Ruthenium Aminophenanthroline Metallopolymer Films

Sep 12, 2008 - The Biomedical Diagnostics Institute, National Centre for Sensor ... Thin electropolymerized films generate electrochemiluminescence (E...
0 downloads 0 Views 235KB Size
Langmuir 2008, 24, 11233-11238

11233

Ruthenium Aminophenanthroline Metallopolymer Films Electropolymerized from an Ionic Liquid: Deposition and Electrochemical and Photonic Properties Anita Venkatanarayanan, Anna-Maria Spehar-De´le`ze, Lynn Dennany, Yann Pellegrin, Tia E. Keyes, and Robert J. Forster* The Biomedical Diagnostics Institute, National Centre for Sensor Research, School of Chemical Sciences, Dublin City UniVersity, Dublin 9, Ireland ReceiVed April 10, 2008. ReVised Manuscript ReceiVed June 24, 2008 The oxidative electropolymerization of [Ru(aphen)3](PF6)2 from an ionic liquid, 1-butyl-2,3-dimethylimidazolium bis[(trifluoromethyl)sulfonyl]imide (BDMITFSI), is reported; aphen is 5-amino-1,10-phenanthroline. The deposition rate in the ionic liquid is more than an order of magnitude faster than in conventional solvents such as anhydrous acetonitrile and aqueous sulfuric acid. The UV-vis absorbance, Raman, and emission spectra of the films grown in ionic liquid, acetonitrile, and sulfuric acid suggest that the polymer formed does not depend on the solvent. However, scanning electron microscopy shows that the film morphologies differ significantly; e.g., films deposited from BDMITFSI have high surface roughness, while films produced in acetonitrile and sulfuric acid are relatively smooth. The rate of homogeneous charge transport through films grown in ionic liquids is (6.4 ( 1.2) × 10-9 cm2 s-1, which is approximately 2 orders of magnitude faster than that found for films deposited from acetonitrile. Thin electropolymerized films generate electrochemiluminescence (ECL) in the presence of tripropylamine as a coreactant. Films produced from sulfuric acid are very thin compared to the ones produced in BDMITFSI; however, they produce an ECL signal of similar intensity. The ECL responses of films produced in anhydrous acetonitrile are significantly less intense. The ECL intensity within the films is approximately 5-fold higher than when they are dissolved and measured in solution.

Introduction Films of functional polymers play pivotal roles in diverse areas ranging from advanced sensors to reactive coatings and displays.1,2 Incorporating transition-metal complexes within these coatings is desirable because of their redox, catalytic and optical properties.3-5 Moreover, the metal complex can be designed to simultaneously fulfill several roles. For example, complexes of the type [M(aphen)3]2+, where M is Ru, Os, Co, or Fe, have the potential to covalently bind an enzyme through peptide bond formation, shuttle electrons to the active site, and produce an interfacial polymer film through oxidative electropolymerization of the amino functionality. The resulting films are superior to those involving an electrostatically immobilized luminophore in a polymer such as Nafion where leaching from the film can occur over time. Moreover, electropolymerization has the advantage that it enables in situ production of thin and stable films of controlled thickness.6-9 Fundamental studies on this type of aminophenanthroline polymer formed from acetonitrile solutions have been previously reported for ruthenium,10 iron,11 and osmium12complexes. However, there are significant challenges to creating relatively thick films with surface coverage greater than about 5 × 10-8 * To whom correspondence should be addressed. (1) Higgins, S. J. Chem. Soc. ReV. 1997, 26, 247. (2) Hjelm, J.; Handel, R. W.; Hagfeldt, A.; Constable, E. C.; Housecroft, C. E.; Forster, R. J. Inorg. Chem. 2005, 44, 1073. (3) Basudam Adhikari, S. M. Prog. Polym. Sci. 2004, 29, 699. (4) Dennany, L.; Hogan, C. F.; Keyes, T. E.; Forster, R. J. Anal. Chem. 2006, 78, 1412. (5) Bertoncello, P.; Dennany, L.; Forster, R. J.; Unwin, P. R. Anal. Chem. 2007, 79, 7549. (6) Takada, K.; Naal, Z.; Abruna, H. D. Langmuir 2003, 19, 5402. (7) Oyama, N.; Anson, F. C. J. Electrochem. Soc. 1980, 127, 247. (8) Belanger, S.; Shikata, K. J. S.; Mudakha, A.; Hupp, J. T. Langmuir 1999, 15, 837. (9) Hjelm, J.; Handel, R. W.; Hagfeldt, A.; Constable, E. C.; Housecroft, C. E.; Forster, R. J. Electrochem. Commun. 2004, 6, 193.

mol cm-2 or approximately 0.5 µm thick, which might improve the limits of detection in analytical applications. When electropolymerization is carried out in acetonitrile, the deposition efficiency decreases as the film grows,8,10 but the origin of this behavior is not fully understood.6 In this paper, we report on electrodeposition of [Ru(aphen)3]2+ from an ionic liquid, 1-butyl2,3-dimethylimidazolium bis[(trifluoromethyl)sulfonyl]imide13 (BDMITFSI); aphen is 5-amino-1,10-phenanthroline. The rate of electrodeposition is considerably faster in the ionic liquid, and thicker films can be formed than in acetonitrile and sulfuric acid. Also, charge can be transferred relatively quickly through the coatings; e.g., the redox state of a 1 µm thick film can be switched within 500 ms. Scanning electron microscopy (SEM) reveals that the morphologies of films deposited in BDMITFSI differ significantly from those deposited in acetonitrile and sulfuric acid. When the polymer films are dissolved off the electrode and optically excited at 488 nm, they emit at approximately 610 nm, indicating that the aromaticity of the complexes is conserved upon electropolymerization. Moreover, electrodeposited films generate electrochemiluminescence (ECL) upon oxidation of the Ru2+ metal centers in the presence of tripropylamine (TPA) as a coreactant.

Experimental Section Materials and Reagents. All chemicals and solvents were purchased from Sigma-Aldrich and were of analytical grade. Aqueous solutions were prepared using Milli-Q water. The solvents used for spectroscopic measurements were of HPLC grade. [Ru(aphen)3]2+ (10) Ellis, C. D.; Margerum, L. D.; Murray, R. W.; Meyer, T. J. Inorg. Chem. 1983, 22, 1283. (11) Pickup, P. G.; Osteryoung, R. A. Inorg. Chem. 1984, 242707. (12) Bachas, L. G.; Cullen, L.; Hutchins, R. S.; Scott, D. L. J. Chem. Soc., Dalton Trans. 1997, 1571. (13) Fredlake, C. P.; Crosthwaite, J. M.; Hert, D. G.; Aki, SNV. K.; Brennecke, J. F. J. Chem. Eng. Data 2004, 49, 954.

10.1021/la8011316 CCC: $40.75  2008 American Chemical Society Published on Web 09/12/2008

11234 Langmuir, Vol. 24, No. 19, 2008 was prepared according to the procedure described in the literature.10 Ion-exchange HPLC revealed a single peak. The structure of the complex was confirmed using 1H and 13C NMR spectra recorded on a Bruker Advance 400 NMR spectrometer and the free induction decay (FID) profiles processed using the Topspin software package. Anal. Calcd for [Ru(aphen)3](PF6)2: C, 44.27; H, 2.79; N,12.91. Found: C, 44.11; H, 2.90; N, 12.78. The ionic liquid BDMITFSI was obtained from Solvent Innovation. Under drybox conditions, the potential window was from -2.0 to +2.0 V. However, under the ambient conditions used here for the electrodeposition, the ionic liquid absorbed small amounts of water from the atmosphere, reducing the available window to -1.8 to +1.8 V. Oxidative Electropolymerization. Films were deposited from 1 mM solutions of the complex dissolved in BDMITFSI, HPLCgrade anhydrous acetonitrile, or 0.5 M sulfuric acid, pH 0.3. The supporting electrolyte used in acetonitrile was 0.1 M tetrabutylammonium tetrafluoroborate (TBATFB). Activated alumina (0.05 g) was added to acetonitrile to ensure removal of trace water. All solutions were deaerated for 20 min by purging with nitrogen prior to electrodeposition. The surface coverages were estimated from the charge passed under slow sweep rate (ν < 10 mV s-1) in the deposition solvent in the absence of the metal complex. Instrumentation. Electrochemical measurements were carried out with a CH Instruments model 660 potentiostat, using a typical three-electrode cell configuration. Electrodeposition in BDMITFSI was performed using a 25 µm radius platinum microelectrode as a working electrode, while in anhydrous acetonitrile and sulfuric acid a 1.5 mm radius glassy carbon electrodes was used. An aqueous or nonaqueous Ag/AgCl electrode was used as a reference depending on the solvent, and in all cases a platinum flag was used as a counter electrode. The working electrodes used were polished with 0.3 and 0.05 µm alumina for 10 min, rinsed thoroughly with deionized water, and dried under a nitrogen stream. The platinum electrode was additionally electrochemically cleaned by cycling in 0.5 M sulfuric acid. All potentials are quoted versus a Ag/AgCl reference electrode, and all measurements were performed at room temperature. The working electrode for the EQCM measurements was an 8 MHz AT-cut quartz crystal with platinum electrodes. The surface of the electrode was masked to give an active area of 0.04 cm2. These crystals were calibrated by characterizing the electrical behavior of the resonator around its resonance frequency as described by Endres and co-workers.14 This calibration involves measuring the real part of the electrical admittance of the quartz crystal with a network analyzer. Calibration using electrodeposited metal indicates that a 1 Hz frequency change corresponded to a 1.2 ng mass change. For the ECL measurements, an Autolab PGSTAT 10 and an Oriel 70680 photomultiplier tube (PMT) equipped with a high-voltage power supply (Oriel, model 70705) which was used at a bias of -850 V and an amplifier/recorder (Oriel, model 70701) were utilized. During the experiments the cell was placed inside a specially constructed holder, which positioned the working electrode in a reproducible manner directly opposite the face of a fiber optic bundle, the other end of which was coupled to the PMT. The entire electrode assembly was contained inside a light-tight box. Absorbance spectra were recorded on a Shimadzu UV-vis-NIR 3100 spectrophotometer and emission spectra on a Perkin-Elmer LS50-B spectrophotometer or Varian Cary Eclipse fluorescence spectrophotometer using a 1 cm optical path length quartz cuvette. The fluorescence properties of the electropolymerized films were investigated with a Zeiss LSM 510 Meta laser module confocal microscope using an excitation wavelength of 488 nm of an argon ion laser and a 10× magnification objective. Resonance Raman spectra were recorded on a HORIBA JobinYvon Labram HR confocal Raman microscope. An argon ion laser (Coherent) was used to excite at 488 nm. The laser was focused onto the film using a 10× objective. (14) Schneider, O.; Bund, A.; Ispas, A.; Borissenko, N.; Zein El Abedin, S.; Endres, F. J. Phys. Chem. B 2005, 109, 7159–7168.

Venkatanarayanan et al.

Figure 1. Interfacial electropolymerization from a 1.0 mM solution of [Ru(aphen)3]2+ on a 25 µm radius platinum microelectrode at a scan rate of 0.1 V s-1. The deposition medium was an ionic liquid, BDMITFSI.

Figure 2. (A) EQCM frequency change recorded during electropolymerization of [Ru(aphen)3]2+. The deposition medium was BDMITFSI, and the scan rate was 0.1 V s-1. (B) Dependence of the Ru2+/3+ peak current at 0.1 V s-1 ([) and the change in frequency (b) on the scan number.

The morphology of the electropolymerized films was investigated using the Hitachi S-3000N scanning electron microscope. Image analysis was carried out using Image J version 1.37d image analysis software.

Results and Discussion Electropolymerization in an Ionic Liquid. Figure 1 illustrates the effect of repeated voltammetric cycling of a 25 µm radius platinum microelectrode immersed in a 1.0 mM solution of [Ru(aphen)3]2+ dissolved in BDMITFSI. Consistent with previous reports10,11 for electropolymerization in acetonitrile, two oxidation processes are observed during the first scan. These correspond to irreversible oxidation of the 5-amino-1,10-phenanthroline ligand at approximately +1.13 V and the electrochemically reversible Ru2+/3+ couple at +1.250 V. Significantly, the peak

[Ru(aphen)3](PF6)2 Films from an Ionic Liquid

Langmuir, Vol. 24, No. 19, 2008 11235

Figure 3. Scan rate dependence of the voltammetric response for a thin film of [Ru(aphen)3]2+ (Γ ) (9.8 ( 2.6) × 10-8 mol cm-2) electropolymerized on a 25 µm radius platinum microelectrode in BDMITFSI. The scan rates are from top to bottom 100, 50, 20, and 10 mV s-1. The inset shows the peak current dependence vs the scan rate.

current associated with the Ru2+/3+ couple increases with an increasing number of voltammetric scans; however, the rate of growth rapidly decreases between successive scans. This behavior is consistent with the formation of a redox-active electropolymerized film on the electrode surface but indicates that the growth rate decreases between successive scans. One possibility for this decreasing electropolymerization efficiency is that the rate of homogeneous charge transport through the film decreases as the film thickness increases. Under these conditions, oxidation of the ruthenium complex in solution becomes progressively less efficient and the rate of film deposition decreases. We address the issue of the rate of charge transport through the films in a later section. To probe the rate of film deposition, EQCM was used. Figure 2A shows that the resonance frequency decreases for potentials more positive than +1.15 V. This decrease in frequency indicates an increased mass on the crystal, which is consistent with the voltammetric response and suggests that electropolymerization occurs when the aminophenanthroline ligand is oxidized. Figure 2B shows that the frequency decreases and the peak current increases with an increasing number of scans in a nonlinear manner. The mass change at the quartz crystal can be determined from the frequency change using the Sauerbrey equation:15,16

∆m ) (-1/2)(f01⁄2)(∆f)A(kF)1⁄2

(1)

where ∆f is the measured frequency change, A is the area of the platinum disk onto which polymer is deposited, F is the density of the crystal, k is the shear modulus of the crystal, and f0 is the oscillation frequency of the crystal. Calibration using electrodeposited metal indicates that a 1 Hz frequency change corresponded to a 1.2 ng mass change. Thus, the total mass deposited during seven voltammetric cycles is 2880 ( 350 ng, which corresponds to a surface coverage of (1.0 ( 0.1) × 10-7 mol cm-2. Significantly, this surface coverage is equivalent to approximately 1000 monolayers deposited within 7 cycles at 100 mV s-1. Figure 3 shows the voltammetric response of the electrodeposited film in a blank BDMITFSI solution at scan rates of 10-100 mV s-1. The inset shows the dependence of the voltammetric response on these scan rates. It can be clearly seen that the peak current, ip, is proportional to the scan rate, which

indicates that the response is under finite diffusion control.17 However, the response deviates from that expected for an electrochemically reversible reaction in that the peak-to-peak separation, ∆Ep, is not zero as theoretically predicted, but is typically between 30 and 45 mV. We have reported similar responses previously18-20 and interpreted them in terms of Feldberg’s unusual quasi-reversibility model21 involving Nshaped free energy curves in which some kinetic processes are slow compared to the voltammetric scan even at scan rates on the order of 1 mV s-1. The charge passed is 0.19 ( 0.05 µC, corresponding to a surface coverage of (9.8 ( 2.6) × 10-8 mol cm-2. This value is experimentally indistinguishable from that found in the in situ EQCM experiments, indicating that essentially the entire polymer formed ends up as an electroactive coating on the electrode. For 100 mV s-1 < ν < 500 mV s-1, the voltammetric response changes significantly and becomes consistent with a reversible response that is under semi-infinite linear diffusion control; i.e., the experimental time scale is now sufficiently short so that only a fraction of the film is oxidized and reduced. Under these circumstances, the ratio of anodic to cathodic peak currents, ipa/ipc, is 1.00 ( 0.06 and ∆Ep is 65 mV, which compares favorably with the 57 mV value expected for an ideally reversible response. Moreover, plots of ip vs ν1/2 are linear. Assuming that the concentration of ruthenium centers within the film is 1.0 M, the slope using the Randles-Sevcik equation yields a homogeneous charge transport diffusion coefficient, DCT, of 6.4 ( 1.2 × 10-9 cm2 s-1. It is perhaps important to note that this represents the lower bound of DCT since any film swelling will decrease the ruthenium concentration, resulting in an underestimate of DCT. Electropolymerization in Acetonitrile and Sulfuric Acid. Electropolymerization in anhydrous acetonitrile and sulfuric acid has also been investigated. Significant research on the electropolymerization of [Ru(aphen)3]2+ in acetonitrile10 has been reported. Our results including the film deposition rate, volta(15) Sauerbrey, G. Z. Z. Phys. 1959, 155, 206. (16) Kankare, J. Langmuir 2002, 18, 7092. (17) Forster, R. J.; Walsh, D. A. Cyclic Voltammetry. Encyclopedia of Analytical Chemistry, 3rd ed.; Academic Press: New York, 2003. (18) Forster, R. J.; Keyes, T. E.; Bond, A. M. J. Phys. Chem. B 2000, 104, 6389. (19) Forster, R. J.; Iqbal, J.; Hjelm, J.; Keyes, T. E. Analyst 2004, 129(12), 1186–92. (20) Forster, R. J.; O’Kelly, J. P. J. Phys. Chem. 1996, 100, 3695. (21) Feldberg, S. W.; Rubinstein, I. J. Electroanal. Chem. 1988, 240, 1.

11236 Langmuir, Vol. 24, No. 19, 2008

Venkatanarayanan et al.

Figure 4. Interfacial electropolymerization from a 1.0 mM solution of [Ru(aphen)3]2+ on a 3 mm diameter glassy carbon electrode at a scan rate of 0.1 V s-1. The deposition medium was aqueous 0.5 M H2SO4, pH 0.3. The inset shows the voltammetric behavior following electrodeposition in blank 0.5 M H2SO4.

mmetric response, and rates of charge transport are consistent with these previous reports. Therefore, we focus the discussion here on electrodeposition from sulfuric acid. Consistent with previous reports,22,23 we find that the optimal pH for electropolymerization of [Ru(aphen)3]2+ in sulfuric acid is 0.3. This is most probably due to the protonation of free 5-aminophenanthroline ligand at low pH. It is known that the uncomplexed ligand is protonated at low pH; i.e., aphenH and aphenH224 are the dominant species at pH < 0.9, and these species appear to favor film growth and to produce films with superior stability. In sharp contrast to the behavior observed in the ionic liquid or acetonitrile, Figure 4 shows that in aqueous sulfuric acid the voltammetric response is irreversible. Oxidation and reduction peaks are observed at +1.5 and +0.5 V, respectively, and ipa/ipc is approximately 0.25 compared to the value of unity expected for a reversible response. Significantly, the response does not become any more ideal at slow scan rates. This suggests that slow heterogeneous electron transfer is not the origin of the irreversibility but rather arises because of a following chemical reaction. While our current data do not allow us to reach a definitive conclusion, the most likely explanation is that the pKa of the aminophenanthroline functionalities depends on the oxidation state of the ruthenium center. Specifically, the basicity of the amino group is likely to decrease significantly on oxidation of the metal site, thus causing the half-wave potentials for oxidation and reduction to be very different from one another. After 60 scans at 0.1 V s-1 the electrode was removed from the cell, rinsed thoroughly, and transferred to a cell containing blank electrolyte solution (0.5 M H2SO4, pH 0.3). The inset of Figure 4 shows that a persistent voltammetric response associated with the Ru2+/3+ couple is observed, suggesting that a redox-active film is formed on the electrode. However, the response is not consistent with that expected for a surface-confined species undergoing a reversible one-electron transfer reaction. For example, the ratio of anodic to cathodic current (ipa/ipc) is approximately 0.6 compared to the value of unity expected for a reversible response. The charge passed under the peak at +1.510 V yields a surface coverage of (4.4 ( 0.6) × 10-8 mol cm-2. (22) Galicia, L.; Rojas-Herna´ndeza, A.; Gomez-Hernandez, M.; Ramirez-Silva, M. T. Sens. Chemom. 2001, 65. (23) Cobos-Murcia, J. A.; Galicia, L.; Rojas-Herna´ndeza, A.; Ramı´rez-Silva, ´ lvarez-Bustamante, R.; Romero-Romo, M.; Rosquete-Pina, G.; PalomarM. T.; A Pardave´, M. Polymer 2005, 46, 9053. (24) Ramı´rez-Silva, M. T.; Go´mez-Herna´ndez, M.; Pacheco-Herna´ndez, M. L.; Rojas-Herna´ndez, A.; Galicia, L. Spectrochim. Acta, A 2004, 60, 781.

It is important to note that this surface coverage is only 40% of that found for deposition in the BDMITFSI although the number of voltammetric scans employed here is 10-fold higher. The charge transfer coefficient (DCT) for films formed in sulfuric acid was found to be 9.6 × 10-9 cm2 s-1, which is of similar order of magnitude compared to those formed in an ionic liquid. In sharp contrast, the DCT in acetonitrile was found to be (4.4 ( 0.8) × 10-11 cm2 s-1; i.e., charge transport through films formed in acetonitrile is approximately 2 orders of magnitude slower than that found for films formed in an ionic liquid or sulfuric acid. Parts A-C of Figure 5 show SEM images of the polymer deposited from BDMITFSI, acetonitrile, and sulfuric acid, respectively. From Figure 5A it can be seen that films electrodeposited in BDMITFSI exhibit significant surface roughness, most likely reflecting the high deposition rate. Wide-area views confirm that the whole of the electrode surface is covered by a continuous film, suggesting that nucleation does not influence the film structure. In contrast, parts B and C of Figure 5 show that films formed in sulfuric acid and acetonitrile are significantly smoother. Again, wide-area views of films formed in acetonitrile also confirm that a continuous film covers the whole electrode surface. The features of high surface coverage, continuous film formation, and high surface area make the films deposited from ionic liquid potentially useful for sensor and other applications.25,26 Spectroscopic Properties. Raman and luminescence spectroscopies have been used to investigate whether the differences in electrochemical behavior of the polymers are due to structural differences in the materials deposited from the different solvents. When polymer films formed on the electrode surface were probed with confocal fluorescence microscope, no emission was detected. To investigate whether this is because of the loss of aromaticity of phenanthroline ligands upon electropolymerization or due to the luminescence quenching by the electrode surface, the metallopolymers were dissolved off the electrode by prolonged sonication in acetonitrile and luminescence spectra of the resulting solutions measured. Significantly, all polymer solutions show a typical ruthenium polypyridine metal-to-ligand charge transfer (MLCT) emission peak at approximately 610 nm upon photoexcitation. It (25) O’Connor, M.; Kim, S. N.; Killard, A. J.; Forster, R. J.; Smyth, M. R.; Papadimitrakopoulos, F.; Rusling, J. F. Analyst 2004, 129, 1176. (26) Dennany, L.; Forster, R. J.; White, B.; Smyth, M.; Rusling, J. F. J. Am. Chem. Soc. 2004, 126, 8835.

[Ru(aphen)3](PF6)2 Films from an Ionic Liquid

Langmuir, Vol. 24, No. 19, 2008 11237

Figure 6. Resonance Raman spectra of the [Ru(aphen)3]2+ monomer drop-cast onto glass (a) and thin metallopolymer films electrodeposited from BDMITFSI (b) and sulfuric acid (c), respectively. Samples were irradiated using argon laser excitation at 488 nm.

Figure 5. SEM images of films of [Ru(aphen)3]2+ electropolymerized on a 25 µm radius platinum microelectrode from BDMITFSI (A) and on 1.5 mm radius glassy carbon electrodes from sulfuric acid (B) and anhydrous acetonitrile (C). In each case, the bar represents 100 µm.

also appears that the relative emission intensities do not depend strongly on the deposition medium. Depending on the mechanism of radical reaction, the NH• radical is thought to either dimerize with an adjacent NH• radical to form a diazo bond or react with a carbon atom on a neighboring complex to form an imine bond.11,27 The latter reaction would be expected to result in a significant change in the conjugation of the complex. That the polymer remains soluble is important as the monomer contains three reactive aphen ligands, which might be expected to lead to the formation of highly cross-linked network polymers. The fact that the polymers emit suggests that the aromaticity of the ligand is preserved upon electropolymerization and that the reaction mechanism is similar in all three processes (Supporting Information). The electrochemical and EQCM data clearly indicate that polymerization proceeds more rapidly in BDMITFSI than in anhydrous acetonitrile or aqueous sulfuric acid. While the metallopolymers exhibit similar absorption and emission spectra irrespective of the deposition medium, vibrational spectroscopy can directly probe the structure of the polymers formed. Figure 6 illustrates the resonance Raman spectra of a dropcast film of the parent monomer (a) and of polymer films formed in BDMITFSI (b) and aqueous sulfuric acid (c). The monomer exhibits Raman spectral features typical of resonance with an MLCT absorbance centered on a phenanthroline ligand with signature phenanthroline modes at 1606 (sh), 1577, 1514, 1500 (sh), 1460, 1435 (sh), 1313, 1190, 520, and 437 cm-1. For parent and electropolymerized films these features remain, albeit with some small shifts in frequency. In addition, there are some new features evident for each metallopolymer, and there are also (27) Nyasulu, F. W. M.; Mottola, H. A. J. Electroanal. Chem. 1988, 239, 175.

small but significant differences between the two metallopolymers. For both polymers an intense and broad new feature appears at approximately1600 cm-1. This feature is superimposed on a weak shoulder centered around 1606 cm-1 in the parent complex. This is likely to be associated with a CdC ring stretch, since a CdN imine bond produced by reaction of 5-imino (iphen•) would be expected to appear below about 1630 cm-1. The appearance of the 1600 cm-1 mode for the polymer is likely to arise from a change in conjugation of the phenanthroline CdC modes on polymerization. Correspondingly, there are a number of shifts and changes in intensity in the manifold of the ring stretch bands between 1400 and 1530 cm-1. The most intense C-C aromatic stretch mode in the parent complex appears at 1577 cm-1, and this mode shifts to slightly lower energy for both polymers at approximately 1574 cm-1. Between 1400 and 1530 cm-1 inplane ring stretches occur, and there are significant differences in the frequencies of these spectral features for the three spectra. For the parent monomer four features are observed at 1514, 1500 (sh), 1460, and 1435 (sh) cm-1. Electropolymerization in BDMITFSI causes a significant shift in the highest energy ring stretch mode to 1486 cm-1, and the remaining features appear at 1486, 1460, and 1420 cm-1; broadly, they appear to be shifted to lower frequency on electropolymerization. Broadly similar structural changes are induced in the polymer formed from sulfuric acid. Interestingly, for the polymer deposited from BDMITFSI a number of new modes appear which are not evident in the parent or the polymer deposited from sulfuric acid, most notably a shoulder at 1558 cm-1 and three peaks at 1274, 1171, and 1031 cm-1. The shoulder is attributed to an aromatic stretch, and the feature at 1171 cm-1 is tentatively attributed to a C-N azo stretch,28 with the mode at 1274 cm-1 tentatively attributed to a combination of C-N and NdN stretch modes.29 No corresponding NdN mode was observed, which would be expected to appear between 1370 and 1440 cm-1. However, this is a congested spectral region, and these new features are comparatively weak, so it is likely that the NdN feature is obscured. Overall, Raman spectral data indicate significant changes to the aromatic ring structures of the polymer compared with the parent monomer. This is a consequence of changes in the extent (28) Trotter, P. J. Appl. Spectrosc. 1977, 31, 30. (29) Andrikopoulos, P. C.; McCarney, K. M.; Armstrong, D. R.; Littleford, R. E.; Graham, D.; Smith, W. E. J. Mol. Struct. 2006, 789, 59. (30) Bard, A. J.; Keszthelyi, C. P.; Tachikawa, H.; Tokel, N. E. Chemilumim. Biolumin., Pap. Int. Conf. 1973, 3, 193. (31) Wallace, W. L.; Bard, A. J. J. Phys. Chem. 1979, 83, 1350.

11238 Langmuir, Vol. 24, No. 19, 2008

Venkatanarayanan et al.

reaction.30 It is the product of the efficiency of populating the excited state (φEX) and the quantum yield of emission from that excited state (φP). Here, [Ru(bpy)3]2+ was used as a relative standard. The relative efficiency was obtained using the relation

φECL ) φECL0(IQf0/QfI0)

Figure 7. Electrochemiluminescence response for a 3 mm glassy carbon electrode modified with a [Ru(aphen)3]2+ film electrodeposited from BDMITFSI (a), acetonitrile (b), sulfuric acid (c), and a blank solution (d) with only acetonitrile, where the coreactant is tripropylamine dissolved in acetonitrile and the scan rate is 0.1 V s-1.

of conjugation arising from imine formation. In the polymer electropolymerized from BDMITFSI, there is also Raman evidence for azo bond formation, which would arise from reaction of two NH• radicals. In contrast, these bands do not appear in the polymer deposited from sulfuric acid, suggesting that the mechanism for polymer termination may differ between BDMITFSI and H2SO4. It is tempting to speculate that the longer lifetime of the NH• radical in an ionic liquid facilitates both types of termination, i.e., imine formation and azo formation, in BDMITFSI whereas imine formation dominates in H2SO4. The retention of luminescence without changing the emission wavelength suggests that, for an individual complex, electropolymerization may involve reaction of a single aminophenanthroline ligand. This conclusion is also consistent with the solubility of the polymer and its persistent luminescence, which suggests that the polymers are not heavily cross-linked. This feature is attractive from the perspective of developing biosensors since the remaining amino-phen moieties can be used to bind biomolecules such as enzymes or antibodies. Electrochemiluminescence. One of the key objectives of this work was to investigate whether electropolymerized films exhibit ECL and to compare the efficiency of the ECL generation of the films electrodeposited from different solvents. Figure 7 shows the potential dependence of the ECL response for films produced in an ionic liquid (a), anhydrous acetonitrile (b), and sulfuric acid (c), all measured in acetonitrile solution containing TPA as a coreactant. All films produced ECL upon electrochemical production of Ru3+ centers in the presence of TPA. However, while the films deposited from both the ionic liquid and sulfuric acid showed intensities of similar magnitude, the intensities are considerably higher than that found for films deposited from acetonitrile. This result is consistent with the higher DCT values obtained for the films formed in BDMITFSI and sulfuric acid, compared to the substantially lower DCT value obtained for the films deposited from acetonitrile. In terms of developing sensors, a key issue is the relative quantum efficiency of the metallopolymer within a film compared to that in solution. To evaluate this parameter, we compared the quantum efficiency of the metallopolymer deposited from BDMITFSI in solution (acetonitrile) and within a thin film. The overall ECL efficiency (φECL) is defined as the number of photons emitted per faradic electron passed during the chemiluminescent

(2)

where φECL0 is the ECL efficiency of Ru(bpy)32+ (1 mM and 0.1 M TBABF4/ACN) via annihilation, taken as 5.0%,31 I and I0 are the integrated photomultiplier tube responses for the polymer and [Ru(bpy)3]2+, respectively, and Qf and Qf0 are the faradic charges passed for the sample and standard. The ECL efficiency of layers formed in an ionic liquid was estimated for the TPA coreactant pathway. Significantly, the overall efficiency of the ECL reaction for the polymer film obtained is 0.11 ( 0.03. This is approximately 5 times higher than the highest value obtained for the polymer dissolved in acetonitrile (0.018 ( 0.02). This enhanced efficiency for the immobilized form of the polymer most likely arises because the polymer matrix protects the electronically excited states from quenching and competing side reactions that occur when the species is freely diffusing in solution. It is also possible that the luminescence quantum yield is higher for the layer as a result of a decreased rate of vibrational relaxation due to the rigidity of the immobilized film compared to solution. These data clearly indicate that the overall ECL efficiency is improved by immobilizing the polymer on the electrode surface.

Conclusions The electropolymerization efficiency and properties of [Ru(aphen)3]2+ films formed by oxidative electropolymerization in the ionic liquid BDMITFSI, anhydrous acetonitrile, and sulfuric acid have been presented. Significantly, electropolymerization in BDMITFSI proceeds considerably more rapidly than in acetonitrile and sulfuric acid, and thicker films can be obtained. Films formed in BDMITFSI have considerably higher surface coverages ((9.8 ( 2.6) × 10-8 mol cm-2) than films formed in sulfuric acid ((4.4 ( 0.6) × 10-8 mol cm-2) or acetonitrile ((2.2 ( 0.6) × 10-8 mol cm-2). The rate of homogeneous charge transport through films deposited in BDMITFSI is relatively fast, and it takes approximately 500 ms to fully oxidize a 1 µm thick film. These metallopolymers exhibit significant ECL when oxidized to the Ru3+ state in the presence of TPA as a coreactant. However, films deposited from acetonitrile have lower emission than those deposited from either sulfuric acid or the ionic liquid. Significantly, the ECL efficiency of the metallopolymer is approximately 5 times higher within a thin film than in solution. This result highlights a key advantage of polymeric luminophores, namely, that the matrix may decrease the rate of radiative decay and block access of quenchers, e.g., molecular oxygen from solution. Acknowledgment. This material is based upon works supported by the Science Foundation Ireland under Grant No. 05/CE3/B754. Supporting Information Available: Figures showing the peak current dependence vs the root of the scan rate and the scan rate dependence of the voltammetric response for a thin film of [Ru(aphen)3]2+ electropolymerized on a 25 µm radius platinum microelectrode in BDMITFSI and emission spectra of the [Ru(aphen)3]2+ monomer in acetonitrile (a), the polymer formed in sulfuric acid (b), the polymer formed in acetonitrile (c), and blank acetonitrile (d). This material is available free of charge via the Internet at http://pubs.acs.org. LA8011316