases significantly from approximately0 zero for GC-NPfs, GC-NP200 (as well as GC-NPY and GC-NP800 , vide infra), and GC-10 -NP to 95 and 120 mV for GC-Ph0 -NP and GCNP200-Ph0 , respectively. The common feature of the electrodes exhibiting nearly reversible behavior (i.e., ΔEp ≈ 0 mV) is that the (29) Nielsen, L. T.; Ceccato, M.; Holm, A. H.; Kristensen, M. V.; Pedersen, S. U.; Daasbjerg, K. Langmuir 2009, 25, 12160–12168. (30) Downard, A. J. Langmuir 2000, 16, 9680–9682.

10816 DOI: 10.1021/la1006428

Figure 2. Successive voltammograms recorded in 0.1 M Bu4NBF4/MeCN at a sweep rate of 2 V s-1 for different NP-modified electrodes (A) GC-NPfs, (B) GC-NPfs-Ph0 , (C) GC-NP200, (D) GC-10 -NP, (E) GC-NP200-Ph0 , and (F) GC-Ph0 NP. (A, E) Some of the voltammograms in the series have been omitted for clarity.

NP groups are present in close proximity to the surface, no matter the thickness of the film. Still, the presence of a thick nonpolar polyphenylene top layer in GC-NP200-Ph0 is able to slow down the rate of electron transfer, presumably by diminishing the permeation rate of the electrolyte into and/or out of the inner NP part of the film but without affecting the size of the current signal (at least on the forward sweep). Surprisingly, there is the almost complete disappearance of the current signal of the related GC-NPfs-Ph0 film, but we suspect that in this case the very thin NP layer in GC-NPfs to a great extent may be chemically affected (i.e., essentially destroyed) during the second grafting step with the benzenediazonium salt involving phenyl radicals. Another way of slowing down the electron-transfer process is to place the NP groups farther away from the electrode surface by introducing an intermediate organic layer. Hence, ΔEp increases by almost 100 mV from GC-10 -NP (dinner=2.1 nm) to GC-Ph0 -NP (dinner= 8.7 nm). Langmuir 2010, 26(13), 10812–10821

Ceccato et al.

Figure 3. First and second cyclic voltammetric sweeps of GCNP200 and GC-NP800 in 0.1 M Bu4NBF4/MeCN.

To shed further light on the observation that even for the relatively thick GC-NP200 film the NP/NP.- pair exhibits almost reversible behavior (i.e., ΔEp ≈ 0 mV), we decided to investigate a series of GC-NPY films in which the film thickness was increased by grafting for 300 s at an increasingly more negative grafting potential [Egraft = Ep 0- Y mV (Y = 200, 400, 600, or 800)]. In addition, a GC-NP800 film was produced by performing electrolysis for a particularly short time (1.9 s) at Egraft=Ep - 800 mV. As already noted, the reversible behavior of GC-NP200 reveals itself only upon performing a potential sweep excursion in the negative direction. The same is true for all other films in this series. In Figure 3, we have collected the two first cycles for the GCNP200 and GC-NP800 films to illustrate the phenomenon that arises upon carrying out potential sweeps in the negative direction. Similar voltammograms for GC-NP400, GC-NP600, GC8000 NP , GC-NP-Ph0 , and GC-Ph0 -NP are available in the Supporting Information (Figure S3). Evidently, the reduction wave pertaining to the NP group on the first cycle is less well defined than that on the second and subsequent cycles, and at the same time it appears at more negative potentials. Hence, the first reductive sweep is sufficient to produce a film that will respond in an electrochemically reproducible way on the reverse oxidative sweep. In general, the differences between the first and second cycles are more pronounced as Y increases in the GC-NPY series, (i.e., as the film thickness increases from 5.2 to 11.4 nm). Furthermore, it can be seen that the ΔEp0 measured at the second cycle for all films, including GC-NP800 , is close to zero. This behavior is consistent with a fast reversible electron-transfer process taking place between the GC surface and the innermost NP groups. The current signals originating from those would be very small Langmuir 2010, 26(13), 10812–10821

Article

unless the layer had a high permeability toward the electrolyte, thereby allowing a self-mediated electron-transfer process to take place throughout the film. In this manner, an amplification of the signal will occur, depending on the total number of solventaccessible redox groups in the film. To extract additional information about the GC-NPY electrodes, we turned our attention toward one of the strongest electrochemical techniques available in the characterization of modified films, that is, electrochemical impedance spectroscopy (EIS) on solution-based redox probes. Electrochemical Impedance Spectroscopy. To obtain as comprehensive a picture as possible, we selected two redox probes in two different solvent systems as our objects of study (i.e., Ru(NH3)63þ in aqueous pH 7 solution and ferrocene in 0.1 M Bu4NBF4/MeCN). The GC-NP200 and GC-NP800 films were investigated both before and after a potential excursion in the negative direction in order to open the grafted layer. Henceforth, 800 these electrodes are denoted GC-NP200 open and GC-NPopen. We also 3-/4as a redox probe, but the attempted to use Fe(CN)6 reproducibility was found to be so low that it was omitted for further studies. In Figures 4 and 5, cyclic voltammograms of Ru(NH3)63þ and ferrocene, respectively, are shown for bare GC, GC-NP200, GC800 NP800, GC-NP200 open, and GC-NPopen along with the corresponding Nyquist plots from EIS measurements. The cyclic voltammograms show the same picture in that the modified films on the GCNP200 and, in particular, the GC-NP800 electrodes are strongly blocking toward the redox probes as evidenced by the large distortions of the voltammograms. However, upon performing a single potential excursion in the negative direction a re-establishment of the original signal pertaining to the reversible electron transfer of the redox couples at a bare GC electrode is by and large observed, especially in the ferrocene case. 800 The Nyquist plots for GC-NP200 open and GC-NPopen show a semicircle with a depressed shape followed by a Warburg line that in most cases could be fitted to a simple equivalent electrical circuit (Supporting Information, Figure S4). The same circuit was used for bare GC electrodes. In contrast, for initially grafted electrodes GC-NP200 and GC-NP800, the Nyquist plots had poorly reproducible shapes with even two semicircles appearing in several cases, which made it impossible to fit the data to the simple circuit. It was therefore decided to make a rough estimation of the charge-transfer resistance, RCT, by extrapolating the semicircle to the intersection with the Z 0 axis. A single potential excursion is sufficient to induce the changes in the film structure because neither a second nor a third excursion leads to an additional decrease in RCT. All extracted values of RCT are collected in Table 2. First, it may be noted that RCT for GC-NP200 and, in particular, for GC-NP800, although not directly extractable from the Nyquist 800 plots, is much larger than for GC-NP200 open and GC-NPopen. This feature is also corroborated by the distorted nature of the cyclic voltammograms seen in the two former cases (Figures 4 and 5). The effect of the potential excursion is a decrease in RCT from values on the order of 105 to 103 Ω. Note that the latter value is only a factor of 2 higher than RCT for ferrocene on a bare GC electrode. This indicates that the permeability of the films to the redox probe becomes not only high but also approximately the same for all thickness values (Table 1) once the potential-induced opening of the grafted layer has occurred. Furthermore, this observation indicates that the film structures of the open films are more porous compared to those of GC-NP200 and GC-NP800, which will facilitate an easy exchange of the redox probes between the DOI: 10.1021/la1006428

10817

Article

Ceccato et al.

800 Figure 4. (A) Cyclic voltammograms recorded at GC ( 3 3 3 ), GC-NP200 (-), GC-NP800 (---), GC-NP200 open, (-) and GC-NPopen (---) electrodes of 5 mM Ru(NH3)63þ (0.1 M KCl phosphate-buffered solution at pH 7) at a sweep rate of 0.05 V s-1. (B) Corresponding Nyquist plots 200 800 obtained for GC-NP200 and GC-NP800. open (b) and GC-NPopen (O); the inset shows plots for GC-NP

800 Figure 5. (A) Cyclic voltammograms recorded at GC ( 3 3 3 ), GC-NP200 (-), GC-NP800 (---), GC-NP200 open (-), and GC-NPopen (---) electrodes -1 200

of 5 mM ferrocene in 0.1 M Bu4NBF4/MeCN at a sweep rate of 0.1 V s . (B) Corresponding Nyquist plots obtained for GC-NPopen (b) and 200 GC-NP800 and GC-NP800. open (O); the inset shows plots for GC-NP Table 2. Charge-Transfer Resistance, RCT, Measured for Ru(NH3)63þ and the Ferrocene Redox Probes at GC, GC-NP200, 800 GC-NP800, GC-NP200 open, and GC-NPopen Electrodes RCT (Ω) film

Ru(NH3)63þa

ferroceneb

GC