Mediated Oxidation of Ascorbic Acid on a Homologous Series of

Langmuir 2007, 23, 4965-4971

4965

Mediated Oxidation of Ascorbic Acid on a Homologous Series of Ferrocene-Terminated Self-Assembled Monolayers Birute˘ Kazakevicˇiene˘ ,† Gintaras Valincius,*,† Gediminas Niaura,† Zita Talaikyte˘ ,† Maryte˘ Kazˇeme˘ kaite˘ ,† Valdemaras Razumas,† Deivis Plausˇinaitis,‡ Ausˇra Teisˇerskiene˘ ,‡ and Vaclovas Lisauskas§ Department of Bioelectrochemistry and Biospectroscopy, Institute of Biochemistry, Mokslininku 12, LT-08662 Vilnius, Department of Physical Chemistry, Faculty of Chemistry, Vilnius UniVersity, Naugarduko 24, LT-03225 Vilnius, and Semiconductor Physics Institute, A. Gosˇtauto 11, LT-01108 Vilnius, Lithuania ReceiVed NoVember 3, 2006. In Final Form: January 10, 2007 The kinetics of electrocatalytic oxidation of ascorbate was studied on a series of redox self-assembled monolayers (SAMs) of the general formula Fc(CH2)4COO(CH2)nSH as electron-transfer mediators, where Fc is the ferrocenyl group and n ) 3, 6, 9, and 11. We show that the rate of electron transfer from ascorbate to the surface-confined Fc+ decreases with increasing n. The rationale for the dependence of the rate of electrocatalytic activity and n, in the presence of ClO4, is obtained from Fourier-transform surface-enhanced Raman spectroscopy (FT-SERS), cyclic voltammetry, and electrochemical quartz crystal microbalance (EQCM) data. In particular, FT-SERS shows decreasing amounts of surface-bound ClO4- upon oxidation of the ferrocene with decreasing n, while EQCM data show the effective electrode mass increase was consistently higher on the shorter chain SAMs. This mass increase is likely due to increasing ferricinium cation hydration. As n decreases, the SAMs become less ordered (FT-SERS data), as is widely known from previous literature. Disorder favors water penetration into the SAM, which, in turn, increases the hydration of the Fc+ (EQCM data). Increased hydration of the Fc+ impedes the formation of Fc+-ClO4- ion pairs (EQCM and FT-SERS data), which, consequently, accelerates the electrocatalytic electron transfer from the solution-dissolved ascorbate.

1. Introduction Ferrocene (Fc)-terminated self-assembled monolayers (SAMs) are extensively used in studies of electron-transfer processes.1 The chemical composition, structure, and length of the molecules linked to the surface via a sulfur-gold bond as well as the dielectric environment on the surface are all crucial factors determining the dynamics of the electron transfer between the ferrocene headgroups and the electrode surface.2 Less is known about the processes of the electron transfer between the surfaceconfined ferrocene SAMs and redox species located outside the Helmholtz layer.3 This particular step of the electrochemical process is of principal importance for understanding the mechanism of electron transfer between biological substrates, including redox proteins and enzymes, and surface-immobilized redox mediators.4 Recently we have reported an unusual ion-gating effect on the electron transfer between the surface-confined 9-mercaptononyl 5′-ferrocenylpentanoate and ascorbic acid.5,6 In particular, we * To whom correspondence should be addressed. Phone: +370-5-2729186. Fax: +370-5-2729196. E-mail: [email protected]. † Institute of Biochemistry. ‡ Vilnius University. § Semiconductor Physics Institute. (1) Chidsey, S. E. D. Science 1991, 251, 919. (2) Rowe, G. K.; Creager, S. E. Langmuir 1991, 7, 2307. Creager, S. E.; Rowe, G. K. Anal. Chim. Acta 1991, 246, 233. Creager, S. E.; Hockett, L. A.; Rowe, G. K. Langmuir 1992, 8, 854. Rowe, G. K; Creager, S. E. J. Phys. Chem. 1994, 98, 5500. (3) Alleman, K. S.; Weber, K.; Creager, S. E. J. Phys. Chem. 1996, 100, 17050. (4) Nanjo, S.; Ishii, K.; Ueki, T.; Imabayashi, S-i.; Watanabe, M.; Kano, K. Anal. Chem. 2005, 77, 4142. (5) Kazakevicˇiene˘ , B.; Valincius, G.; Niaura, G.; Talaikyte˘ , Z.; Kazˇeme˘ kaite˘ , M.; Razumas, V. J. Phys. Chem. B 2003, 107, 6661. (6) Valincius, G.; Niaura, G.; Kazakevicˇiene˘ , B.; Talaikyte˘ , Z.; Kazˇeme˘ kaite˘ , M.; Butkus, E.; Razumas, V. Langmuir 2004, 20, 6631.

found that the electron transfer from ascorbate to the surfacebound Fc group is inhibited by “hydrophobic” (low hydration energy) anions such as perchlorate and accelerated by “hydrophilic” (high hydration energy) ions such as fluoride.6 Supported by spectroelectrochemical data, we attributed this effect to possible structural changes in the Fc-SAM and suggested that strong ion pairing occurs between the oxidized ferricinium (Fc+) monolayers and hydrophobic anions (e.g., PF6-, ClO4-, BF4-), forming a rigid two-dimensional ionic layer that impedes the electrocatalytic electron transfer. On the other hand, for systems in which the anions are hydrophilic, ion pairs with Fc+ in the redox SAMs do not form; consequently, SAMs are structurally less ordered and exhibit enhanced electron-mediatory properties.6 It is well-known that shorter alkyl chain SAMs are less ordered relative to longer chain analogues.7,8 Consequently, by changing the length of the carbon chain one can deliberately control the order of the system, making it possible to further validate the above hypothesis that the electrocatalytic electron-transfer rate is modulated by the order of the surface-bound electron-transfermediating catalyst. To this end, we synthesized a series of ω-mercaptoalkyl 5′-ferrocenylpentanoates of the general formula Fc(CH2)4COO(CH2)nSH, in which n ) 3, 6, 9, and 11, and prepared their corresponding SAMs on gold, henceforth, for the sake of brevity, denoted as FcC4COOCnSH. Using cyclic voltammetry, Fourier-transform surface-enhanced Raman spectroscopy (FT-SERS), and electrochemical quartz crystal microbalance (EQCM) techniques, we herein present evidence of a correlation between the SAM’s order and the rate of electrocatalytic electron transfer. (7) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559. (8) Benitez, G.; Vericat, C.; Tanco, S.; Remes Lenicov, F.; Castez, M. F.; Vela, M. E.; Salvarezza, R. C. Langmuir 2004, 20, 5030.

10.1021/la0632169 CCC: $37.00 © 2007 American Chemical Society Published on Web 03/23/2007

KazakeVicˇ iene˘ et al.

4966 Langmuir, Vol. 23, No. 9, 2007 Scheme 1

2. Experimental Section 2.1. Synthesis of Ferrocene-Terminated Mercapto Compounds. Ferrocene derivatives containing an ω-mercaptoalkyl pentanoate group were synthesized from 5-ferrocenylpentanoic acid, obtained as reported earlier,9 and ω-bromoalkanols (Scheme 1). Esterification of acid 1 with ω-bromoalkanols afforded compounds 2a-d, and the Fc-mercapto compounds 3a-d were obtained by the conversion of the corresponding bromides into terminal mercapto derivatives with thiocarbamide. This transformation was accomplished using the elaborated reaction conditions for 9-mercaptononyl 5′-ferrocenylpentanoate (3c),10 since the usual procedures for the decomposition of isothiouronium salts employ ethanol and potassium hydroxide in ethanol and these reaction conditions could not be used in this case owing to possible hydrolysis of the ester group. Compounds 3a-d were obtained in good yields by using thiocarbamide in acetone and performing the decomposition of the isothiouronium salt in a water/chloroform heterogeneous phase containing Na2S2O5. 2.1.1. General Procedure for the Synthesis of ω-Bromoalkyl 5′Ferrocenylpentanoates 2a-d. A mixture of 5-ferrocenylpentanoic acid (1),10 ω-bromoalkanol (1.0 mmol), and p-toluenesulfonic acid (1.0 mmol) in benzene (30 mL) was refluxed for 6 h in a flask with a Dean-Stark head under Ar. The reaction mixture was washed with water, the organic phase was dried over MgSO4, and after evaporation of solvent under reduced pressure, the residue was purified by silica gel column chromatography (eluent CH2Cl2). Data for 3-bromopropyl 5′-ferrocenylpentanoate (2a): yellow oil; yield 51%; 1H NMR δ (ppm) 1.45-1.85 (m, 4H, 2CH2), 2.02.55 (m, 6H, CH2, CH2CO, FcCH2), 3.47 (t, J ) 6.2 Hz, 2H, CH2Br), 3.9-4.2 (m, 11H, OCH2, Fc ring); IR (cm-1) 1731 (CdO). Data for 6-bromohexyl 5′-ferrocenylpentanoate (2b): yellow oil; yield 48%; 1H NMR δ (ppm) 1.1-1.9 (m, 12H, 2CH2, 4CH2), 2.22.5 (m, 4H, CH2CO, FcCH2), 3.4 (t, J ) 6.2 Hz, 2H, CH2Br), 3.94.2 (m, 11H, OCH2, Fc ring), IR (cm-1) 1731 (CdO). 9-Bromononyl 5′-ferrocenylpentanoate (2c) was obtained as reported previously.10 Data for 11-bromoundecyl 5′-ferrocenylpentanoate (2d): yellow oil; yield 46%; 1H NMR δ (ppm) 1.1-2.0 (m, 22H, 2CH2, 9CH2), 2.2-2.5 (m, 4H, CH2CO, FcCH2), 3.4 (t, J ) 6.2 Hz, 2H, CH2Br), 3.9-4.2 (m, 11H, OCH2, Fc ring); IR (cm-1) 1731 (CdO). 2.1.2. General Procedure for the Synthesis of 3a-d. ω-Bromoesters 2a-d (1.0 mmol) and thiocarbamide (4.0 mmol) in dry acetone (30 mL) were refluxed for about 70 h under Ar. The solvent was evaporated in vacuo, and the crude isothiouronium salt and excess of thiocarbamide were poured into the mixture of CHCl3 (40 mL) and water (20 mL). After addition of solid Na2S2O5 (2.0 mmol), the reaction mixture was vigorously stirred and refluxed for 2.5-3 h under Ar. The organic phase was washed with water (2×) and dried over MgSO4, and after evaporation in vacuo the residue was purified by silica gel column chromatography (eluent CH2Cl2/C6H14, 1:1). Data for 3-mercaptopropyl 5′-ferrocenylpentanoate (3a): orange oil; yield 52%; 1H NMR δ (ppm) 1.2-1.7 (m, 5H, 2CH2, SH), 1.8-2.8 (m, 8H, CH2S, CH2, CH2CO, FcCH2), 3.9-4.2 (m, 11H, OCH2, Fc ring); IR (cm-1) 1734 (CdO), 2567 (SH). (9) Rinehart, K. L., Jr.; Curby, R. J., Jr.; Gustafson, D. H. J. Am. Chem. Soc. 1962, 84, 3263. (10) Kazˇeme˘ kaite˘ , M.; Bulovas, A.; Smirnovas, V.; Niaura, G.; Butkus, E.; Razumas, V. Tetrahedron Lett. 2001, 42, 7691.

Data for 6-mercaptohexyl 5′-ferrocenylpentanoate (3b): orange oil; yield 48%; 1H NMR δ (ppm) 1.1-1.9 (m, 13H, 2CH2, 4CH2, SH), 2.2-2.7 (m, 6H, CH2S, CH2CO, FcCH2), 3.9-4.2 (m, 11H, OCH2, Fc ring); IR (cm-1) 1734 (CdO), 2569 (SH). 9-Mercaptononyl 5′-ferrocenylpentanoate (3c) was obtained as reported previously.10 Data for 11-mercaptoundecyl 5′-ferrocenylpentanoate (3d): orange oil; yield 45%; 1H NMR δ (ppm) 1.1-2.0 (m, 23H, 2CH2, 9CH2, SH), 2.2-2.7 (m, 6H, CH2S, CH2CO, FcCH2), 3.9-4.2 (m, 11H, OCH2, Fc ring); IR (cm-1) 1734 (CdO), 2570 (SH). 2.2. Other Chemicals. Sodium salts were used throughout the work. Most salts, inorganic acids, alkalies, and ascorbic acid were ACS reagent grade and were purchased from Sigma-Aldrich Chemie GmbH (Germany). The Millipore purified (18.2 MΩ·cm) water was used throughout the work. 2.3. Electrode Preparation. For the electrochemical measurements, SAMs were formed on polycrystalline gold electrodes (BAS, West Lafayette, IN) with a geometric surface area of 0.02 cm2. Before each experiment, the electrodes were polished on a 0.05 µm alumina (Struers, Denmark) slurry and then (i) sonicated for 10 min in a 1:1 mixture of water and ethanol, (ii) etched in aqua regia for 2 min and sonicated for 10 min in water, and (iii) potentiostatically scanned (100 mV/s) for about 10 min in 1.0 M sulfuric acid solution in the potential range between 0 and 1.5 V (hereafter, the potentials are relative to that of a saturated sodium chloride calomel reference electrode, SSCE). After being rinsed with water and ethanol, the electrodes were transferred into a 0.1 mM solution of one of the thiols 3a-d in ethanol (95%) and incubated there for 12-15 h. For SERS experiments, the electrodes were prepared in the same way as for the electrochemical measurements except for the electrochemical roughening step, which was carried out in accordance with the procedure described previously.5,11 All basic features of the voltammograms and electrocatalytic effects were qualitatively the same on all the electrodes. Gold layers for the EQCM experiments were magnetron-sputtered on quartz disks (15 mm diameter, 0.3 mm thick, 5 MHz mainresonance frequency crystals, Valley Fisher) using a high-vacuum station (VUP-5M, Ukraine). The thickness of the gold layer was 200 ( 20 nm. A thin (approximately 20 nm) chromium adhesion underlayer was magnetron-sputtered onto quartz substrates prior to gold deposition. The basic vacuum was at least 2 × 10-4 Pa while the argon pressure was kept at about 1 Pa during the sputtering process. The gold films were stable and sustained multiple oxidationreduction cycles from 0 to 1.5 V in 1 M sulfuric acid solution, which were carried out prior to thiol deposition (vide ultra). 2.4. Electrochemical Measurements. The electrochemical part of the work was carried out on an EG&G Versastat computerized potentiostat system (Princeton Applied Research, Princeton, NJ). A three-electrode conventional cell was used for the measurements. A coiled platinum wire served as an auxiliary electrode. As noted in section 2.3, SSCE was used as a reference with potential E ) 239 mV vs a standard hydrogen electrode. The real surface area of the working electrode was estimated from the integration of the Au surface oxidation charge. The integration was carried out in the potential range from the onset of the oxidation until the “Burstein minimum” potential.12 In 1 M sulfuric acid, this range spanned from (11) Gao, P.; Gosztola, D.; Leung, L.-W. H.; Weaver, M. J. J. Electroanal. Chem. 1987, 233, 211. (12) Burstein, R. K. Elektrokhimiya 1967, 3, 349.

Mediated Oxidation of Ascorbic Acid on SAMs

Langmuir, Vol. 23, No. 9, 2007 4967

Figure 1. Cyclic voltammograms of the FcC4COOCnSH SAMs on gold in 0.01 M sodium phosphate buffer (pH 7.0, 25 °C) containing 0.1 M NaClO4. Sweep rate 100 mV s-1. about 0.9 to 1.45 V. The measurements were carried out at 25 °C, under an argon (AGA, Sweden) flow. 2.5. Spectroscopic Experiments. FT-SERS measurements were carried out with an FT-Raman spectrometer (Perkin-Elmer, model Spectrum GX) equipped with an InGaAs detector operating at room temperature. An air-cooled diode-pumped Nd:YAG laser provided the excitation with an emission wavelength of 1064 nm. The laser beam was focused to a spot of ca. 1 mm2 area, and the laser power at the sample was set to 300 mW. To reduce photo- and/or thermoeffects, the spectroelectrochemical cell, together with the working electrode, was moved linearly with respect to the laser beam (ca. 20 mm/s).13 The experiments were carried out in a 180° geometry. The spectral resolution was set at 4 cm-1, and the wavenumber increment per data point was 1 cm-1. Spectra were acquired by accumulation of 200-600 scans. None of the spectra presented have been smoothed. The Raman spectrum of FcC4COOC9SH dissolved in CCl4 was recorded with a dispersive f/5.6 spectrometer in a 90° geometry. Excitation was provided by 632.8 nm radiation of a He-Ne laser (Spectra-Physics). The power at the sample was 15 mW. A cooled photomultiplier tube and photon counting system were used for signal detection. NMR spectra were recorded on a Varian Unity Inova 300 spectrometer (300 MHz) with tetramethylsilane (TMS) as the internal standard in CDCl3. IR spectra of liquid samples spread on a KBr window were recorded on an FT-IR spectrometer (Perkin-Elmer, model Spectrum GX) equipped with a DGTS detector. All spectroscopic measurements were performed at 25 °C. 2.6. Electrochemical Quartz Crystal Microbalance Experiments. EQCM measurements were carried out in a 5 mL sealed poly(tetrafluoroethylene) cell. The quartz-supported working electrode was placed in a holder between two silicone washers and fixed vertically. The sensitivity constant of the EQCM was calculated from the Sauerbray equation. It was found to be 12 ng/Hz. A crystal was connected to an oscillator circuit, which operated at the parallel frequency of the quartz crystal.14 A CAMAC15 interface (13) Niaura, G.; Gaigalas, A. K.; Vilker, V. L. J. Raman Spectrosc. 1997, 28, 1009. (14) Daujotis, V.; Raudonis, R.; Kubilius, V.; Jasaitis, D. Russ. J. Electrochem. 1993, 29, 1005.

Table 1. Electrochemical Parameters (Mean Values ( SE) of the Homologous Series of SAMs of Different Mecaptoalkyl Ferrocenylpentanoatesa SAM-forming substance

E°′, mV vs SSCE

fwhm, mV

Γmax × 1010, mol/cm2

FcC4COOC3SH FcC4COOC6SH FcC4COOC9SHb FcC4COOC11SH

343 ( 6 317 ( 3 316 ( 5 324 ( 3

180 ( 10 135 ( 7 96 ( 8 105 ( 6

3.62 ( 0.24 4.28 ( 0.22 4.46 ( 0.20 4.45 ( 0.24

a Measurements were carried out in 0.1 M NaClO , 0.01 M sodium 4 phosphate buffer (pH 7.0, 25 °C). b Data from ref 6.

provided a link among the PC, potentiostat, and customized frequency counter. CAMAC modular 12 bit A/D (FK4411 Vikama, Lithuania) and 16 bit D/A converters (FK70 Vikama, Lithuania) sampled the current and controlled the working electrode potential program. The software created in the Delphi programming environment drove all instruments. In the EQCM experiments, 22 frequency data points per second were made with an accuracy of (0.01 Hz.

3. Results and Discussion 3.1. Electrochemical Properties of the FcC4COOCnSH SAMs. Figure 1 displays cyclic voltammetry curves of the n ) 3, 6, 9, and 11 ferrocene-terminated SAMs. The electrochemical parameters obtained from these curves are summarized in Table 1. Several specific trends are visible. First, all SAMs exhibit almost the same formal redox potentials (E°′), except for the n ) 3 SAM, which demonstrates consistently higher values of this parameter. Second, the full width at half-maximum of the current peaks (fwhm), which for an ideal Nernstian electron transfer should be 90.6 mV at 25 °C, is dependent on the SAM chain length (Table 1, column 3). The fwhm decreases from 180 to ca. 100 mV with a chain length increase. The widening of the fwhm is indicative of the deviations of the surface-confined redox processes from ideality,16 which might include lateral repulsive (15) CAMAC, an acronym for computer-automated measurement and control, is an IEEE-583 standard, modular, high-performance, real-time data acquisition and control system.


4968 Langmuir, Vol. 23, No. 9, 2007

Table 2. Rate Constants (Mean Values ( SE) of the Electrocatalytic Electron Transfer from Ascorbic Acid to Ferrocene SAMs as a Function of the Chain Lengtha SAM-forming thiol

kcat, M-1 s-1

SAM-forming thiol

kcat, M-1 s-1

FcC4COOC3SH FcC4COOC6SH

1105 ( 114 553 ( 58

FcC4COOC9SHb FcC4COOC11SH

489 ( 59 493 ( 112

a Measured in 1 M NaClO4, 0.01 M sodium phosphate buffer (pH 7.0, 25 °C) containing 1 mM ascorbic acid. b Data from ref 6.

Figure 2. Background-subtracted voltammetry curves of electrocatalytic oxidation of ascorbate on different length FcC4COOCnSH SAMs in 0.01 M sodium phosphate buffer, containing 1 M NaClO4, (pH 7.0, 25 °C) and 1 mM ascorbic acid. Number of carbon atoms n: 1, 3; 2, 6; 3, 9; 4, 11. Sweep rate 100 mV s-1.

interaction between the surface redox groups,17 and/or monolayer heterogeneity that leads to broadening of the redox peaks. In a series of homologous SAMs, the strongest manifestation of these effects is observed for the FcC4COOC3SH monolayers. However, at this point, it is unclear whether the lateral repulsion or monolayer heterogeneity or both play the decisive role in the variation of the fwhm, E°′, and peak shape with the length of the SAM molecule. Table 1 indicates that the fwhm correlates not only with the formal redox potential of the SAMs but also with the maximum surface concentration (Γmax, calculated by integration of the anodic current peaks) of the oxidizible ferrocene groups. In particular, for the FcC4COOC3SH SAM, Γmax is approximately 20% lower than the values for FcC4COOC9SH and FcC4COOC11SH. A smaller Γmax in conjunction with a larger fwhm points to a higher degree of disorder and possibly electrostatic repulsion between Fc+ ions in the FcC4COOC3SH monolayer compared to that of SAMs formed by thiols with longer chains. 3.2. Electrooxidation of Ascorbate on SAMs of Different Lengths. Figure 2 displays the background-subtracted linearscan voltammograms of the electrocatalytic oxidation of ascorbate on different length SAMs. The S-shaped voltammograms are due to the Fc-mediated electron transfer through the monolayers (see the Supporting Information of refs 5 and 6). The current waves exhibit different heights, which suggest different electrontransfer rates18 on different length SAMs. Calculated rate constants of the electrocatalytic electron transfer from the ascorbate to Fc+ (kcat; mediated electron-transfer constants were calculated in accordance with Andreaux and Saveant;18 see also the Supporting Information of ref 6) demonstrate a tendency to increase as the length of the SAM molecule decreases (Table 2). The electron transfer from ascorbate to the surface-confined ferricinium group occurs more than 2 times faster on the FcC4COOC3SH SAM than on the FcC4COOC9SH and FcC4COOC11SH SAMs. Such an effect was not observed in the buffers containing hydrophilic anions (e.g., fluoride, data not shown) that exhibit a weak (16) Brown, A. P.; Anson, F. C. Anal. Chem. 1977, 49, 1589. (17) Laviron, E. J. Electroanal. Chem. 1979, 101, 19. (18) Andrieux, C. P.; Saveant, J. M. J. Electroanal. Chem. 1978, 93, 163.

Figure 3. Comparison of the FT-SERS spectra of n ) 3 and 9 FcC4COOCnSH SAMs on a gold electrode at E ) 0.6 V vs SSCE in 0.01 M sodium phosphate buffer, containing 0.1 M NaClO4 (pH 7.0, 25 °C). Spectra were normalized by the intensity of the stretching vibration of Fc rings at 1113 cm-1. The excitation wavelength is 1064 nm. The laser power at the sample is 300 mW.

propensity for ion pairing with Fc+. This hints that the modulation of the electron-transfer rate may be related to differences in the ion-pairing strength between the ferricinium and perchlorate on different length SAMs. To substantiate a possible link among the ion-pairing ability, structure, and electrocatalytic electrontransfer rates, we investigated our SAM systems with Fouriertransform surface-enhanced Raman spectroscopy and electrochemical quartz microbalance techniques. 3.3. Fourier Transform Surface-Enhanced Raman Spectra of the FcC4COOCnSH SAMs. In situ FT-SERS allowed us to obtain molecular level information on (i) the ion pairing between the terminal ferricinium cation and the ClO4- ion and (ii) the order and orientation of the monolayer. Figure 3 shows the FTSERS spectra of the n ) 3 and 9 SAMs in their oxidized state (E ) 0.6 V; see Figure 1) from 900 to 950 cm-1 and from 1050 to 1140 cm-1. These spectra are similar to those reported previously.6 The 932 cm-1 band is the symmetric stretching band of the ClO4- ion, and the 1113 cm-1 band is the symmetric vibration band of the Fc+ ring.5,6,19,20 The ratio of integrated (19) Nishiyama, K.; Ueba, A.; Tanoue, S.; Koga, T.; Taniguchi, I. Chem. Lett. 2000, 930. (20) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordinated Compounds; Wiley: New York, 1997.


Langmuir, Vol. 23, No. 9, 2007 4969

Table 3. Dependence of the Relative Integrated FT-SERS Intensities (AClO4-/AFc+; Mean Values ( SE) on the Length of the FcC4COOCnSH-Based SAMsa SAM-forming substance

AClO4-/AFc+

SAM-forming substance

AClO4-/AFc+


0.56 ( 0.06 0.63 ( 0.05


0.73 ( 0.15b 0.69 ( 0.06

a Measurements performed in 0.01 M sodium phosphate buffer (pH 7.0, 25 °C) containing 0.1 M NaClO4. The electrode Potential is 0.6 V vs SSCE. b From ref 6.

Figure 4. FT-SERS spectra in the C-H stretching frequency region of the FcC4COOC9SH SAM on gold at 0.0 (a) and 0.6 (b) V vs SSCE recorded in a perchlorate-dominant solution. The difference spectrum (c) is also shown. The solution composition is 0.01 M sodium phosphate buffer and 0.1 M NaClO4 (pH 7.0, 25 °C). The excitation wavelength is 1064 nm. The laser power at the sample is 300 mW.

intensities AClO4-/AFc+ is a measure of the amount of ClO4- ionpaired with the Fc+ ring.5,6 Simple inspection of the 932 cm-1 bands (Figure 3) indicates that the relative intensity of ClO4- is less for the n ) 3 SAM than that found for the n ) 9 SAM. The measured AClO4-/AFc+ ratios (Table 3), obtained by averaging the results from 8-12 experiments, clearly show that AClO4-/AFc+ increases as n increases, confirming the electrochemical data (fwhm and E°′ values) indicating that as the SAM length increases ion pairing with Fc+ increases. FT-SERS data for the FcC4COOC9SH SAMs in the 27803220 cm-1 region reveals the C-H stretching vibrations of the methylene groups and the Fc rings in the reduced (Fc, spectrum a, E ) 0.00 V) and the oxidized (Fc+, spectrum b, E ) +0.60 V) states in the presence of ClO4- (Figure 4). Within this region, the Fc and Fc+ ring vibrational modes and the methylene vibrational modes are found from 3100 to 3150 cm-1 and from 2800 to 3000 cm-1, respectively. As can be seen in Figure 4a, the bands for the reduced Fc ring, ν(C-H)Fc, centered at 3105 cm-1, undergo a significant shift to 3124 cm-1 (Figure 4b). This 19-wavenumber shift is also seen as the negative peak in the difference spectrum (Figure 4c). Spectra a-c of Figure 4 also show that the methylene vibrational modes in the 2840-2920 cm-1 spectral region are

dependent on the applied voltage. First, the intensities of both the symmetric and asymmetric (labeled as νs(CH2) and νas(CH2), respectively) modes decrease at 0.60 V. This observation indicates an orientational transformation within the monolayer. According to the SERS selection rules,21 the relative band intensities depend, among other factors, on the orientation of the molecular moieties with respect to the surface. The most enhanced bands are related to the modes containing polarizability tensor components perpendicular to the surface. For the idealized all-trans conformation of alkyl chains, the intensity of the νas(CH2) mode should decrease as the orientation of the alkyl chain approaches the surface normal. The band intensity decrease in Figure 4b demonstrates that, upon oxidation of the terminal Fc group, the alkyl chains adopt a more perpendicular orientation with respect to the electrode surface as compared to that in the monolayer in the reduced state. Similar potential-dependent intensity changes were observed by in situ Fourier-transform infrared reflectionabsorption spectroscopy (FT-IRRAS) for the FcC11H22SH monolayer22,23 and were correlated with a more perpendicular orientation of the alkyl chains upon oxidation of Fc. However, infrared spectroelectrochemical studied of other Fc-terminated monolayers, namely, FcCOOC11H22SH24 and FcCOCnHn+2SH with n e 9,25 did not reveal potential-induced intensity changes for the methylene stretching modes. In this paper, we report for the first time in situ FT-SERS observation of potential-induced intensity changes of νs(CH2) and νas(CH2) modes for the FcC4COOCnSH (n ) 3, 6, 9, 11) SAMs. Figure 4b also shows a potential-dependent shift in the position of the νas(CH2) mode from 2913 (spectrum a) to 2916 cm-1. The shift to higher wavenumbers is also evidenced in the difference spectrum (Figure 4c), where a positive band is expressed at 2907 cm-1. It is well documented that the band positions of the symmetric and asymmetric stretching modes of the CH2 group depend on the conformational order and interchain coupling between the polymethylene chains, decreasing with increasing order and coupling.7,26,27 Thus, the increased νas(CH2) frequency upon oxidation of the Fc groups is indicative of induced conformational disorder and interchain decoupling. Figure 5 shows the dependence of the νas(CH2) mode frequency on the length of the coupled thiol molecules at E ) 0.00 and 0.60 V for the entire series. A frequency decrease is evident for both oxidation states as the length of the SAM molecule increases. This is consistent with the general view that monolayers of longer alkyl chain thiols are more ordered.26 In addition, Figure 5 shows that the magnitude of the potential-induced shift in the band position is noticeably higher for the short-chain SAMs (ca. 5 cm-1 for FcC4COOC3SH) as compared to the long-chain SAMs (ca. 3 cm-1 for FcC4COOC11SH). This could possibly be attributed to the ordering stimulus of ion pairing acting along with the van der Waals interaction, which is stronger for the long-chain SAMs (Table 3). Finally, because the Fc+ monolayers exhibit more disorder and poorer coupling between the alkyl chains, one can expect that these less ordered films will allow penetration of water into the SAM, concomitantly increasing the mean distance between (21) Creighton, J. A., Clark, R. J. H., Hester, R. E., Eds. Spectroscopy of Surfaces; John Wiley & Sons Ltd.: New York, 1988; p 37. (22) Ye, S.; Sato, Y.; Uosaki, K. Langmuir 1997, 13, 3157. (23) Ye, S.; Haba, T.; Sato, Y.; Shimazu, K.; Uosaki, K. Phys. Chem. Chem. Phys. 1999, 1, 3653. (24) Popenoe, D. D.; Deinhammer, R. S.; Porter, M. D. Langmuir 1992, 8, 2521. (25) Viana, A. S.; Jones, A. H.; Abrantes, L. M.; Kalaji, M. J. Electroanal. Chem. 2001, 500, 290. (26) Bryant, M. A.; Pemberton, J. E. J. Am. Chem. Soc. 1991, 113, 8284. (27) Orendorff, C. J.; Ducey, M. W., Jr.; Pemberton, J. E. J. Phys. Chem. A 2002, 106, 6991.


4970 Langmuir, Vol. 23, No. 9, 2007

Figure 5. Dependence of the νas(CH2) mode frequency on the total number of chain atoms (6 + n) in FcC4COOCnSH monolayers on gold at 0.6 (b) and 0.0 (O) V vs SSCE. Measurements were performed in 0.01 M sodium phosphate buffer (pH 7.0, 25 °C) containing 0.1 M NaClO4. The frequencies of the νas(CH2) mode were determined after decomposition of the spectral contour in the frequency range of 2780-3220 cm-1 into the mixed Gausian-Lorentzian components.

the counterions. Unfortunately, FT-SERS is not able to directly detect the presence of water in the SAMs because of the low Raman scattering cross-section of water. However, we obtained indirect spectroscopic evidence for water penetration into the monolayer, which was then supported by the EQCM technique (vide infra). Let us consider the CdO stretching vibration band in the 1650-1820 cm-1 spectral region (Figure 6) recorded for the FcC4COOC9SH-based SAM. The position of the CdO band for this SAM (1732 cm-1) was found to be very close to that of the CdO vibration of FcC4COOC9SH dissolved in CCl4 (1731 cm-1), demonstrating that, in the reduced state, the carbonyl group buried in the SAM is not involved in hydrogen-bonding interaction. However, oxidation of the terminal Fc group at +0.60 V results in a noticeable red shift of the CdO frequency from 1732 to 1727 cm-1 (Figure 6c). It is well documented that the carbonyl stretching frequency of an ester group decreases upon formation of a hydrogen bond, due to a decrease of the double bond character in the CdO group.28 Thus, these FT-SERS data suggest that oxidation of the terminal Fc group induces alkyl chain disordering and the ability of some of the water molecules to penetrate within the monolayer and participate in hydrogen-bonding interaction with the CdO group. To the best of our knowledge, this is the first direct FT-SERS observation of such a type of interaction in SAMs. 3.4. Electrochemical Quartz Crystal Microbalance Study. The EQCM technique is sensitive to the mass change at the solid electrode surface and can be utilized to quantify water penetration into the Fc-terminated monolayers. Shimatzu et al.29-31 have (28) Socrates, G. Infrared and Raman Characteristic Group Frequencies, 3rd ed.; John Wiley & Sons, Ltd.: Chichester, U.K., 2004.

Figure 6. Comparison of CdO stretching bands of the FcC4COOC9SH ester group in CCl4 (a) and the FcC4COOC9SH SAM on gold electrode at 0.0 (b) and 0.6 (c) V vs SSCE. For the FT-SERS spectra, measurements were carried out in 0.01 M sodium phosphate buffer containing 0.1 M NaClO4 (pH 7.0, 25 °C). Spectrum a was recorded with a dispersive spectrometer using 632.8 nm excitation, while spectra b and c were recorded with an FT-Raman spectrometer using 1064 nm excitation. Table 4. EQCM Mass Increase per Mole of Oxidized Ferrocene (Mean Values ( SE) and the Effective Number of Water Molecules Coadsorbed with Perchlorate upon Oxidation of FcC4COOCnSH (n ) 3, 6, 9, 11) SAMsa SAM-forming substance

mass increase per mole of oxidized ferrocene, g/mol

effective no. of water molecules coadsorbed with perchlorate

FcC4COOC3SH FcC4COOC6SH FcC4COOC9SH FcC4COOC11SH

144 ( 3 121 ( 2 115 ( 1 114 ( 2

2.4 1.2 0.8 0.8

a Measurements were carried out in 0.01 M sodium phosphate buffer (pH 7.0, 21 °C) containing 0.1 M NaClO4. The electrode potential is 0.6 V vs SSCE.

demonstrated that, in perchlorate-dominant buffers, the effective mass increase is close to 100 g per mole of Fc+. However, the authors pointed out that the theoretical mass increase of 100 g/mol is observed only on well-packed, ordered monolayers, while EQCM responses above 100 g/faraday are indicative of packing distortions and the formation of looser monolayer structures.30 Using the EQCM technique, we estimated electrode mass changes during the oxidation of the Fc moiety of different length SAMs. As the data in Table 4 indicate, in all cases the electrode mass increase was more than 100 g/faraday, suggesting that simultaneously with perchlorate ions water molecules are attracted to the monolayer upon Fc oxidation. To assess the effective number of water molecules penetrating the SAMs, the perchlorate anion mass, 100 g/mol, was subtracted from the total EQCM mass change and the difference was divided by 18 (molecular (29) Shimatzu, K.; Ye, S.; Sato, Y.; Uosaki, K. J. Electroanal. Chem. 1994, 375, 409. (30) Kawaguchi, T.; Tada, K.; Shimatzu, K. J. Electroanal. Chem. 2003, 543, 41. (31) Shimatzu, K.; Yagi, I.; Sato, Y.; Uosaki, K. Langmuir 1992, 8, 1385.


weight of water). These data are presented in Table 4. In the sequence from FcC4COOC3SH to FcC4COOC11SH, a decrease of the effective number of coadsorbed water molecules from 2.4 to 0.8 per oxidized Fc group was observed. This result strongly supports the idea that the shortest SAMs in the series are more disordered and hydrated around the Fc group and down the polymethylene chain. The hydration inhibits ion pairing with ClO4-, thus ensuring the highest electron-transfer rates from the ascorbate to the Fc group in this series of SAMs.

4. Conclusions The effect of ion pairing on the electron-transfer rate in solution32 and on electrodes33 is known; however, the experimental evidence of this phenomenon is not abundant. Electrochemical quartz microbalance, cyclic voltammogram, and FTSERS data in this study show that, in the presence of ion-pairing anions such as ClO4-, the rate of catalytic electron transfer from ascorbate, the solution-dissolved electron donor, to SAMs with the general formula Fc(CH2)4COO(CH2)nSH, where Fc+ is an (32) Marcus, R. A. J. Phys. Chem. B 1998, 102, 10071. (33) Saveant, J.-M. J. Phys. Chem. B 2001, 105, 8995.

Langmuir, Vol. 23, No. 9, 2007 4971

oxidized surface-bound ferricinium group (electron acceptor), changes inversely with the length of the tethering polymethylene chain (CH2)n. Longer (CH2)n SAMs are more ordered and, importantly, inhibit the penetration of water into the SAM. In an environment of decreased hydration, lower dielectric permittivity favors stronger ionic interactions between ions of opposite charge, resulting in lower free energies of the Fc+ groups, which, in turn, may cause a higher activation energy barrier and lower rates for electrocatalytic electron transfer. Alternatively, the increased electron-transfer rate to the shorter Fc-SAMs may be the result of an increase in the ascorbate surface concentration due to the increased surface potential created by the unbalanced charges of Fc+ in the less ordered, hydrated films. Regardless of which of these factors dominates, or whether they act simultaneously, the modulation of the electron-transfer rate by the length of the polymethylene chain provides additional support to our earlier conclusion5,6 that factors diminishing the ion-pairing ability of surface-bound electron acceptors tend to accelerate electron transfer from the solution-dissolved donor and vice versa. LA0632169

Mediated Oxidation of Ascorbic Acid on a Homologous Series of

Recommend Documents