Electrochemistry and Electrochemiluminescence of [Ru(II)-tris

Jul 25, 2008 - Juan López-Gejo , Álvaro Navarro-Tobar , Antonio Arranz , Carlos Palacio ... Leopoldo Della Ciana , Simone Zanarini , Rossana Perciacca...
3 downloads 0 Views 377KB Size
Download PDF
10188

J. Phys. Chem. B 2008, 112, 10188–10193

Electrochemistry and Electrochemiluminescence of [Ru(II)-tris(bathophenanthroline-disulfonate)]4- in Aprotic Conditions and Aqueous Buffers Simone Zanarini,*,† Leopoldo Della Ciana,*,‡ Massimo Marcaccio,† Ettore Marzocchi,‡ Francesco Paolucci,† and Luca Prodi† Dipartimento di Chimica “G. Ciamician”, UniVersita’ di Bologna, Via Selmi 2, 40126 Bologna, Italy, and Cyanagen srl Via Stradelli Guelfi, 40/c, 40138 Bologna, Italy ReceiVed: April 29, 2008; ReVised Manuscript ReceiVed: June 10, 2008

In this work, the electrochemical and ECL properties of tris[1,10-phenanthrolinediyl-4,7-di(benzenesulfonate)]Ru(II) ([Ru(BPS)3]4-) have been addressed in both strictly aprotic conditions and aqueous buffers. A combined theoretical and experimental approach is presented to focus thermodynamics and kinetic effects of electro-generated species possessing highly negative charge. The complex, prepared as the sodium salt by using a newly developed procedure, was subsequently converted to the tetrabutylammonium salt by ion exchange, thus making it soluble in organic media and allowing, for the first time, its thorough electrochemical investigation in ultra-dry aprotic media. The electrochemically induced luminescence (ECL) of Na4[Ru(BPS)3] in phosphate buffer, using the co-reactant method (tripropylamine), was investigated as a function of the electrode material and halide addition, and ECL intensities six times higher than that of [Ru(bpy)3]2+ were found. In addition, the ECL behavior of this promising dye for biomolecule recognition was investigated in aprotic media and, for the first time, the direct radical anion-radical cation annihilation ECL was obtained. Introduction The improvement of water solubility of Ru(II) and Ir(III) complexes is a possible strategy to transfer their electrochemical and photophysical properties in biological buffers enabling bioanalytical applications. As it was similarly done for diphenylanthracene,1 the sulfonation of phenyl groups into ligands was a successful method to make the tris [1,10-phenanthrolinediyl-4,7-di(benzenesulfonate)] ruthenate(II) ([Ru(BPS)3]4-; see Chart 1 for structure) complex soluble and stable in aqueous media.2 Bis-chelating disulfonate bathophenanthroline ligand and the corresponding metal complexes have found many practical applications.3 [Ru(BPS)3]4- and the BPS2- ligand have been effectively used as chromogenic compounds in the colorimetric measurement of hepatic4 and blood serum iron concentration5 and as staining agents for protein detection electrophoresis gel.6 The [Ru(BPS)3]4-, as the widely used unsulfonated complex, has been used to prepare a fluorimetric oxygen sensor.7 However, because of its relatively low solubility in organic media, there is still a gap in the electrochemical characterization of such a highly negatively charged Ru(II) complex. This work concerns the investigation of the electrochemical and electrochemiluminescent properties of [Ru(BPS)3]4- aimed to elucidate the effect of -SO3- groups on the behavior of the complex with respect to the previously reported unsulfonated homologue.8 On the other hand, a combined theoretical and experimental approach is presented to understand the thermodynamic and kinetic effects of the high negative charge of electrogenerated species on the formation of excited state. The introduction of tetrabutylammonium (TBA) counter-cations (by ion-exchange chromatography) enabled us to carry out, for the first time, a thorough voltammetric characterization of [Ru(BPS)3]4- in * Corresponding author. E-mail: [email protected] (S.Z.) and [email protected] (L.D.C.). † Universita’ di Bologna. ‡ Cyanagen srl via Stradelli Guelfi.

CHART 1: Chemical Structure of [RuII(bathophenanthroline-disulfonate)3]4([Ru(BPS)3]4-)a

a The commercially available ligand is obtained by sulfonating the phenyl rings of bathophenantroline in the para and meta positions; for this reason, the complex synthesized from this ligand is actually a mixture of isomers, which are however not distinguishable electrochemically or spectroscopically.

aprotic media. By using ultra-dry acetonitrile solutions of (TBA)4[Ru(BPS)3], both the metal-centered oxidation and several ligand-centered reduction processes were then observed that were not accessible in aqueous media, thus allowing a comparison with the unsulfonated homologue. Additionally, taking advantage of the wider potential window available in such a medium, ECL emission was successfully generated, for the first time, by direct cation-anion annihilation, that is, without the use of any co-reactant.2 Experimental Section Materials. RuCl3 · ×H2O and Na2(bathophenanthrolinedisulfonate) · 3H2O were purchased from Fluka. [Ru(bpy)3](PF6)2

10.1021/jp803757y CCC: $40.75  2008 American Chemical Society Published on Web 07/25/2008

[Ru(II)-tris(bathophenanthroline-disulfonate)]4(from Cyanagen srl) was recrystallized twice from acetone and used as a standard in fluorescence and ECL measurements. Piranha solution was prepared by mixing 98% H2SO4 and 30% H2O2 in 3:1 ratio (v/v). Synthesis of Tetrasodium and Tetrakis-tetrabutylammonium [1,10-Phenanthrolinediyl-4,7-di(benzenesulfonate)] Ruthenate(II)(Na4[Ru(BPS)3]and(TBA)4)3[Ru(BPS)3]).RuCl3 · × H2O (50 mg, 0.191 mmol) and Na2(bathophenanthrolinedisulfonate) · 3H2O (350 mg, 0.593 mmol, 3.1 equiv) were dissolved in 10 mL of argon degassed ethylene glycol. The dark red solution was refluxed under argon for 4.5 h. After cooling to room temperature, 20 mL of absolute ethanol (EtOH) were added. The resulting orange solution was added dropwise to 250 mL of vigorously stirred diethyl ether. A reddish brown tar separated, and the solvent was decanted. The residue was dissolved in the minimum amount of water and purified twice by size-exclusion chromatography (1.5 × 22 cm LH-20 Sephadex column; eluent: water). The red-fluorescent band was collected and evaporated to give a reddish-brown solid which was dried in vacuo over P2O5. Yield: 284 mg (84%) for C72H42N6Na4O18RuS6 · 6H2O, MW 1772.57. The water of crystallization was determined by differential thermogravimetry, with a Nietzsch STA 409 PC/PG instrument (expected mass change for loss of 6 H2O, 6.10%; found, 6.05%). The tetrabutylammonium salt was prepared by ion exchange. A Dowex 50WX8-200, 1 × 7 cm column, sodium form, was exchanged with tetrabutylammonium bromide 1 M and subsequently washed with water. A 25 mg sample of the ruthenium complex dissolved in 300 µL of water was loaded onto the column. The orange product was elueted with water, collected, and evaporated under reduced pressure. Yield: 96%, MW 2650.54, (hexahydrate). Electrochemistry, Electrochemiluminescence (ECL), and Photophysical Measurements. Electrochemical one-compartment airtight cells with O-ring sealed high-vacuum glass stopcocks were used for electrochemical and ECL measurements in strictly aprotic conditions. The connections to the highvacuum line and to the Schlenk flask containing the solvent were made by spherical joints fitted with Kalrez O-rings. The working electrode consisted of a Pt disk ultramicroelectrode (UME, with diameter of 125 µm) or, alternatively, a sideoriented 2 mm diameter Pt disk sealed in glass. The counter electrode consisted of a platinum spiral and the quasi-reference electrode was a silver spiral. Further details about the electrochemical cell were described elsewhere.9 Acetonitrile (MeCN, spectroscopy grade from Merck), after being refluxed over CaH2, was distilled under vacuum at room temperature with a high refluxing ratio, utilizing 1 m length distillation column filled with glass rings. Dichloromethane (DCM, from Fluka) was refluxed over and successively distilled from P2O5 and activated 4 Å molecular sieves. Solvents were stored in specially designed Schlenk flasks over 3 Å activated molecular sieves, protected from light, and kept under vacuum prior to use. Tetrabutylammonium hexafluorophosphate (TBAH, from Sigma-Aldrich of electrochemical grade) was used without further purification as supporting electrolyte in organic media. Further details on the electrochemical instrumentation used with UME were reported elsewhere.10 ECL and electrochemical measurements with 2 mm Pt disk electrode were carried out with an AUTOLAB electrochemical station (Ecochemie). The E1/2 values for reductive and oxidative processes were calculated by adding ferrocene as an internal standard. In aqueous media, phosphate buffer solution (PBS) was used as supporting electrolyte; the buffer was prepared using equimolar quantities of Na2HPO4 and KH2PO4

J. Phys. Chem. B, Vol. 112, No. 33, 2008 10189 (from Sigma-Aldrich). The concentration of PBS was 0.1 M, and pH was adjusted to 7.5 by adding small amounts of concentrated H3PO4 or a 3 M solution of NaOH. TPA was added as oxidative coreactant at the concentration of 3 × 10-2 M (i.e., the saturated solution at pH ) 7.5). ECL was obtained in a single oxidative step by generating at the same time the TPA and [Ru(BPS)3]4- oxidized form according to well-established methods.11 The ECL signal during cyclic voltammetry was measured with a photomultiplier tube (PMT, Hamamatsu model R928) placed at a few millimeters distance in front of the working electrode, inside a darkbox. To register light/current/ voltage curves, the PMT output signal was sent to an ultralow noise current preamplifier (Acton research model 181). After electronic processing, the signal was directly sent to the second input channel of AUTOLAB. The ECL spectrum was recorded by inserting the same PMT in a dual exit monochromator (Acton Research model Spectra Pro 2300i). Photocurrent detected at PMT was accumulated for 3-5 s for monochromator step depending on emission intensity; entrance and exit slits were fixed to the maximum value of 3 mm. PMT was biased at 750 V. Absorbance spectra were collected with a Cary 5 UV-visNIR spectrophotometer and the photoluminescence was investigated with a Varian (model Cary Eclipse) fluorimeter by using a 10 mm quartz cuvette with fittings for Ar degassing and for connection to a vacuum line. Results and Discussion Tris[1,10-phenanthrolinediyl-4,7-di(benzenesulfonate)] Ru(II) tetrasodium was prepared by a new, efficient method employing ethylene glycol (instead of the conventional ethanol/water mixture) as both solvent and reducing agent due to the Ru(II) limited reactivity with BPS4- ligand. A similar method was previously routinely used in the synthesis of Os(II) complexes from Os(III) salts.12 It is however noteworthy that Os(II) complexes unlike RuBPS are cationic, crystalline, and easily purified by precipitation from aqueous solution with perchlorate or hexafluorophosphate. The successive exchange of sodium counterions with tetrabutylammonium ions (TBA) greatly improves the solubility of [Ru(BPS)3]4- in organic media, thus allowing the subsequent electrochemical characterization. Electrochemical Behavior The voltammetric study was carried out in MeCN (Figure 1) and DCM solvents, under strictly aprotic conditions. The E1/2 values for oxidative and reductive processes in the two solvents are summarized in Table 1. These data can be compared with those obtained for the unsulfonated homologue, [Ru(4,7diphenyl-l,10-phenanthroline)3]2+ ([Ru(BP)3]2+), in MeCN.8,13 Not unexpectedly, [Ru(BPS)3]4- is, at the same time, slightly easier to oxidize (∆E1/2,I Ox ) -0.07 V) and slightly more difficult to reduce (∆E1/2,1 Red ) -0.02 V, ∆E1/2,2 Red ) -0.04, ∆E1/2, 3,red ) -0.04 V) than [Ru(BP)3]2+, as a likely consequence of its high negative charge (4-), vis-a`-vis the positive charge of [Ru(BP)3]2+. The anodic peak associated to the Ru(II) center one-electron oxidation displayed in both solvents a slightly larger height than the ligand-based reduction peaks. This behavior was attributed to the occurrence of adsorption phenomena involving the oxidized species, and accordingly, the oxidation peak height increased almost linearly with scan rate as typically observed for processes involving redox species immobilized on the electrode surface.14 This hypothesis, in line with reported adsorption and electro-induced crystallization of Ru(bpy)3-like complexes,15 was confirmed by the voltammetric curve recorded at low temperature (Figure S1, Supporting Information) that

10190 J. Phys. Chem. B, Vol. 112, No. 33, 2008

Zanarini et al. standard.19 Much lower solubility of Na4[Ru(BPS)3] with respect to the latter complex in organic solvents was observed as an effect of the sulfonate groups (the MeCN solution was in fact visibly cloudy). The formation of colloidal aggregates was confirmed by the red shift of MLCT band and consequently by the lower luminescence quantum efficiency in MeCN with respect to H2O. After exchanging the sodium counterion with TBA, the MeCN solution appeared clear; the red shift of MLCT band disappeared, and the luminescence efficiency became similar to that obtained in aqueous media. Electrochemiluminescence

Figure 1. Cyclic voltammetry of 0.1 mM [Ru(BPS)3](TBA)4 in 0.1 M TBAH/MeCN solution. Scan rate: 0.5 V/s. Working electrode: Pt disk (diameter ) 0.125 mm). T ) 25 °C. Dashed line: cyclic voltammogram under the same experimental conditions including the fourth reduction process occurring at the edge of the solvent discharge. Reversing the scan after the fourth reduction makes the anodic counterpart of peak of 1, 2, and 3 less reversible.

TABLE 1: Summary of the Electrochemical Processes of [Ru(BPS)3](TBA)4 in strictly Aprotic Conditions (T ) 25 °C, scan rate ) 0.5 V/s) MeCN, Pt electrode φ ) 0.125 mm process

E1/2/V (vs Fc/Fc+)

I ox. 1 red. 2 red. 3 red. 4 red.

+0.77 -1.71 -1.86 -2.07 (-2.65)

comments partially irrev.

estimated by simulation of voltammetric curve

DCM, Pt electrode φ ) 3 mm I ox. 1 red. 2 red. 3 red.

+0.77 -1.72 -1.91 -2.12

partially irrev. partially irrev.

clearly displays the triangular shape of the oxidative process and larger heights (about two times) with respect to reduction peaks. Interestingly, the E1/2 values of the first reduction and oxidation processes were almost identical in MeCN and DCM, while relatively larger (negative) values were found for the second and third reductions in DCM, because of its lower dielectric constant (Table 1).16 At more negative potentials, close to the baseline discharge limit, a fourth reduction peak is observed. The E1/2 of such a partly irreversible process, -2.65 V (estimated by digital simulation of the CV curves), allows to identify such a process as the introduction of a second electron in one of the one-electron reduced ligands: the 600 mV gap between the third and the fourth peaks is, in fact, typical of bpy-like ligands and is associated to the coupling energy of two electrons within the same redox orbital.17,18 Photophysical Properties in MeCN and H2O Photophysical data for Na4[Ru(BPS)3] and (TBA)4[Ru(BPS)3] are summarized in Table 2 where [Ru(bpy)3]2+ was used as a

Radical Cation-Radical Anion Annihilation ECL Generation. Because of the redox properties of [Ru(BPS)3]4described above and reported in Table 1 and of its photophysical properties (Table 2), electrochemically generated luminescence by direct (metal-centered) radical cation (ligand-centered)-radical anion annihilation is thermodynamically allowed for [Ru(BPS)3]4-. In fact, the free energy of such a process (∆GES) can be calculated according to the following equations11,20

∆GES ≈ ∆Gannihil +

1240.8 EMLCT(nm)

∆Gannihil ) - (E1⁄2,Ox - E1⁄2,Red) + ∆Gel

(1) (2)

Equation 2 includes the ∆Gel term that accounts for the electrostatic interactions between the partners of the annihilation reaction when such species are both charged. Such a term may represent a significant fraction of the free energy involved in the process and was therefore taken into account in the calculation of ∆GES. The term is positive in the present case (i.e., repulsion increases from charge transfer between reactants). However, assuming two spherical ions with radius 1.1 nm, in contact with each other in a dielectric continuum with a relative permittivity of 35.94,21 we found that ∆Gel corresponds to ∼0.18 eV, that is, a relatively small fraction of the electrochemical term in eq 2 (2.48 eV). Since the MLCT excited-state energy 00 (EMLCT ) of TBA4[Ru(BPS)3] species, in Ar degassed MeCN,22 was 2.03 eV (corresponding to the emission centered at 618 nm at room temperature; see Table 2), by using eqs 1 and 2 and the above electrochemical data (Table 1), it follows that ∆GES (MeCN) ) -0.27 eV. This value suggests that there is an energy excess in the electron transfer process and that, as a consequence, the thermodynamic ECL criterion is met. ECL of [Ru(BPS)3]4- by such a generating mechanism has, however, never been reported. By contrast, ECL produced by [Ru(BPS)3]4- oxidation in the presence of TPA has been reported.2 In this case, ∆GES resulted to be -0.61 eV, where the estimated reducing power of deprotonated electro-generated TPA+ cation is introduced as E1/2,Red.23 The E1/2, Red of -1.7 V versus SCE in MeCN has been converted to -2.05 versus Fc/ Fc+ considering a mean value of literature data.24 Notice that, because of the very large reducing power of the coreactant, generation of [Ru(BPS)3]5-, either directly from the parent [Ru(BPS)3]4- or indirectly from the electrochemically produced [Ru(BPS)3]3- species, may also take place; in such cases, light emission would also occur via the of [Ru(BPS)3]5- and [Ru(BPS)3]3- annihilation reaction. In a first series of experiments, the ECL of (TBA)4[Ru(BPS)3] was obtained in MeCN by cation-anion annihilation under strictly aprotic conditions (Figure 2 and Figure 3). The ECL signal was about 3 orders of magnitude less intense than that registered in the same conditions with [Ru(bpy)3]2+ (φECL ) 0.0032% for [Ru(BPS)3]4- and 4.0% for [Ru(bpy)3]2+). A

[Ru(II)-tris(bathophenanthroline-disulfonate)]4-

J. Phys. Chem. B, Vol. 112, No. 33, 2008 10191

TABLE 2: Summary of Photophysical Properties of [Ru(BPS)3](Na)4, [Ru(BPS)3](TBA)4, and [Ru(bpy)3]2+ in MeCN and H2Oa [Ru(BPS)3](Na)4 [Ru(BPS)3](TBA)4 [Ru(bpy)3]2+

solvent

MLCT absbλmax/nm

(Lbcm-1bmol-1)

fluorescence λmax /nm

ΦPLc

MeCNd

480 460 464 465 450 455

27 000 27 000 26 300 27 600 14 000 18 000

626 615 618 620 615 615

2.7% 17.6% 14.5% 22.1% 5.9% 4.2%

H2 O MeCN H2O MeCN H2O

a Ar degassed solutions. [Ru] ) 10-5 M. b Instrument: Varian Cary 5. c Istrument: Varian Cary Eclipse. PMT bias was set at 600 V. Five nanometer excitation and emission slits were selected. Excitation wavelength ) max. abs MLCT. [Ru(bpy)3]2+ was taken as reference according to literature data.33 d Solution is visibly cloudy due to the formation of colloidal aggregates; the reduced solubility is also suggested by the red shift of abs. max. and confirmed by the reduced fluorescence quantum efficiency in MeCN with respect to water.

Figure 2. Light/current/time profiles for the cation-anion direct annihilation of 0.1 mM [Ru(BPS)3](TBA)4 in 0.1 M TBAH/DCM solution. Double potential step program: E1 ) +1.8 V (first oxidation), E2 ) -1.12 (first reduction) V, t1 ) t2 ) 1 s. All of the potentials are vs Ag wire. Sample time: 1 ms. PMT bias: 750 V. ECL current range: 10-6 A/V.

Figure 3. Cation-anion annihilation ECL spectrum of 0.1 mM [Ru(BPS)3](TBA)4 species in 0.1 M TBAH/DCM solution. Potential step program as described in Figure 2. PMT bias: 750 V. Integration time: 5 s for each monochromator step.

possible explanation of such a lack of efficiency of the process, in spite of its highly favorable thermodynamics, is the electrostatic repulsion between oxidized and reduced form of the complex, both carrying a high negative charge. This may

kinetically limit the coupling of electrogenerated [Ru(BPS)3]3and [Ru(BPS)3]5- and consequently the excited-state generation. In fact, for [Ru(bpy)3]2+, the charge of electrogenerated acceptor and donor species is 3+ and 1+, respectively, while for [Ru(BPS)3]4-, the corresponding species bear charges 3- and 5-, respectively. Note also that, by cyclic voltammetry at 1 V/s, ECL emission was produced only during [Ru(BPS)3]4reduction, while, when the scan rate was progressively increased to 10 V/s, a weak ECL emission was also recorded during the oxidation step, although with a significantly lower intensity with respect to that produced during the reduction one. On the other hand, the use of fast alternated reductive and oxidative pulses (Figure 2) with the progressive inclusion of the second and third reduction processes or the increase of the upper positive potential did not succeed in giving comparable ECL intensities in the reduction step vis-a`-vis the oxidation one. A possible reason for the relatively lower ECL intensity during the oxidation step is the formation of a resistive layer onto the electrode surface because of the adsorption/crystallization of the oxidized complex as already described above. Such a film would instead be removed upon the application of negative potentials, thus restoring the fully active electrode area and allowing higher efficiency of the ECL generating process. Finally, the ECL spectrum obtained by direct cation-anion annihilation (Figure 3) was equivalent to that registered by photoexcitation (maximum wavelength at ∼620 nm), thus showing that the same excited state is populated and the [Ru(BPS)3]3- and [Ru(BPS)3]5- species do not undergo any electro-induced decomposition and confirming the prevailing electrostatic nature of the filming product. Moreover, experimental evidence shows that the MLCT electro-generated emission energy is unaffected by the sulfonation of the phenyl groups on the ligands.25 Aqueous Buffers. The ECL behavior of [Ru(BPS)3]4- was also investigated in phosphate buffer solution by oxidizing the complex in the presence of TPA. By comparing the light/current/ potential profiles of [Ru(bpy)3]2+ and [Ru(BPS)3]4- (Figure 4), the ECL maximum intensity of the latter is approximately six times higher,26 a value very close to the relative fluorescence quantum efficiency recorded in water (see Table 2). The current/ potential profile (Figure 4) shows three peaks in the positive potential region, labeled as I, II, and 1, respectively. Peak I (E ) 0.77 V vs Ag/AgCl) can be confidently assigned to the formation of an oxide on the working electrode surface that is successively reduced in the reverse scan (process 1). The potential value is, in fact, too low to be associated with the irreversible oxidation of TPA.27 Recently, an analogous behavior was reported and discussed in detail.28 The peak marked as II is associated with Ru(II) oxidation and maximum ECL intensity. The voltammograms of [Ru(bpy)3]2+ and [Ru(BPS)3]4- show very similar curves and currents, with an irreversible process

10192 J. Phys. Chem. B, Vol. 112, No. 33, 2008

Figure 4. Typical air equilibrated light/current/voltage profiles of 10-5 M [Ru(bpy)3]2+ (dashed line) and [Ru(BPS)3]Na4 (solid line) in 0.1 M PBS/H2O solution containing 3 × 10-2 M TPA as coreactant. PMT bias ) 750 V; ECL current range ) 10-6 A/V; scan rate ) 0.5 V/s; working electrode Pt disk (diameter ) 3 mm); potentials are vs saturated Ag/AgCl.

close to E) +1.2 V. The oxidation potential of the single complex can be however more accurately estimated by the position of the ECL emission peaks (E1/2,I,Ox([Ru(bpy)3]2+) ) +1.15 V; E1/2,I,Ox([Ru(BPS)3]4-) ) +1.19 V vs Ag/AgCl). These values are, as expected, very similar, thus confirming that the delay of ECL emission with respect to the electrochemical stimulus can be satisfactorily neglected, especially considering the relatively slow scan rate. Stability of ECL Activity of [Ru(BPS)3]4- PBS Solutions. A simple test has been performed to check whether high ECL efficiency of [Ru(BPS)3]4- is retained in a PBS 0.1 M solution (pH ) 7.5) for a six months period. The results presented in Figure S1 show that ECL maximum intensity, registered in the same conditions and with the same potential program already described in Figure 4, remains within standard deviation considering three repetitions of each measurement. Effect of Bromide Addition and Electrode Material into Light/Current/Potential Profiles. The effect on the ECL of the Pt electrode surface conditioning by KBr was monitored in a 0.1 M PBS aqueous solution containing 3 × 10-2 M TPA. Two distinct modifications were observed comparing the light/ current/potential profiles of [Ru(BPS)3]4- upon KBr addition (see Figure 5a). The first was a moderate increase, approximately 25%, in ECL intensity maximum during Ru(II) oxidization without any appreciable shift in the potential (E ) 1.19 V, peak marked II in Figure 5a). The second effect was the appearance, and the progressive increase, of a second ECL emission peak (E ) 0.90 V vs Ag/AgCl, peak I). This potential value is in agreement with that of TPA in phosphate buffer.29 Such an effect can thus suggest that the same mechanism for the excited-state generation is active for both [Ru(bpy)3]2+ and [Ru(BPS)3]4-. This process, as already reported, is based on the oxidation of TPA without the electrochemical generation of the oxidized [Ru(BPS)3]4-.30 As previously demonstrated,28a the addition of small amounts of halide is an effective method to partially avoid the formation of surface oxides that can slow down the kinetics of TPA oxidation and inhibit the oxidation of the [Ru(BPS)3]4-. By replacing the Pt electrode with a glassy carbon (GC) disk electrode of equivalent geometric area31 an increase in ECL intensity was observed, consistent with the complete absence of oxide on GC electrodes. (Figure 5b; 100% increase of intensity on the peak I and 60% on the peak II with respect to Pt electrode). The potentials of the ECL maxima (Figure 5; peak

Zanarini et al.

Figure 5. (a) Effect of KBr addition into light/current/potential profile of [Ru(BPS)3]4- registered with a Pt disk electrode; without KBr (blue line), 5 µM KBr (black line), and 10 µM KBr (red line). (b) Comparison of light/current/potential profile for Pt (solid line) and glassy carbon (dashed line) in the presence of 10 µM KBr. Experimental conditions for all tests were 10-5 M RuBPS in 0.1 M PBS/H2O solution (pH ) 7.5), containing 3 × 10-2 M TPA. PMT bias ) 750 V; ECL current range ) 10-6 A/V; scan rate ) 0.5 V/s. Air equilibrated cell; potential is referred to saturated Ag/AgCl. Both GC and Pt working electrodes are 3 mm diameter disk. Peaks labeled as I and I′ correspond approximately to the potential of TPA oxidation while II and II′ are in correspondence of [Ru(BPS)3]4- first oxidative process.

I′, E ) + 1.05 V; peak II′, E ) + 1.26 V vs Ag/AgCl) are slightly shifted to more positive values, as expected from the different nature of GC with respect to Pt. Concluding Remarks For its practical use in biological assays the [Ru(BPS)3]4complex displays a remarkably higher IECL,R in PBS buffer, with respect to [Ru(bpy)3]2+, together with a very high water solubility (>1 M). Moreover, an high robustness is proved by the unchanged ECL activity of the PBS solution over a sixmonth period. These results are very interesting for practical applications, since they can further decrease the already very low detection limit offered by ECL methods. Furthermore, these studies can open up a very interesting perspective. Recently, many iridium(III) complexes, exhibiting excellent ECL yields at different wavelengths, have in fact been described.32,33 Thus, as suggested by the present investigation, the sulfonated derivatives of these water insoluble Ir(III) complexes might be expected to be promising dyes for biomolecule recognition. Finally, in view of the discussed effects of highly negative charge on electrochemical and ECL properties, in conceiving new materials for ECL applications our results suggest that it is likely convenient to reduce the number of -SO3- groups per molecule to an optimum value. This reduction, in fact, while retaining high water solubility, can help to limit the formation of adsorbed/crystalline films onto the electrode surface upon oxidation and eventually speed up the direct radical cation-anion annihilation ECL generation. Acknowledgment. This work was accomplished in the framework of LATEMAR (www.latemar.polito.it), Centre of Excellence funded by MIUR (Italian Ministry for Education, University and Research) Grants FIRB 2003-2004 for public/ private structures involved in research fields characterized by strategic value. Supporting Information Available: Cyclic voltammogram of (TBA)4[Ru(BPS)3] in MeCN at low temperature; plot of ECL intensity registered from a [Ru(BPS)3]4- 10-5 M solution in PBS 0.1 M (pH ) 7.5) in function of the time from preparation. This material is available free of charge via the Internet at http:// pubs.acs.org.

[Ru(II)-tris(bathophenanthroline-disulfonate)]4References and Notes (1) Richards, T. C.; Bard, A. J. Anal. Chem. 1995, 67 (18), 3140. (2) Blanchard, R. M.; Martin, A. F.; Nieman, T. A.; Guerrero, D. J.; Ferraris, J. P. Mikrochim. Acta 1998, 130, 55. (3) For the first reported synthesis of the ligand and relative Ru(II) complex, please refer to. (a) Anderson, S.; Constable, E. C.; Seddon, K. R.; Turp, J. E.; Baggott, J. E.; Pilling, M. J. J. Chem. Soc., Dalton Trans. 1985, 2247. (b) Cryberg, R. L.; Diehl, H. Proc. Natl. Acad. Sci. 1963, 70, 184. (4) Pieroni, L.; Khalil, L.; Charlotte, F.; Poynard, T.; Piton, A.; Hainque, B.; Imbert-Bismut, F. Clin. Chem. 2001, 47, 2059. (5) (a) Deadre, J. J.; Harold, W. L. Clin. Chim. Acta 1990, 189 (2), 199. (b) Reagents for colorimetric determination of iron in blood serum; patent nr. JP 59050364 of 1984/03/23 and nr. JP 03003911 of 1991/01/21. (6) (a) Radioactive staining of gels to identify proteins; Patent nr. US 4459356 of 1984/07/10. (b) Method for the reversible staining of proteins using cationic surfactants and metal chelates; Patent nr. DE 102004022463 of 2005/12/01. (c) Rabilloud, T.; Strub, J-M.; Luche, S.; Dorsselaer, A.; Lunari, J. Proteomics 2001, 1, 699–704. (d) Lamanda, A.; Zahn, A.; Roeder, D. H.; Langen, Proteomics 2004, 4, 599–608. (7) Castellano, F. N.; Lakowicz, J. R. Photochem. Photobiol. 1998, 67 (2), 179. (8) (a) McCord, P.; Bard, A. J. J. Electroanal. Chem. 1991, 318, 91. (b) Kapturkiewicz, A. Chem. Phys. Lett. 1995, 236, 389. (c) Kapturkiewicz, A.; Szrebowaty, P. Dalton Trans. 2002, 3219. (9) Cecchet, F.; Gioacchini, A. M.; Marcaccio, M.; Paolucci, F.; Roffia, S.; Alebbi, M.; Bignozzi, C. A. J. Phys. Chem. B 2002, 106, 3926. (10) (a) Bruno, C.; Marcaccio, M.; Paolucci, D.; Castellarin-Cudia, C.; Goldoni, A.; Streletskii, A. V.; Drewello, T.; Barison, S.; Venturini, A.; Zerbetto, F.; Paolucci, F. J. Am. Chem. Soc. 2008, 130 (12), 3788. (b) La Pense´e, A. A.; Bickley, J.; Higgins, S. J.; Marcaccio, M.; Paolucci, F.; Roffia, S.; Charnock, J. M. Dalton Trans. 2002, 4095. (11) (a) Bard, A. J., Ed.; Electrogenerated Chemiluminescence; Marcel Dekker: New York, 2004. (b) Pyati, R.; and Richter, M. M. Annu. Rep. Prog. Chem., Sect. C 2007, 12. (c) Marquette, C. A.; Blum, L. J. Anal. Bioanal. Chem. 2008, 390, 155. (d) Miao, W.; Electrogenerated Chemiluminescence and Its Biorelated Applications. Chem. Rev.,web release May 28, 2008. (12) Della Ciana, L.; Dressick, W. J.; Sandrini, D.; Maestri, M.; Ciano, M. Inorg. Chem. 1990, 29, 2792. (13) E1/2, I Ox ) +0.84 V vs. Fc/Fc+; E1/2, 1 Red )-1.69 V, E1/2, 2 Red )-1.82 V and E1/2, 3 Red )-2.03 vs. Fc/Fc+. (14) Bard, A. J.; Faulkner, L. R. Electrochemical methods: fundamentals and applications; Wiley: New York, 2001. (15) (a) Echegoyen, L.; DeCian, A.; Fischer, J.; Lehn, J.-M. Angew. Chem.Int. Ed. 1991, 30 (7), 838. (b) Pe`rez-Corderoi, E.; Buigas, R.; Brady, N.; Echegoyen, L. HelV. Chim. Acta 1994, 77 (5), 1222.

J. Phys. Chem. B, Vol. 112, No. 33, 2008 10193 (16) This different behavior upon solvent change can be explained by considering the effect of ion pairs that is discussed in detail in a submitted work. (17) Marcaccio, M.; Paolucci, F.; Roffia, S. Chapter 8 in Trends in Molecular Electrochemistry; Marcel Dekker: New York, 2004. (18) (a) Marcaccio, M.; Paolucci, F.; Paradisi, C.; Carano, M.; Roffia, S.; Fontanesi, C.; Yellowlees, L. J.; Serroni, S.; Campagna, S.; Balzani, V. J. Electroanal. Chem. 2002, 532, 99. (b) Marcaccio, M.; Paolucci, F.; Fontanesi, C.; Fioravanti, G.; Zanarini, S. Inorg. Chim. Acta 2007, 360, 1154. (19) Previous photophysical studies focused on energy and electron transfers involving [Ru(BPS)3]4- (a) Demas, J. N.; Harris, E. W.; McBride, R. P. J. Am. Chem. Soc. 1976, 99, 3547. (b) Kamat, P. V.; Ford, W. E. J. Phys. Chem. 1989, 93, 1405. (20) Kapturkiewicz, A.; Angulo, G. Dalton. Trans. 2003, 3907. (21) Maggini, M.; Guldi, D. M.; Mondini, S.; Scorrano, G.; Paolucci, F.; Ceroni, P.; Soffia, S. Chem. Eur. J. 1998, 4, 1992. 00 (22) The EMLCT was estimated by the fluorescence maximum in MeCN at 77 K (e¨max ) 610 nm). (23) Lai, R. Y.; Bard, A. J. J Phys. Chem. A 2003, 107 (18), 3335. (24) Pavlishchuk, V. V.; Addison, A. W. Inorg. Chim. Acta 1999, 298, 97. (25) For a comparison of fluorescence maxima with respect to [Ru(BP)3]2+, see ref 12c. (26) This value of relative ECL intensity can not be rigorously called ECL quantum efficiency but relative emission intensity IECL,R. (27) For an estimation of the TPA oxidation potential by varying electrode materials and surface conditioning, vide infra. (28) (a) Zu, Y.; Bard, A. J. Anal. Chem. 2000, 72 (14), 3223. (b) Zanarini, S.; Rampazzo, E.; Bich, D.; Canteri, R.; Della Ciana, L.; Marcaccio, M.; Marzocchi, E.; Montalti, M.; Panciatichi, C.; Pederzolli, C.; Paolucci, F.; Prodi, L.; Vanzetti, L. J. Phys. Chem. C 2008, 112 (8), 2949. (29) Please see ref 11 page 232 and ref 19a. (30) See ref 11 page 233. (31) Because of the different roughnesses of the two materials, the geometric surface of the GC electrode was normalized to that of the platinum electrode (having the same diameter) by measuring the charge exchanged in a separate chronoamperometric experiment with the same ferrocene solution in ACN. (32) (a) Kapturkiewicz, A.; Chen, T.-M.; Laskar, I. R.; Nowacki, J. Electrochem. Commun. 2004, 6, 827. (b) Kapturkiewicz, A.; Nowacki, J.; Borowicz, P. Electrochim. Acta 2005, 50, 3395. (c) Kapturkiewicz, A.; Nowacki, J.; Borowicz, P. Z. Phys. Chem. 2006, 220, 525. (33) Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T. Handbook of Photochemistry 3rd edition; CRC Press: Boca Raton, 2006.

JP803757Y