Dawson Polyoxomolybdate α - American Chemical Society

Article pubs.acs.org/Langmuir

Ruthenium Metallopolymer: Dawson Polyoxomolybdate α‑[Mo18O54(SO4)2]4− Adduct Films: Sensitization for Visible Photoelectrocatalysis Jie Zhu,† James J. Walsh,† Alan M. Bond,‡ Tia E. Keyes,*,† and Robert J. Forster*,† †

School of Chemical Science, Dublin City University, Dublin, Ireland School of Chemistry, Monash University, Clayton, Victoria 3800, Australia

‡

ABSTRACT: Thin films of the adduct formed from the electrostatic association of the metallopolymer, [Ru(bpy)2(PVP)10]2+, and the Dawson polyoxomolybdate α[Mo18O54(SO4)2]4−, POMo, have been formed on ITO electrodes using an alternate immersion approach. The Ru/ POMo ratio is 4.5:1, which exceeds the 2:1 ratio expected on the basis of the charges of the Ru2+ and POMo4− building blocks. This behavior arises because of the polymeric character of the cation. In the presence of a substrate that has an abstractable proton such as benzyl alcohol, these rutheniumsensitized polyoxomolybdate films generate significant photocurrents under visible irradiation. Significantly, increasing the surface coverage of the adduct from 1.4 × 10−10 to 8.1 × 10−10 mol cm−2 does not measurably increase the photocurrent observed. Scan-rate-dependent cyclic voltammetry reveals that the rate of homogeneous charge transport through the film is slow, which most likely results in only a fraction of the film thickness being active for photoelectrocatalysis. The photocurrent increases markedly when the driving force for the oxidation of POMo5−, created by the photoelectrocatalytic oxidation of benzyl alcohol, is increased. This result is consistent with the dynamics of heterogeneous electron transfer being centrally important to the regeneration of the photoelectrocatalyst. A system in which the surface coverage and applied overpotential are optimized produces a photocurrent density of 190 ± 18 nA cm−2 under 480 ± 5 nm irradiation.

■

INTRODUCTION Polyoxometalates, POMs, are inorganic metal oxide anions1 that have very attractive photocatalytic and electrocatalytic properties.2,3 Although much POM-based photocatalysis is typically performed in solution,4 there is significant interest in surface-immobilized thin films that often require novel hybrid materials.5,6 POM-based thin films can be conveniently prepared via the Langmuir−Blodgett7 or a layer-by-layer (LBL) deposition method provided that a suitable cation is employed (e.g., transition-metal polypyridyl complexes). For example, multilayers of Dawson phosphotungstate α/β[W18O54(PO4)2]6− and [Ru(bpy)3]2+ have been built on a wide variety of substrates for use in electrocatalysis, electrochromism, and sensing.8−10 α/β-[W18O54(PO4)2]6− has also been employed in electrocatalytic multilayer assemblies with [Fe(bpy)3]2+ and [Os(bpy)3]2+.11,12 A wide variety of other POMo-based multilayer systems using [Ru(bpy)3]2+ as the counterion have been reported and include α/β[SiW12O40]4−,13 [PMo12O40]3−,14,15 [W10O32]4−,16 and [Eu(SiMo9W2O39)2]13−.17 In the majority of cases reported, these multilayers were employed as electrocatalysts for the reduction of small ions such as nitrate and nitrite, and photoelectrochemical applications have, to date, been relatively rare. © 2012 American Chemical Society

Dawson-type polyoxometalates have been extensively employed in the photocatalysis of both alcohols and water.4,18 However, these anions typically absorb in the UV region only and have relatively low extinction coefficients, which limits their overall catalytic efficiency. This issue can be addressed by using a visible-light sensitizer such as [Ru(bpy)3]2+ to enhance the POM photochemistry. For example, adducts of [Ru(bpy)3]2+ and Dawson ions γ[W 18 O 5 4 (SO 4 ) 2 ] 4− , α-[Mo 1 8 O 54 (SO 4 ) 2 ] 4− , and α/β[Mo18O54(SO3)2]4− have been reported.19−22 Some of these systems exhibit a new optical transition that has been identified as an intermolecular charge-transfer transition using resonance Raman. However, this transition is typically weak and depends strongly on the structure of the metal complex. For example, the quantum yield at 420 nm for the photo-oxidation of benzyl alcohol, BnOH, by [Ru(bpy)3]2α-[Mo18O54(SO4)2] is 40 times higher than that found for [(Hex)4N]4α-[Mo18O54(SO4)2] under the same conditions.23 Metallopolymers containing coordinated metal complexes represent a useful approach to creating sensitized POM Received: May 7, 2012 Revised: July 13, 2012 Published: July 19, 2012 13536

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541

Langmuir

Article

white-light source was an Oriel 68811 arc lamp employing a 350 W Xe bulb focused onto a 1 cm2 region of the film. A narrow-band optical filter (Spectrogon UK Ltd.) was used to irradiate the film selectively at 480 ± 5 nm so as to preferentially excite the new optical transition present in the adduct.24

adducts. For example, we reported on the properties of thin films of α-[Mo18O54(SO4)2]4− or α-[W18O54(PO4)2]6−/[Ru(bpy)2(PVP)10]2+ adducts and showed that thin films produce significant photocurrents under 480 nm irradiation that are not observed for the component ions alone; bpy is 2,2′-bipyridyl, and PVP is poly(4-vinyl pyridine).21,24,34 There are a significant number of factors that can influence the overall photoelectrocatalytic performance of these thin films, including the mass transport of the substrate, the efficiency with which the electronically excited state is created, the rate of reoxidation of the catalyst, and the energetics of the catalyst−substrate reaction. In this article, we report on those factors that influence the photocurrent for adducts of [Ru(bpy)2(PVP)10]2+ and α-[Mo18O54(SO4)2]4−, POMo, where the substrate is benzyl alcohol. Significantly, these investigations reveal the importance of the heterogeneous electron-transfer rate on the magnitude of the photocurrent observed.

■

■

RESULTS AND DISCUSSION Polyoxometalates, POMs, are powerful (photo)electrocatalysts capable of oxidizing recalcitrant organics.28 However, to achieve photoelectrocatalysis under visible irradiation, a sensitizer that absorbs in the visible region of the spectrum, such as [Ru(bpy)3]2+, is required; bpy is 2,2′-bipyridyl.21,23 The catalytic adducts exhibit a new optical transition, centered around 480 nm,19,20,29 and can photocatalytically oxidize proton donors such as benzyl alcohol under visible light (480 ± 5 nm). Recently, we demonstrated that metallopolymer [Ru(bpy)2(PVP)10]2+ sensitizes α-[S2Mo18O62]4− within thin films created using the layer-by-layer (LBL) alternate immersion technique; PVP is poly(4-vinyl pyridine).24 Here, the photoelectrocatalytic behavior of the metallopolymer− POW adduct is investigated for thin films in contact with acetonitrile. The parent metallopolymer is sparingly soluble in acetonitrile because of the high molecular weight of its backbone and the coordination of the ruthenium centers that introduce cross-linking. The ruthenium centers undergo a welldefined Ru2+/3+ oxidation process at approximately +1.00 V. The pKa of the pyridine moieties30 is approximately 2.2, and they are not expected to be protonated under the conditions used. Figure 1 illustrates the voltammetric behavior of [Ru(bpy) 2 (PVP) 10 ] 4.5 α-[Mo 18 O 54 (SO 4 ) 2 ], Ru/POMo, films

EXPERIMENTAL SECTION

Materials. Benzyl alcohol, ethanol, and acetonitrile (spectrophotometric-grade acetonitrile, Aldrich) were dried over molecular sieves (3 Å) prior to use. All aqueous solutions were made using deionized Millipore water. Tetrabutylammonium tetrafluoroborate, (But)4NBF4, was used as purchased (Fluka). Indium-doped tin oxide (ITO) electrodes were purchased from Delta Technologies Ltd. (Stillwater, MN, US) and were sonicated in ethanol for 10 min and dried under a stream of nitrogen before use. [(Hex)4N]4α-[Mo18O54(SO4)2] and [Ru(bpy)2(PVP)10](NO3)2 were synthesized as described previously and were characterized by ESI-MS, cyclic voltammetry, and IR spectroscopy.25,26 Electrospray ionization mass spectra were obtained using a Bruker Daltonics Esquire ESI ion-trap mass spectrometer operating in negative mode. Preparation of Composite Thin Films. Bis-substituted metallopolymer [Ru(bpy)2(PVP)10](ClO4)2 was prepared by refluxing cis[Ru(bpy)2(H2O)](ClO4)2 with a 10-fold excess of polyvinyl pyridine (PVP, MW 280 000 g mol−1, prepared by free radical polymerization and not fractionated, i.e., 1 ruthenium per 10 pyridine units) for 3 days. The preparation and characterization were reported in detail previously.27 Thin films were formed using alternate immersion in solutions of [Ru(bpy)2(PVP)10] followed by α-[Mo18O54(SO4)2]. Specifically, a clean 1 cm3 ITO electrode was immersed in a solution of [Ru(bpy)2(PVP)10](NO3)2 (0.1 mM in 4:1 EtOH/H2O). After 20 min, the electrode was removed, rinsed with blank 4:1 EtOH/H2O, and allowed to air dry. The metallopolymer-modified electrode was then immersed in a α-[Mo18O54(SO4)2] solution (1 mM in MeCN) for 20 min. The electrode was removed, rinsed with blank MeCN, and allowed to air dry. This cycle was repeated to create thicker films. The films were characterized using EDX, absorption, and emission spectroscopy as well as cyclic voltammetry. ATR-FTIR spectra were recorded using a Varian 610-IR FTIR microscope. Spectra were acquired using an attenuated total reflectance (ATR) accessory with a Ge crystal tip. In each case, 1064 spectra were collected and averaged. Instrumentation. Cyclic voltammetry was performed in a conventional three-electrode cell with a CH Instruments model 660a electrochemical workstation at 22 ± 2 °C. All solutions were deoxygenated with nitrogen for 20 min prior to use. The working electrode was a 1 cm2 ITO flag, and the pseudoreference electrode was Ag/Ag+, which was calibrated versus the ferrocene/ferrocenium (Fc/ Fc+) couple. The counter electrode was a large-area platinum flag. The surface coverages were determined by graphical integration of the background-corrected cyclic voltammograms using scan rates of 1 to 5 mV s−1. The total cell resistance was measured using a potential step where no faradaic processes occur. Amperometric i−t curves were used to measure the photocurrent response of the films. A quartz cuvette was used to accommodate the ITO electrode coated with a thin film of the adduct, a large-area platinum coil counter electrode, and a nonaqueous reference electrode (Ag/Ag+). The supporting electrolyte was 0.1 M (But)4NBF4. The

Figure 1. Cyclic voltammograms for a 1 cm2 ITO electrode following (A) one, (B) three, and (C) five deposition cycles of [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2]. The supporting electrolyte is 0.1 M (But)4NBF4 dissolved in acetonitrile, and the scan rate is 300 mV s−1. The inset shows a tapping mode AFM image of the film produced using three deposition cycles.

formed using one, three, and five deposition cycles where the scan rate is 300 mV s−1 and the supporting electrolyte is 0.1 M (But)4NBF4 in acetonitrile. Clearly defined redox waves were observed for the [Ru(bpy)2(PVP)10]2+/3+ redox processes at approximately +0.965 V, and the α-[Mo18O54(SO4)2]4−/5− couple is observed between +0.070 and 0.375 V. It is perhaps important that at this scan rate the response is under linear diffusion control (i.e., the depletion layer thickness is smaller than the film thickness) and the current observed does not depend on the surface coverage. The charge associated with the 13537

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541

Langmuir

Article

oxidation of α-[Mo18O54(SO4)2]5− is significantly larger than that observed for the corresponding reduction, but the peak currents decrease by less than 10% over a 4 h period of continuous cycling. The “additional” charge passed under the oxidation peak increases significantly when micromolar concentrations of water are added to the acetonitrile and is considered to arise from the photoelectrocatalytic oxidation of trace quantities of water in the supporting electrolyte solution. The formal potentials, E°′, for the Ru2+/3+ process are independent of the surface coverage whereas for the αPOM4−/5− couple they shift in a positive potential direction as the number of dip cycles is increased from one to three. The observation that it is thermodynamically easier to reduce the αPOM4− in thicker films reflects a change in the local microenvironment (e.g., a change in solvation). An AFM image of the film produced by a single deposition cycle is shown in the inset, and the rms roughness measured across an area of 1 μm2 is 25 ± 5 nm. The surface coverage, Γ, of each ion was calculated by the graphical integration of cyclic voltammograms recorded at scan rates of between 1 and 5 mVs−1. The surface coverage is directly proportional to the number of deposition cycles, with each cycle depositing 1.6 × 10−10 mol cm−2 of ruthenium centers. Significantly, the ratio of Ru to polyoxomolybdate centers in the composite film is 4.5 ± 0.2/1, which is higher than that found for adducts formed using monomeric complexes where the ratio corresponds exactly to full charge compensation at 2:1 or 3:1.11,8,12,31 This behavior most likely arises because of the polymeric nature of the cation (i.e., extra ruthenium centers are bound to the polymeric chain that provides the two ruthenium centers needed to neutralize the charge on the POMo). The metallopolymer/POMo adducts generate significant photocurrent densities when irradiated with visible light in the presence of a wide range of recalcitrant organics such as benzyl alcohol and DMF, provided they have an abstractable proton. When photoelectrocatalysis is performed using benzyl alcohol, the preliminary analysis of the solution and headspace using HPLC and GC, respectively, indicates that benzaldehyde and benzoic acid are produced. However, they are not present in sufficient quantities to account for all of the benzyl alcohol consumed, suggesting that a significant fraction is converted to carbon dioxide and water. Thus, whereas there are additional catalytic pathways involved, the reaction sequence for the photo-oxidation of benzyl alcohol by the [POMo-Ru] films is shown in eq 1 and the relative energy levels are shown in Chart 1: [POMo 4 −−Ru] → [POMo 4 −−Ru]*

Chart 1

benzyl alcohol concentration, and applied potential were systematically varied. Homogeneous Charge Transport Diffusion Coefficient. The photocurrent observed will ultimately depend on the number of [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] centers available, which will be influenced by the rate of homogeneous charge transport, DCT. The rate of charge transport through systems of this type tend to be limited by ion diffusion, which is influenced by the concentration of the supporting electrolyte. However, electrolytes such as perchlorate and even chloride tend to disrupt the electrostatic binding of Ru complexes to POMs, making electrolyte concentration studies challenging.20,32 The peak current for both the ruthenium and POMo centers increases linearly with the square root of the scan rate, υ, for 50 ≤ υ ≤ 500 mV s−1, which is consistent with a redox reaction under semi-infinite linear diffusion control. The Randles− Sevçik equation was used to determine DCT1/2C, where C is the concentration of the redox-active species within the film. To determine DCT, the concentration of ruthenium and POMo within the films was estimated using two different approaches. First, the concentration was estimated to be 0.02 M using the layer thickness from AFM and the surface coverage determined using cyclic voltammetry. Second, the density of a single solid grain of the adduct (approximately 0.2 mm3) was measured by flotation in nonsolvents and yielded a concentration of 0.70 M. The large difference in apparent concentrations most likely arises because the adduct is highly porous/swollen when in contact with acetonitrile. In the density measurements, the solvent will flood any pores and the measured concentration reflects the concentration of material within the solid “scaffold”, ignoring the pores. In contrast, the AFM approach treats the film as a monolithic block averaging the concentration over the solid scaffold and the pores. Specifically, the results suggest that the adduct swells by a factor of approximately three in all directions ((0.7 M/0.02 M)0.33) when in contact with acetonitrile. Significantly, density measurements for the pure parent metallopolymer yield a concentration of 0.80 M, which compares favorably with that found for the adduct. Given that charge transport mediated by the ruthenium centers will not occur within the pores, the concentration of 0.7 M has been used to estimate the homogeneous charge transport diffusion coefficient as (2.4 ± 0.6) × 10−14 cm2 s−1 for both the anodic and cathodic processes and is approximately independent of the surface coverage. Significantly, this diffusion coefficient is approximately 2 orders of magnitude lower that than found for the parent metallopolymer.33 This result indicates that either the interstrand separation has increased or that chargecompensating ions move more slowly through films of the adduct than the parent metallopolymer. From an applications perspective, this DCT indicates that it will take between

(1)

2[POMo 4 −−Ru]* + C6H5CH 2OH → 2[POMo5 −−Ru] + 2H+ + C6H5CHO

reoxidation: [POMo5 −−Ru] − e− → [POMo 4 −−Ru]

Current generation involves a number of steps, including mass transport of the substrate to/through the modifying film, (re)generation of the adduct that contains the reduced POMo5− and the Ru2+ centers, and charge transport through the modifying film.22 Also, the photocurrent density may depend on the quantity of the photoelectrocatalyst deposited on the ITO electrode. To probe the influence of these processes on the photocurrent density, the surface coverage, 13538

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541

Langmuir

Article

approximately 2 and 50 s to fully electrolyze the films formed using three to five deposition cycles. Raman spectroscopy reveals that the Ru-PVP/POMo films exhibit an additional vibrational mode at 900 cm−1 that is not present in either of the components, suggesting significant electronic communication between the ruthenium centers and the polyoxomolybdate.24 Moreover, UV−visible spectroscopy reveals that there is a new optical transition in the adduct centered at approximately 480 nm that is not present in either of the components. The influence of the surface coverage on the photocurrent generated was investigated in the presence of BnOH using a xenon arc lamp source and a 480 ± 5 nm narrow-band filter. The potential of the underlying ITO electrode was held so that the overpotential for the oxidation of the [POM5−/Ru] species generated during the oxidation of benzyl alcohol was +0.450 V. This potential maintains the ruthenium centers in the 2+ oxidation state. Figure 2 shows the

alcohol concentration. Moreover, no photocurrent is observed for pure anhydrous acetonitrile, which is consistent with it not having an abstractable proton. The magnitude of the photocurrent observed does not depend on the benzyl alcohol concentration as it is varied from 20 to 100%. This result confirms that at these high concentrations the mass transport of benzyl alcohol is not rate-limiting with respect to photocurrent production. Moreover, it suggests that the film structure does not change significantly as the substrate concentration is varied. Potential Dependence. The rate of heterogeneous electron transfer across the film/electrode interface may influence the magnitude of the photocurrent observed. This rate depends on the applied potential (i.e., because the potential is made more positive with respect to E°′ for the POM4−/5− couple, the rate of photocatalyst regeneration may increase). Figure 3 shows the photocurrent transients observed when the adduct films are irradiated in the presence of benzyl alcohol

Figure 2. Photocurrent−time curves generated for the [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] films on a 1 cm2 ITO electrode using one (···), three (---), and five (−) deposition cycles. The supporting electrolyte is 0.1 M (But)4NBF4 dissolved in acetonitrile, and the overpotential is +0.450 V. A narrow-band filter was used to irradiate the film selectively at 480 ± 5 nm.

Figure 3. Dependence of the photocurrent−time curves on the overpotential for the reoxidation of POM5− generated within [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] films by the oxidation of benzyl alcohol. The films were formed using a single deposition cycle. From top to bottom at 150 s, the overpotentials are 0.635, 0.537, 0.437, 0.337, and 0.238 V. The supporting electrolyte is 0.1 M (But)4NBF4 dissolved in 100% benzyl alcohol. The electrode/ illumination area is 1 cm2, and the surface coverage of ruthenium is 1.6 × 10−10 mol cm−2. A narrow-band filter was used to irradiate the film selectively at 480 ± 5 nm.

photocurrent−time curves generated for the [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] films where the surface coverage of ruthenium is 1.4 × 10−10, 4.7 × 10−10, and 8.1 × 10−10 mol cm−2. Consistent with slow charge transport through the layer, the photocurrent takes on the order of 30 to 75 s to reach a time-independent value. Significantly, the photocurrent observed does not increase appreciably with increasing surface coverage, with photocurrents of 69 ± 2 nA being observed despite the surface coverage increasing by a factor of approximately 6. This result may indicate that a slow homogeneous charge-transport rate results in only a fraction of the total thickness contributing to the photocurrent generation (i.e., the reaction with benzyl alcohol does not occur throughout the film thickness) or that mass transport in solution limits the overall rate. Concentration of Benzyl Alcohol. The impact of different concentrations of benzyl alcohol on the voltammetry and photocurrent behavior of [Ru(bpy) 2 (PVP) 10 ] 4.5 α[Mo18O54(SO4)2] films has thus been investigated. It is important to note that when the films are not irradiated no catalytic oxidation is observed at the applied potentials used in the photocatalytic experiments irrespective of the benzyl

with 480 ± 5 nm light as the overpotential is systematically varied from 0.238 to 0.635 V. The overpotential is measured with respect to the reoxidation of the POM5− species generated by the photo-oxidation of the benzyl alcohol. The most positive potential applied, +0.765 V versus the calibrated Ag/Ag+ pseudoreference electrode, maintains the ruthenium centers in the reduced (i.e., Ru2+) state. Overpotentials are quoted because, as shown in Figure 1, the formal potential for the α[Mo18O54(SO4)2]4−/5− couple shifts from +0.070 to 0.375 V as the surface coverage is increased, and using the overpotential allows the photocurrents to be directly compared as the surface coverage is varied. In all cases, irradiation produces a significant photocurrent that increases with increasing overpotential. The applied potential cannot change the driving force for the electrocatalytic reaction itself because this is controlled by the 13539

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541

Langmuir

Article

oxidation potential of the substrate (e.g., benzyl alcohol) and the reduction potential of the electronically excited state. However, the applied potential will control the rate of heterogeneous electron transfer to the reduced species (i.e., POM5−) and so influence the catalyst regeneration rate at least for centers relatively close to the electrode surface (i.e., a larger driving force regenerates the photocatalyst more rapidly, and a greater number of catalytic cycles occur per unit time). Figure 4

Figure 5. Photocurrent density vs time curve for the photo-oxidation of benzyl alcohol (60% v/v in acetonitrile) by a [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] adduct film. The film was formed using three deposition cycles, and the surface coverage of ruthenium is 4.8 × 10−10 mol cm−2. The supporting electrolyte is 0.1 M (But)4NBF4. The applied potential is 0.800 V. A narrow-band filter was used selectively to irradiate the film at 480 ± 5 nm.

POM5− product following the oxidation of the benzyl alcohol. Significantly, the turnover number (moles of product per hour divided by moles of catalyst) is approximately 11 million under optimized conditions. However, the rate of homogeneous charge transport associated with the Ru2+/3+ couple is very slow, 2.4 ± 0.6 × 10−14 cm2 s−1. This means that the surface coverage cannot be increased significantly to increase the current density because only a fraction of a thicker film will be used for photoelectrocatalysis. Under these circumstances, the dynamics of heterogeneous electron transfer influence the photocurrent to the greatest extent. The overall performance may be improved by increasing the conductivity of the adduct layer (e.g., by incorporating metal nanoparticles at loadings above the percolation threshold so as to increase the homogeneous charge transport diffusion coefficient).

Figure 4. Dependence of ln(jphoto) on the overpotential for the oxidation of the POM5− product produced within the metallopolymer−POMo adduct by the photooxidation of benzyl alcohol. Conditions are the same as in Figure 3.

shows the dependence of the semilog photocurrent density, jphoto, on the overpotential, η (i.e., the difference between the applied potential and the formal potential, E°′, of the POM4−/5− redox couple) for thin films of the adduct. The linear response is consistent with the Butler−Volmer formulation of electrode kinetics and clearly indicates that the magnitude of the photocurrent generated depends on the rate of heterogeneous electron transfer. Optimized System. Having optimized the key parameters affecting the magnitude of the photocurrent individually, we thought it important to probe the system properties using the parameters in a single experiment. Figure 5 shows the photocurrent responses obtained at +0.800 V for [Ru(bpy)2(PVP)10]4.5α-[Mo18O54(SO4)2] adduct layers formed using three deposition cycles in 60% (v/v) benzyl alcohol. The optimized system produces a photocurrent of 193.6 ± 17.6 nA cm−2. Taken in conjunction with the data presented in Figure 4, it is evident that optimizing the applied potential so as to increase the rate of photoelectrocatalyst regeneration by heterogeneous electron transfer increases the photocurrent density by more than a factor of 3 irrespective of the surface coverage.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected], [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This material is based upon work supported by the Science Foundation Ireland under the Research Frontiers Programme, award no. 07/RFP/MASF386. We sincerely thank Mr. Damien McGuirk and Dr. Sharon Murphy (DCU) for the acquisition of ESI mass spectra. We also thank Dr. Colm Mallon (DCU) for many helpful discussions and Mr. Shane O’Carroll for synthesizing the [Ru(bpy)2(PVP)10](NO3)2.

■

■

CONCLUSIONS Thin films of the adduct formed between the metallopolymer, [Ru(bpy) 2 (PVP) 10 ] 4.5 , and the polyoxomolybdate, α[Mo18O54(SO4)2], generate photocurrents under visible irradiation in the presence of benzyl alcohol. The sensitization of the polyoxometallate with the metallopolymer eliminates the barriers posed by the need to use UV excitation for the parent polyoxometallate. The photocurrent generated in the presence of benzyl alcohol with irradiation at 480 nm increases significantly with overpotential for the reoxidation of the

REFERENCES

(1) Long, D.-L.; Burkholder, E.; Cronin, L. Chem. Soc. Rev. 2007, 36, 105. (2) Yin, Q.; Tan, J. M.; Besson, C.; Geletii, Y. V.; Musaev, D. G.; Kuznetsov, A. E.; Luo, Z.; Hardcastle, K. I.; Hill, C. L. Science 2010, 328, 342. (3) Bi, L.-H.; Foster, K.; McCormac, T.; Dempsey, E. J. Electroanal. Chem. 2007, 605, 24. (4) Rüther, T.; Bond, A. M.; Jackson, W. R. Green Chem. 2003, 5, 364.

13540

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541

Langmuir

Article

(5) Pan, D.; Chen, J.; Tao, W.; Nie, L.; Yao, S. Langmuir 2006, 22, 5872. (6) Kida, T. Langmuir 2008, 24, 7648. (7) Clemente-León, M.; Ito, T.; Yashiro, H.; Yamase, T.; Coronado, E. Langmuir 2007, 23, 4042. (8) Fay, N.; Dempsey, E.; Kennedy, A.; McCormac, T J. Electroanal. Chem. 2003, 556, 63. (9) Wang, B.; Bi, L.-H.; Wu, L.-X. J. Mater. Chem. 2011, 21, 69. (10) Ammam, M.; Easton, E. B. J. Mat. Chem. 2011, 21, 7886. (11) Fay, N.; Dempsey, E.; Kennedy, A.; McCormac, T. Electrochim. Acta 2005, 51, 281. (12) Fay, N.; Dempsey, E.; McCormac, T. J. Electroanal. Chem. 2005, 574, 359. (13) Zhang, J.; Goh, J. K.; Tan, W. T.; Bond, A. M. Inorg. Chem. 2006, 45, 3732. (14) Ma, H.; Peng, J.; Chen, Y.; Feng, Y.; Wang, E. J. Solid State Chem. 2004, 177, 3333. (15) Wang, X.-L.; Han, Z.-B.; Wang, E.; Zhang, H.; Hu, C.-W. Electroanalysis 2003, 15, 1460. (16) Bi, L.; Wang, H.; Shen, Y.; Wang, E.; Dong, S. Electrochem. Commun. 2003, 5, 913. (17) Dong, T.; Ma, H.; Zhang, W.; Gong, L.; Wang, F.; Li, C. J. Colloid Interface Sci. 2007, 311, 523. (18) Bernardini, G.; Zhao, C.; Wedd, A. G.; Bond, A. M. Inorg. Chem. 2011, 50, 5899. (19) Seery, M. K.; Guerin, L.; Forster, R. J.; Gicquel, E.; Hultgren, V. M.; Bond, A. M.; Wedd, A. G.; Keyes, T. E. J. Phys. Chem. A 2004, 108, 7399. (20) Keyes, T. E.; Gicquel, E.; Guerin, L.; Forster, R. J.; Hultgren, V. M.; Bond, A. M.; Wedd, A. G. Inorg. Chem. 2003, 42, 7897. (21) Walsh, J. J.; Long, D.-L.; Cronin, L.; Bond, A. M.; Forster, R. J.; Keyes, T. E. Dalton Trans. 2011, 40, 2038. (22) Ruhlmann, L.; Costa-Coquelard, C.; Hao, J.; Jiang, S.; He, C.; Sun, L.; Lampre, I. Can. J. Chem. 2008, 86, 1034−1043. (23) Fay, N.; Hultgren, V. M.; Wedd, A. G.; Keyes, T. E.; Forster, R. J.; Leane, D.; Bond, A. M. Dalton Trans. 2006, 32, 4218. (24) Zhu, J.; Zeng, Q.; O’Carroll, S.; Keyes, T. E.; Forster, R. J. Electrochem. Commun. 2011, 13, 899. (25) Cooper, J. B.; Way, D. M.; Bond, A. M.; Wedd, A. G. Inorg. Chem. 1993, 32, 2416. (26) Forster, R. J.; Vos, J. G. Macromolecules 1990, 23, 4372. (27) Hogan, C. F.; Forster, R. J. Anal. Chim. Acta 1999, 396, 13−21. (28) Long, D. L.; Tsunashima, R.; Cronin, L. Angew. Chem., Int. Ed. 2010, 49, 1736. (29) Seery, M. K.; Fay, N.; McCormac, T.; Dempsey, E.; Forster, R. J.; Keyes, T. E. Phys. Chem. Chem. Phys. 2005, 7, 3426−3433. (30) Franck-Lacaze, L.; Sistat, P.; Huguet, P. J. Membr. Sci. 2009, 326, 650. (31) Wang, B.; Bi, L.-H.; Wu, L.-X. J. Mater. Chem. 2011, 21, 69. (32) Ballardini, R.; Gandolfi, M. T.; Balzani, V. Inorg. Chem. 1987, 26, 862. (33) Devadoss, A.; Spehar-Deleze, A.-M.; Tanner, D. A.; Bertoncello, P.; Marthi, R.; Keyes, T. E.; Forster, R. J. Langmuir 2010, 26, 2130. (34) Walsh, J. J.; Mallon, C.; Bond, A. M.; Keyes, T. E.; Forster, R. J. Chem. Commun. 2012, 48, 3593.

13541

dx.doi.org/10.1021/la300886s | Langmuir 2012, 28, 13536−13541