Electron- and Energy-Transfer Processes in a Photocatalytic SystemBased on an Ir(III)-Photosensitizer and an Iron CatalystAntje Neubauer,*,† Gilbert Grell,‡ Aleksej Friedrich,† Sergey I. Bokarev,*,‡ Patrick Schwarzbach,†

Felix Gartner,§ Annette-E. Surkus,§ Henrik Junge,§ Matthias Beller,§ Oliver Kuhn,‡

and Stefan Lochbrunner†

†Institute for Physics, University of Rostock, Universitatsplatz 3, 18055 Rostock, Germany‡Institute for Physics, University of Rostock, Wismarsche Str. 43-45, 18057 Rostock, Germany§Leibniz-Institute for Catalysis, Albert-Einstein-Str. 29a, 18059 Rostock, Germany

*S Supporting Information

ABSTRACT: The reaction pathways of bis-(2-phenylpyridinato-)(2,2′-bipyridine)-iridium(III)hexafluorophosphate [Ir(ppy)2(bpy)]PF6 within a photocatalytic waterreduction system for hydrogen generation based on an iron-catalyst were investigatedby employing time-resolved photoluminescence spectroscopy and time-dependent densityfunctional theory. Electron transfer (ET) from the sacrificial reagent to the photoexcited Ircomplex has a surprisingly low probability of 0.4% per collision. Hence, this step limits theefficiency of the overall system. The calculations show that ET takes place only for specificencounter geometries. At the same time, the presence of the iron-catalyst represents anenergy loss channel due to a triplet−triplet energy transfer of Dexter type. This losschannel is kept small by the employed concentration ratios, thus favoring the reductive ETnecessary for the water reduction. The elucidated reaction mechanisms underline the further need to improve the sun light’senergy pathway to the catalyst to increase the efficiency of the photocatalytic system.

SECTION: Spectroscopy, Photochemistry, and Excited States

For the large-scale utilization of sunlight as a sustainable,greenhouse-gas-emission-neutral, and secure energy

source, an efficient storage of solar energy in chemical fuels isessential.1 Therefore, water splitting as one possible way toachieve this goal has gained high scientific attention in recentyears. A broad variety of heterogeneous and homogeneousphotocatalytic systems for hydrogen generation has beeninvestigated. (See refs 1−6 and references therein.) Becauseof the complexity of the processes involved, each of the halfreactions, water oxidation and reduction, is often studiedseparately by adding a sacrificial oxidant or reductant,respectively. Thus, homogeneous photocatalytic systems forhydrogen generation contain a light-absorbing photosensitizer,optionally an electron relay to facilitate charge separation, acatalyst which reduces the aqueous protons to hydrogen, and asacrificial reductant (SR). Surprisingly, despite the diversity ofthe investigated systems based on these components, detailedmechanistic studies are scarce.4,7−9 Eisenberg et al.7,8

investigated systems with cobaloxime complexes as catalystsand Pt terpyridyl acetylide chromophores as photosensitizers.For these systems, oxidative quenching of the Pt photo-sensitizer by the cobaloximes are described, and rate constantsbetween 4 × 107 and 2 × 109 M−1 s−1 were reported for thisprocess. The latter rate constants are close to the diffusion-controlled limit. Mechanistic studies of a homogeneousphotocatalytic system with heteroleptic Ir complexes asphotosensitizer, including the here studied bis-(2-phenyl-

pyridinato-) (2,2′-bipyridine)iridium(III)hexafluorophosphate[Ir(ppy)2(bpy)]PF6 (IrPS), [Co(bpy)3]

2+ as electron relay,and triethanolamine as SR, revealed two quenching pathways.9

Therein, photoluminescence (PL) quenching rates due to theelectron relay were described as about one order of magnitudehigher than those due to the SR. For a similar system withoutelectron relay and employing Ir complexes as PS and Pt colloidsas catalyst, a reductive quenching mechanism is described.10

However, mechanistic insights for different pathways on amolecular level are still missing. Recently, some of us developeda homogeneous photocatalytic system based on IrPS asphotosensitizer, [HFe3(CO)11][NEt3H] as catalyst (Fe-cat)replacing the noble-metal complexes, and triethylamine (TEA)as SR (Scheme 1).11−14 The catalytically active species andintermediates for this system have been studied by in situelectron paramagnetic resonance, in situ infrared spectroscopyand DFT, as well as CASSCF/CASPT2 calculations.12,15 Forinstance, the reduced IrPS− species has been observed underirradiation of IrPS in a THF/TEA/water mixture. However, inthe presence of the Fe-cat, this species could not be detecteddue to limited time resolution.15,16

This paper addresses the primary photophysical reactionsteps of the iridium photosensitizer [Ir(ppy)2(bpy)]PF6 with its

Received: February 28, 2014Accepted: March 26, 2014Published: March 26, 2014

Letter

pubs.acs.org/JPCL

© 2014 American Chemical Society 1355 dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−1360

reaction partners, that is, the SR and the catalyst in a THF orTHF/water solution. It aims at mechanistic insights into theindividual reaction steps on a molecular level and at identifyingloss channels on relevant time scales for these processes. Inparticular, by means of steady-state and time-resolved PLspectroscopy as well as DFT/TD-DFT (density functionaltheory/time-dependent density functional theory) calculations,we focus on the following issues: (i) electron transfer (ET)from the SR TEA to the excited IrPS, (ii) excited-statequenching of the IrPS by the iron catalyst as loss channel in thephotocatalytic system, and (iii) the dynamics of these reactionsteps in the overall photocatalytic system. Finally, the relevanceof the interplay of these reaction steps for the photocatalyticsystem is discussed.Quenching by Sacrif icial Reductant. The ET from the SR TEA

to the excited IrPS was studied by PL quenching experiments inTHF and THF/water mixture. The experimental conditions,for example, concentrations of the components, were chosen tobe comparable to previous activity studies of the photocatalyticsystem12 but also to minimize spectroscopically problematiceffects such as inner filter phenomena.17 The determined PLquantum yield, φPL, and lifetime, τ0, of IrPS in degassed THFare 0.15 (±0.03) and 370 ns (±30 ns), respectively. PLlifetimes reported in literature9,14,18−22 range from values ofabout 18019 to 565 ns,22 demonstrating the high sensitivity ofthe IrPS PL to solvatochromic and quenching effects. In fact,this is the lifetime of the lowest triplet excited electronic state,which has been proven to be relevant for photocatalysisaccording to spectroscopic9,14,23 and computational studies.24

By adding TEA as SR, the PL lifetime of IrPS decreases, forexample, to 50 ns for a 5 vol% TEA concentration (0.4 M,Figure 1a). All measured PL decay signals exhibit amonoexponential behavior. Figure 1b shows the Stern−Volmerplot of the quantum yield and lifetime for quenching with TEAconcentrations in the range of 0.3 to 2.6 M. The dependenciesof lifetime and quantum yield ratios on concentration are linearwithin experimental accuracy and for the chosen concentrationregime. Both plots of lifetime and corresponding yields coincidenicely, indicating the high reliability of the extractedinformation. Thus, the quenching rate constant kq wasdetermined from the slope of the fit according to the Stern−Volmer equation (eq 1)25

ττ

φφ

τ= = + k1 [Q]q0 0

0(1)

where τ0 and φ0 are the PL lifetime and quantum yield in theabsence of the quencher and τ and φ are PL lifetime andquantum yield at the quencher concentration [Q]. Thefollowing discussion is based on the mean values for kq derivedfrom the lifetime and quantum yield ratios.Interestingly, the quenching rate constant in a THF/water

mixture (17 vol% water content), kq,water, is reduced by a factorof ∼5 (Figure 1b) compared with the quenching rate constant

kq in THF. The here-derived PL quenching rate of 1.3 × 107

M−1 s−1 for the excited state of IrPS in THF/water is in goodagreement with the ones determined by Bernhard et al. inacetonitrile/water as solvent with TEA and triethanolamine asquencher, where kq was determined to be 1.9 × 107 and 6 × 106

M−1 s−1, respectively.9,23 The reduced quenching rate in THF/water mixture compared with pure THF can probably beattributed to the reaction of TEA as a weak base with water orshielding by water molecules surrounding the positivelycharged IrPS from interaction with the quencher TEA.Additionally, a shorter excited-state lifetime can also reducethe evaluated quenching rate constant. However, the impact ofthis effect is very low here due to a relatively long measured PLlifetime of 230 ns in THF/water mixture (Figure 1a). Thesequenching experiments also suggest that increasing the amountof SR leads to a higher performance of the photocatalyticsystem. However, previous hydrogen evolution experiments ina solvent mixture of THF/water showed a maximum activity at33 vol% TEA (∼2.4 M), which was due to the limited solubilityof TEA in the THF/water mixture.12 The measured rates canbe compared with the diffusion limit (Supporting Information,section B). The calculated diffusion rate constant kD for IrPSand TEA in THF as solvent is 1.4 × 1010 L mol−1 s−1 and hence∼200 times higher than the observed quenching rate constant.

Scheme 1. Scheme of the Investigated Photocatalytic System

Figure 1. (a) Photoluminescence decay curves and monoexponentiallyfitted time constants of IrPS (1.3 × 10−5 M) in degassed THF withoutTEA (black triangles) and with 5 vol% TEA (black circles) as well as indegassed THF/17 vol% water solution without TEA (red triangles)and with 5 vol% TEA (red circles). (b) Stern−Volmer plots: Ratios ofluminescence lifetimes (triangles) and quantum yields (squares) versusTEA concentration in degassed THF solution without (open symbols)and with 17 vol% water (filled symbols). Mean quenching rateconstants, determined by the slope of the linear fit curves for thelifetimes (solid line) and quantum yields (dashed line), are also given.

The Journal of Physical Chemistry Letters Letter

dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−13601356

This indicates that the quenching efficiency per encounter issurprisingly low with about 0.4 and 0.1% in THF and THF/water mixture, respectively. To elucidate the origin of the lowquenching efficiencies on a molecular level, we performed TD-DFT calculations.According to the oxidation and reduction potentials (Scheme

SI-1 in the Supporting Information) and the zero-zero energyE00 of IrPS with 2.36 eV (525 nm), as determined by theinterception of absorption and luminescence spectra, the tripletexcited state of IrPS (T1) with an redox potential of +1.27 Vversus NHE is capable of oxidizing TEA (+1.09 V vs NHE),thus being reduced to IrPS−, where the spin density ispredominantly located on the bipyridine ligand.16 Hence, thereductive luminescence quenching of IrPS in the triplet excitedstate by TEA corresponds to an ET from TEA in the groundstate (S0) to IrPS (T1), resulting in radical species in theirdoublet ground states (D0), as given in eq 2 (conservation ofthe total spin is allowed for)

The charge-transfer (CT) behavior of the system IrPS/TEAand the binding energies of the corresponding collisioncomplexes were investigated at the DFT/TD-DFT level. (Seethe Supporting Information for computational details.) Theenergy of the D0+D0 state should be lower than that of theT1+S0 state to make the reaction (eq 2) thermodynamicallypossible. For analysis of the CT behavior, the lowest five tripletstates of the joint IrPS/TEA system were calculated.In Figure 2a, the angular dependence of the binding energy

between TEA and IrPS is plotted for a distance of R = 7 Åbetween the nitrogen atom of TEA and the iridium atomapproximately corresponding to the potential energy minimumof the joint system. (For results with R = 8 Å, see theSupporting Information.) The plot represents a sphericalsurface and regions, where the polar angles are φ = 0 and πcontract into one point. The spherical grid is represented byblack dots, and the projections of ligand atoms onto the angularcoordinates are represented by red dots for the 2,2-bipyridineligand and green or blue dots for the two 2-phenylpyridineligands. The grid points, where the energy of the CT (D0+D0)state is lower than the energy of the localized one (T1+S0), aremarked with large orange circles in Figure 2a.The number of favorable orientations is only about 5−10%

of the total 152 points for distances of 7 to 8 Å between IrPSand TEA. The CT process is favored by a closer distance andthe inclusion of solvent effects but in general only possible for avery limited number of configurations.In addition, the calculated binding energies for IrPS and TEA

are of the same order of magnitude as the binding energiesbetween the solvent molecules calculated with the samemethod (Table SI-1 and Figure SI-5 in the SupportingInformation). Hence, no stable complexes of TEA and IrPSare expected to form in the multicomponent reaction mixture.Furthermore, the orientations favorable for CT do notsystematically correspond to the regions with the highestbinding energy, but the CT state is lower in energy when TEAis close to the ligands. Weak bonding and the small number oforientations favoring CT thus explain the relatively low rates forreductive quenching observed in the experiments because onlya small fraction of collisions leads to the products.

Quenching by Iron Catalyst. The PL of IrPS is also quenchedby the presence of Fe-cat in the reaction solution. For instance,the PL lifetime of 370 ns for IrPS in pure degassed THF isreduced to 280 ns for an Fe-cat concentration of 1.6 × 10−5 M(Table 1). To investigate the reaction between IrPS and Fe-cat,

we performed corresponding luminescence quenching experi-ments (Figure 3a). All measured PL decay curves exhibit amonoexponential behavior. Figure 3b shows the steady-stateabsorption spectra of IrPS and Fe-cat as well as the normalizedluminescence spectrum of IrPS. Because of the spectral overlapof the absorption spectrum of the iron catalyst and theluminescence spectrum of the iridium photosensitizer, the

Figure 2. (a) Angular dependence (θ and φ are azimuthal and polarangles in radians, as described in panel b) of the binding energy (inelectronvolts, contour plot). TEA was placed on a spherical grid (blackpoints), centered at the iridium atom with a distance of R = 7 Åbetween the iridium atom of IrPS and the nitrogen atom of TEA. Red,green, and blue dots correspond to projections of ligand atoms ontothe sphere. Orange circles denote orientations energetically favorablefor CT reaction (eq 2). (c) Density difference plot of the CT statecorresponding to n(TEA) + dx2−y2(Ir) + π(ppy) → π*(bpy) excitation.

Table 1. PL Lifetimes of IrPS (c = 1.3 × 10−5 M) in DifferentSolvent Mixtures (degassed) and without and with Fe-cat (c= 1.6 × 10−5 M)a

solvent componentsτPL [ns]

without Fe-catτPL [ns]

with Fe-cat

THF 370 280THF/TEA (5/1) 13 14THF/water (5/1) 230 210THF/TEA/water (4/1/1) 42 42

aIf added, the amount of TEA or water was kept constant at aproportion of ∼17 vol %.

The Journal of Physical Chemistry Letters Letter

dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−13601357

procedure for determining the Stern−Volmer constant wasslightly changed. Instead of determining the intensity ratios forthe whole luminescence spectrum, the considered spectralrange was limited to 650−750 nm, where the spectral overlap isnegligible. Additionally, the obtained intensity values werecorrected for the absorption of incident light by the Fe-cat(Figure SI-1 in the Supporting Information).The concentration of IrPS was 1.3 × 10−5 M in degassed

THF as in the experiments with TEA (see previous). The Fe-cat concentration was varied between 3.7 × 10−6 M and 1 ×10−4 M. The obtained Stern−Volmer plots of the quantumyield and the lifetime are again in quite good agreement (Figure3a) yielding values of (7.8 and 5.3) × 1010 M−1 s−1 for kq,respectively. Hence, the averaged quenching rate constant is 6.6

× 1010 M−1 s−1. Thus, in degassed THF, the quenching by theFe-cat is 1000 times more efficient than quenching by TEA.Because of the charges of each partner of this quenching pair,the Debye equation25,26 has to be applied for the theoreticaldiffusion rate kD,i, resulting in a value of 1.5 × 1011 M−1 s−1.Hence, quenching by the Fe-cat seems to be nearly diffusion-limited. ET that would be favorable for photocatalysis can beexcluded due to the reduction potentials. The reductionpotential of the excited state of the IrPS is with −0.82 Vversus NHE (Scheme SI-1 in the Supporting Information), notlarge enough to reduce the Fe-cat species [HFe3(CO)11]

−

(−1.09 V vs NHE). The calculations, applying the same schemeas before for the IrPS/Fe-cat system at a distance of R = 10 Å,also suggest that the Fe-cat is neither oxidized nor reduced bythe excited IrPS, because local states are always lower in energythan the CT ones. Thus, in accordance with other photo-catalytic systems based on IrPS as PS,10 the oxidative ETmechanism does not play a role in our investigated system.Note that the reduced IrPS, in contrast with the oxidized formdiscussed here, is capable of reducing the Fe-cat, which is aprerequisite for catalytic water splitting in our system. Mostlikely, the efficient quenching of the PL of IrPS* by the Fe-catis due to an energy transfer, which is known for quite a fewsimilar iridium complexes in multicomponent systems.27−29

According to eq 3

+ ‐ → + ‐T S S TIrPS( ) Fe cat( ) IrPS( ) Fe cat( )x1 0 0 (3)

and to the calculated energies of the involved species, theenergy of the lowest singlet−triplet transition for IrPS is largerthan the energies of several of the lowest singlet−triplettransitions of the Fe-cat with a maximal energy difference of 1.2eV (Figure 3c). Hence, the spectral overlap between theabsorption of the Fe-cat and the PL of the IrPS, the energies ofthe possibly involved states, and the quenching rate close todiffusion limit are strong indications for a triplet−triplet energytransfer of Dexter type.30 In consequence, quenching of IrPS*by the Fe-cat is an energy-loss process in this photocatalyticsystem.Photocatalytic Half-Cell. The PL lifetimes τPL of IrPS obtained

for different combinations of the components in Scheme 1 withconcentration ratios similar to the photocatalytic system11 aresummarized in Table 1. For comparison, a typical catalysisexperiment in ref 11 was performed by employing 7.5 × 10−4·M IrPS. The IrPS/Fe3-species concentration ratio was variedbetween 0.4 and 4.8, and a ratio of 4:1:1 for the THF/TEA/water solvent mixture. The highest turnover number for thechosen Fe3-species was observed at a concentration ratio of0.62 of IrPS/Fe3-species. In our experiments, we chosecomparable concentration ratios by employing an 1.3 × 10−5

M IrPS concentration, an IrPS/Fe-cat ratio of 0.8, and a ratio of4:1:1 for the THF/TEA/water solvent mixture.At a TEA concentration of 17 vol %, PL lifetime is only

marginally affected by adding the Fe-cat, for example, 13 versus14 ns in THF/TEA and 42 ns in THF/TEA/water for both theabsence and presence of the Fe-cat. Thus, the loss process dueto the energy transfer from the IrPS* to the Fe-cat plays only aminor role in the overall photocatalytic system. Considering theconcentration ratios within this system with a huge excess ofSR, this is not very surprising. Even though the ET probabilityper collision with TEA is very low, this process outruns theenergy transfer to the Fe-cat due to the high collision rate.Furthermore, the PL lifetime of IrPS in presence of Fe-cat isincreased from 14 to 42 ns by the addition of water in THF/

Figure 3. (a) Stern−Volmer plots: Ratios of IrPS luminescencelifetimes (triangles) and quantum yields (squares) in dependence ofFe-cat concentration in degassed THF solution and linear fit curves forthe lifetimes (solid line) and quantum yields (dashed lines). Thedetermined quenching rate constant is based on the mean value forthese two slopes according to eq 1. (b) UV/vis absorption spectra ofIrPS (c = 1.3 × 10−5 M, black line), Fe-cat (c = 4 × 10−5 M, blue line),and normalized luminescence spectrum of IrPS in degassed THF(black dashed line, λexc = 388 nm). (c) Schematic representation ofenergetic positions of IrPS and Fe-cat local triplet states as predictedby theoretical calculations. The rectangles correspond to the variationof excitation energies depending on the mutual orientation of IrPS andFe-cat.

The Journal of Physical Chemistry Letters Letter

dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−13601358

TEA and decreased from 280 to 210 ns by the addition of waterin pure THF (Table 1). This is in accordance with the Stern−Volmer constants determined in the TEA quenching experi-ments with and without water (Figure 1b).In conclusion, the presented results on PL quenching show

that the ET from TEA to the photoexcited IrPS with a yield ofonly 0.4% per collision is surprisingly inefficient. This result isexplained well by the weak binding between TEA and IrPScompared with the binding energies of the solutioncomponents water, THF and TEA with each other. In addition,the DFT/TD-DFT calculations show that an ET from TEA toIrPS is, in general, improbable. For distances of 7 to 8 Å, EToccurs only in 5−10% of all investigated geometries. It onlyhappens if the molecules accidentally collide in the rightorientation. Hence, replacement of the sacrificial reagent hastwo positive effects: (i) the ET step from the excitedphotosensitizer to the electron donor has large potential forimprovement and (ii) substituting the sacrificial electron donorby the oxidation reaction of the water splitting is essential for alarge-scale application. Photoluminescence quenching of thephotoexcited IrPS by the Fe-cat can be a loss channel for thephotocatalytic hydrogen-generating system. This is in accord-ance with previous experimental studies on this photocatalyticsystem, which showed a decrease in turnover numbers for bothFe-cat and IrPS with increasing Fe-cat amounts.11 The absenceof CT between the photoexcited IrPS and the Fe-cat does notallow an ET between these two species. Instead, triplet−tripletenergy transfer by a Dexter-type mechanism takes place, asindicated by the high quenching efficiency and the spectraloverlap of the emission and absorption spectra of IrPS and Fe-cat. In the overall photocatalytic system, this loss channel isminimized by the chosen conditions for the photocatalytichydrogen generation, for example, the concentration ratios. Inaddition, modified photocatalytic systems with employedphosphines and different ligand substituents of the IrPS reachincident photon to hydrogen yields up to 16%,12,13 which arerather high for organometallic PS despite the unfavorablereaction steps. Our results show that for further efficiencyimprovement, the pathway of the absorbed light to the catalysthas to be modified. Hence, keeping the here elucidated energyloss channel ineffective might be an issue in future designs.

■ ASSOCIATED CONTENT

*S Supporting InformationExperimental and computational details. Calculation ofdiffusion rate constants and determination of oxidation andreduction potentials. This material is available free of charge viathe Internet at http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Authors*E-mail: antje.neubauer@uni-rostock.de.*E-mail: sergey.bokarev@uni-rostock.de.

NotesThe authors declare no competing financial interests.

■ ACKNOWLEDGMENTS

We acknowledge the financial support for the project“Light2Hydrogen” by the BMBF program “Spitzenforschung& Innovation in den Neuen Landern”.

■ REFERENCES(1) Teets, T. S.; Nocera, D. G. Photocatalytic Hydrogen Production.Chem. Commun. 2011, 47, 9268−9274.(2) Maeda, K.; Domen, K. Photocatalytic Water Splitting: RecentProgress and Future Challenges. J. Phys. Chem. Lett. 2010, 1, 2655−2661.(3) Osterloh, F. E. Inorganic Materials as Catalysts for PhotochemicalSplitting of Water. Chem. Mater. 2008, 20, 35−54.(4) Eckenhoff, W. T.; McNamara, W. R.; Du, P.; Eisenberg, R. CobaltComplexes as Artificial Hydrogenases for the Reductive Side of WaterSplitting. Biochim. Biophys. Acta 2013, 1827, 958−973.(5) Losse, S.; Vos, J. G.; Rau, S. Catalytic Hydrogen Production atCobalt Centres. Coord. Chem. Rev. 2010, 254, 2492−2504.(6) Esswein, A. J.; Nocera, D. G. Hydrogen Production by MolecularPhotocatalysis. Chem. Rev. 2007, 107, 4022−4047.(7) Du, P.; Schneider, J.; Luo, G.; Brennessel, W. W.; Eisenberg, R.Visible Light-Driven Hydrogen Production from Aqueous ProtonsCatalyzed by Molecular Cobaloxime Catalysts. Inorg. Chem. 2009, 48,4952−4962.(8) Du, P.; Knowles, K.; Eisenberg, R. A Homogeneous System forthe Photogeneration of Hydrogen from Water Based on a Platinum-(II) Terpyridyl Acetylide Chromophore and a Molecular CobaltCatalyst. J. Am. Chem. Soc. 2008, 130, 12576−12577.(9) Goldsmith, J. I.; Hudson, W. R.; Lowry, M. S.; Anderson, T. H.;Bernhard, S. Discovery and High-Throughput Screening of Hetero-leptic Iridium Complexes for Photoinduced Hydrogen Production. J.Am. Chem. Soc. 2005, 127, 7502−7510.(10) Tinker, L. L.; McDaniel, N. D.; Curtin, P. N.; Smith, C. K.;Ireland, M. J.; Bernhard, S. Visible Light Induced Catalytic WaterReduction without an Electron Relay. Chem.Eur. J. 2007, 13, 8726−8732.(11) Gartner, F.; Sundararaju, B.; Surkus, A.-E.; Boddien, A.; Loges,B.; Junge, H.; Dixneuf, P. H.; Beller, M. Light-Driven HydrogenGeneration: Efficient Iron-Based Water Reduction Catalysts. Angew.Chem., Int. Ed. 2009, 48, 9962−9965.(12) Gartner, F.; Boddien, A.; Barsch, E.; Fumino, K.; Losse, S.;Junge, H.; Hollmann, D.; Bruckner, A.; Ludwig, R.; Beller, M.Photocatalytic Hydrogen Generation from Water with Iron CarbonylPhosphine Complexes: Improved Water Reduction Catalysts andMechanistic Insights. Chem.Eur. J. 2011, 17, 6425−6436.(13) Gartner, F.; Cozzula, D.; Losse, S.; Boddien, A.; Anikulmar, G.;Junge, H.; Schulz, T.; Marquet, N.; Spannenberg, A.; Gladiali, S.; et al.Synthesis, Characterization and Application of Iridium(III)-Photo-sensitizers for Catalytic Water Reduction. Chem.Eur. J. 2011, 17,6998−7006.(14) Gartner, F.; Denurra, S.; Losse, S.; Neubauer, A.; Boddien, A.;Anikulmar, G.; Spannenberg, A.; Junge, H.; Lochbrunner, S.; Blug, M.;et al. Synthesis and Characterization of New Iridium Photosensitizersfor Catalytic Hydrogen Generation from Water. Chem.Eur. J. 2012,18, 3220−3225.(15) Hollmann, D.; Gartner, F.; Ludwig, R.; Barsch, E.; Junge, H.;Blug, M.; Hoch, S.; Beller, M.; Bruckner, A. Insights into theMechanism of Photocatalytic Water Reduction by DFT-Supported in-situ EPR/Raman Spectroscopy. Angew. Chem., Int. Ed. 2011, 50,10246−10250.(16) Bokarev, S. I.; Hollmann, D.; Pazidis, A.; Neubauer, A.; Radnik,J.; Kuhn, O.; Lochbrunner, S.; Junge, H.; Beller, M.; Bruckner, A. SpinDensity Distribution after Electron Transfer from Triethylamine to an[Ir(ppy)2(bpy)]

+ Photosensitizer during Photocatalytic Water Reduc-tion. Phys. Chem. Chem. Phys. 2014, 16, 4789−4796.(17) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;Springer: New York, 2006.(18) Lowry, M. S.; Hudson, W. R.; Pascal, R. A.; Bernhard, S.Accelerated Luminophore Discovery through Combinatorial Syn-thesis. J. Am. Chem. Soc. 2004, 126, 14129−14135.(19) Wilde, A. P.; King, K. A.; Watts, R. J. Resolution and Analysis ofthe Components in Dual Emission of Mixed-Chelate/Ortho-MetalateComplexes of Iridium(III). J. Phys. Chem. 1991, 95, 629−634.

The Journal of Physical Chemistry Letters Letter

dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−13601359

(20) Ohsawa, Y.; Sprouse, S.; King, K. A.; DeArmond, M. K.; Hanck,K. W.; Watts, R. J. Electrochemistry and Spectroscopy of Ortho-Metalated Complexes of Iridium(III) and Rhodium(III). J. Phys. Chem.1987, 91, 1047−1054.(21) Ladouceur, S.; Fortin, D.; Zysman-Colman, E. EnhancedLuminescent Iridium(III) Complexes Bearing Aryltriazole Cyclo-metallated Ligands. Inorg. Chem. 2011, 50, 11514−11526.(22) Costa, R. D.; Monti, F.; Accorsi, G.; Barbieri, A.; Bolink, H. J.;Ortí, E.; Armaroli, N. Photophysical Properties of Charged Cyclo-metalated Ir(III) Complexes: A Joint Theoretical and ExperimentalStudy. Inorg. Chem. 2011, 50, 7229−7238.(23) Curtin, P. N.; Tinker, L. L.; Burgess, C. M.; Cline, E. D.;Bernhard, S. Structure-Activity Correlations Among Iridium(III)Photosensitizers in a Robust Water-Reducing System. Inorg. Chem.2009, 48, 10498−10506.(24) Bokarev, S. I.; Bokareva, O. S.; Kuhn, O. Electronic ExcitationSpectrum of the Photosensitizer [Ir(ppy)2(bpy)]

+. J. Chem. Phys. 2012,136, 214305−214310.(25) Balzani, V.; Moggi, L.; Manfrin, M. F.; Bolletta, F.; Laurence, G.S. Quenching and Sensitization Processes of Coordination Com-pounds. Coord. Chem. Rev. 1975, 15, 321−433.(26) Debye, P. Reaction Rates in Ionic Solutions. Trans. Electrochem.Soc. 1942, 82, 265−272.(27) Flamigni, L.; Barbieri, A.; Sabatini, C.; Ventura, B.; Barigelletti,F. Photochemistry and Photophysics of Coordination Compounds:Iridium. Top. Curr. Chem. 2007, 281, 143−203.(28) Dixon, I. M.; Collin, J.-P.; Sauvage, J.-P.; Flamigni, L.; Encinas,S.; Barigelletti, F. A Family of Luminescent Coordination Compounds:Iridium(III) Polyimine Complexes. Chem. Soc. Rev. 2000, 29, 385−391.(29) You, Y.; Nam, W. Photofunctional Triplet Excited States ofCyclometalated Ir(III) Complexes: Beyond Electroluminescence.Chem. Soc. Rev. 2012, 41, 7061−7084.(30) Turro, N. J. Modern Molecular Photochemistry; The Benjamin/Cummings Publishing Company, Inc: Menlo Park, CA, 1978.

The Journal of Physical Chemistry Letters Letter

dx.doi.org/10.1021/jz5004318 | J. Phys. Chem. Lett. 2014, 5, 1355−13601360