Influence of Triplet Surface Properties on Excited-State Deactivation of

Article pubs.acs.org/JPCA

Cite This: J. Phys. Chem. A 2019, 123, 5293−5299

Influence of Triplet Surface Properties on Excited-State Deactivation of Expanded Cage Bis(tridentate)Ruthenium(II) Complexes Lisa A. Fredin*,† and Petter Persson Chemistry Department, Theoretical Chemistry Division, Lund University, Box 124, SE-22100 Lund, Sweden

Downloaded via BUFFALO STATE on August 1, 2019 at 17:13:36 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Calculations of excited-state potential energy surfaces (PESs) are useful to predict key properties relating to the deactivation cascade of transition-metal complexes. Here, we first perform full free optimizations of the relevant excitedstate minima, followed by extensive two-dimensional PES calculations based on the minima of interest. Maps of the lowest triplet excited-state surfaces of two bistridentate RuIIcomplexes, [Ru(DQP)2]2+ and [Ru(DQzP)2]2+, are used to explain recent experimental findings including an unexpected order of magnitude difference in an excited-state lifetime. The calculations reveal significant differences in the fundamental shapes and spin transitions of the lowest triplet excited-state energy surfaces of the two complexes and, in particular, show that the metal-to-ligand charge transfer (MLCT) surface region of [Ru(DQzP)2]2+ with a shorter excited-state lifetime is significantly smaller than that of [Ru(DQP)2]2+. This leads to a minimum energy crossing between the triplet and singlet surfaces near the MLCT for [Ru(DQzP)2]2+ or near the MC for [Ru(DQP)2]2+. These results indicate that the experimentally observed difference in the excited-state lifetime is closely related to the set of energetically accessible 3MLCT conformations.

■

INTRODUCTION Much research effort has been applied to finding new ruthenium polypyridyl complexes for photochemical applications for several decades1 including a surge of interest in sensitization applications since O’Regan and Grätzel’s 1991 Nature paper showing >7% efficiency combining a Ru-dye with mesoporous titanium dioxide.2 Complexes like trisbidentate [Ru(bpy)3]2+ (bpy is 2,2′-bipyridine), bistridentate [Ru(tpy)2]2+ (tpy is 2,2′:6′,2″-terpyridine), and the benchmark Ru−N3 dye (cis-bis(isothiocyanato)bis(2,2′-bipyridyl-4,4′-dicarboxylato ruthenium(II)) have proved useful for photosensitization purposes,3,4 as sensors and local environment probes,5,6 and in molecular electronics applications.7 These sensitizers collect sunlight through excitation from the Ru center to the ligands in a metal-to-ligand charge transfer (MLCT) excitation and quick intersystem crossing to the triplet MLCT state.1,8−10 Relaxation of this 3MLCT state in the above-mentioned complexes leads to a stable excited-state lifetime of 250 ps to 1 μs at room temperature. Nonradiative activated decay via a metal-centered (MC) triplet state is believed to be a majority loss channel of the 3MLCT state for most RuII-complexes. This often leads to a significant loss in energy, accompanied by a significant lengthening of bonds around the metal center to stabilize the two unpaired spins (Chart 1), with no significant photochemistry occurring in the MC state.11−13 A large number of alternative complexes have been developed in the hope of increased efficiencies arising from © 2019 American Chemical Society

Chart 1. Schematic Depiction of the Excited-State Surfaces of RuII-Complexes and the Relevant Differences in Structure and Energya

Note that So often crosses the 3TMC surface, and ΔQMLCT/GS and ΔEMLCT/MC can be positive or negative.

a

increased excited-state lifetimes or better injection of the excited electron to an acceptor. The electrochemical and Received: March 28, 2019 Revised: May 31, 2019 Published: June 5, 2019 5293

DOI: 10.1021/acs.jpca.9b02927 J. Phys. Chem. A 2019, 123, 5293−5299

Article

The Journal of Physical Chemistry A

tions.24,25,41,42 Thus, we have here employed a step-by-step strategy of first conducting full free optimization of the relevant excited-state minima and then identifying relevant axes for further more extensive two-dimensional PES calculations to elucidate the shape of the energy landscape connecting the minima of interest. In this study, we investigate the lowest triplet excited-state surfaces of two experimentally examined complexes, [Ru(DQP)2]2+ and [Ru(DQzP)2]2+,26 (where DQzP is 2,6-di(quinoxaline-5-yl)pyridine) which has an order of magnitude difference in lifetime in spite of having very similar metal−ligand coordination and similar calculated 3 MC energies.

photophysical properties have been proven to be highly sensitive to the ligand structure both experimentally and computationally. Thus, the goal of higher efficiencies has been achieved by changing the ligand framework and tuning the energy levels of promising frameworks by the addition of substituents on the parent framework. The C2 axis of bistridentate RuII-complexes, such as [RuII(tpy)2]2+,14 allow these complexes to form linear assemblies that help when forming molecular arrays. However, the room-temperature excited-state lifetime of 250 ps makes [RuII(tpy)2]2+ less useful as a photosensitizer than [Ru(bpy)3]2+,1,15 with its 1 μs lifetime. The rapid 3MLCT → 3 MC conversion in terpyridine is believed to be caused by poor bite angles,11 which are unfavorable for the formation of ideal octahedral complexes. Therefore, terpyridine ligands fail to destabilize the lowest unoccupied metal d orbital relative to the lowest unoccupied ligand-centered orbital, which influences the relative stabilities of the 3MLCT and 3MC excited states negatively. Efforts to improve bistridentate complex lifetimes have targeted ligand substitutions at the 4 and 4′ positions16 and incorporation of different heterocycles in biheteroaromatic ligands, which influence both the electronic properties and the coordination geometry of the complexes.11,12,17−19 One of the most striking examples is two 8-quinolinyl groups and a pyridine ring forming the tridentate ligand 2,6-bis(8quinolinyl)-pyridine (DQP)12 with its expanded six-membered chelate rings, whose homoleptic RuII-complexes obtained nearly perfect octahedral coordination and microseconds excited-state lifetimes. Many calculations of RuII-polypyridyl complexes have been published,12,20−26 including a number of investigations of ground-state properties and vertical excitations,27 including investigations of solvent effects28,29 and counter ions,30 as well as supramolecular31 and heterogeneous interactions.32 To investigate the thermal deactivation of the 3MLCT state, it is critical to calculate the relaxed triplet-state properties beyond the Franck−Condon region, especially the 3MLCT and 3MC state relaxed minima. This is becoming more common and has been done for many RuII-complexes.21−23,33−37 Such optimizations of bistridentate RuII−polypyridyl complexes have provided ample evidence that 3MC states typically have energy minima characterized by significantly longer Ru−N bonds compared to the 3MLCT states, something which has been difficult to characterize experimentally because of the efficient nonradiative decay of the 3MC states.38 While these optimized 3 MLCT and 3MC states provide a picture of the energetic balance between the two states, further insights into the thermal 3MLCT/3MC activated transitions can be gained from theoretical exploration of the reaction profiles. The few metal complexes for which this kind of exploration has been performed show comparatively flat excited-state potential energy surfaces (PESs) where identification of formal transition states using standard transition state searches remains difficult.22,24,25,39,40 In addition, the strategy used to choose the geometric axes along which to scan the PES and relaxation of the ligand system around the metal center are critical to comprehensively capture key PES features, including the different types of minima, in a realistic fashion on such flat PESs. Calculations of RuII-bistridentate complexes have shown the importance of considering multidimensional excited-state relaxation pathways and full unrestricted geometry optimiza-

■

RESULTS AND DISCUSSION In order to understand the order of magnitude difference in the excited-state lifetimes of [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ (Chart 2), 3000 and 255 ns, respectively,26 we Chart 2. [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ structures where q1 and q2 are the averages of the blue and red bonds, respectively.

examine the structure and nature of their excited states and excited-state surfaces. The GS (electronic structure in Figures S1 and S2 and Tables S1, S8, and S9), 3MLCT, and 3MC geometries of both complexes are similar, with distortion of the ligands around the Ru in the 3MLCT resulting in a less octahedral structure (a larger octahedricity value, O) and population of a dz2 in the lowest energy fully relaxed MC state (Table 1). Each geometry was optimized freely using PBE0,43−45 which is known to match experimental geometries particularly well for Ru-complexes,18,46−48 in conjunction with standard Gaussian type orbital basis sets of double-ζ quality, 631G(d,p), and the SDD Stuttgart/Dresden effective core potential (ECP) was used to provide an ECP for Ru.49 All calculations were performed using the Gaussian 09 program50 with a total charge on the complex (2+), a complete polarizable continuum model (PCM) solvent description for acetonitrile,51 and no symmetry constraints were applied, allowing for possible Jahn−Teller effects. Projected PESs (PPESs) provide a quantitative way to map the freely optimized energy minima and the energies of other spin surfaces of interest projected along an axis of structural change. As the optimized minima are quite similar, projecting [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ along almost any axis of change, for example, average of all Ru−ligand bond distances (R), octahedricity (O, the deviation of the angels around Ru from a perfect octahedral geometry O = 0°), average dihedral angle of the ligands (D), or specific Ru−ligand bond distances from Ru−Py, Ru−Q, or bonds on one ligand or the other, produces a very similar picture. The most obvious of these projections are structural axes that separate the structures in a way that describes as much of the geometric change as 5294


Article


Table 1. Calculated Structural Properties of [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ for the GS, 3MLCT, and 3MC States Using PBE0|6-31G(d,p)-SDD|PCM(Acetonitrile)a,b [Ru(DQP)2]

2+

[Ru(DQzP)2]2+

GS 3 MLCT 3 MC GS 3 MLCT 3 MC

E (eV)

q1c

q2c

0.00 2.18 2.05 0.00 1.86 2.05

2.08 2.10 2.38 2.07 2.07 2.30

2.08 2.10 2.11 2.07 2.08 2.10

2.08 2.09 2.20 2.07 2.07 2.16

Rd

Oe

De

± ± ± ± ± ±

0.84 2.70 11.09 0.74 2.08 4.47

37.57 37.61 41.65 36.70 36.72 40.03

0.01 0.02 0.15 0.005 0.04 0.11

Distances in Å and angles in degrees. Deviations are calculated as σn-values. bCalculated from crystallographic data in ref 52. cq1 and q2 are the composite reaction coordinates, which are the average of the Ru−N distances for the 2 ligands, respectively. dR is the average of all metal coordinating atom bond distances, where the error is the standard deviation. eAverage deviation = (∑|ideal angle − measured angle|)/n, where the ideal octahedral angle (O) is 90° and dihedral (D) is 0°.

a

Figure 1. PPESs of [Ru(DQP)2]2+ (A) and [Ru(DQzP)2]2+ (B) structures projected along the Ru−Q bonds of ligand 1 (q1). Red points are optimized minima and black points are single point energies calculated at the minimum geometries. The gray lines schematically show the PPESs based on the potential energy curves calculated in Figure S4. The inset indicates the differences between calculated minima geometries in both q1 and q2.

Figure 2. Calculated 3D (A,D) and 2D (B,E) representations of the triplet potential energy and spin (C,F) surfaces of [Ru(DQP)2]2+ (A−C) and [Ru(DQzP)2]2+ (D−F) surfaces between the 3MLCT and 3MC geometries from the points in Tables S3−S6. 3D spin surfaces are shown in Figure S6.

possible. Here, we use the average Ru−quinoline bond distances on the two ligands (q1 and q2, color coded in

Chart 2) as this is a structural parameter that changes most between the GS, 3MLCT, and 3MC (Figure 1), as well as best 5295


Article


[Ru(DQP)2]2+ clearly shows a small sharp transition region between the 3MLCT (blue in Figure 2C) and the 3MC (red in Figure 2C). In contrast, the MLCT to MC transition of [Ru(DQzP)2]2+ is more gradual with a large region of mixed MLCT/MC triplets (Figures 2F and S6C). . The [Ru(DQzP)2]2+ PES (Figure 2D) shows a 2.0 eV 3MLCT region with an increase of about ∼0.1 eV to a very asymmetric 3MC. While a 3MC minimum was able to be located through a fully unconstrained optimization, the 1D scan (Figure S4) and 2D PES show basically no additional activation barrier (i.e., there is no activation energy higher than the 3MLCT/3MC energy difference) between the MLCT and MC. This points to a very destabilized MC. While the fully optimized 3MC is calculated to have populated a dz2 orbital (Figure S3), and this is the lowest energy structure in the MC region, most calculated structures with MC character on the 2D PES correspond to the population of a dx2−y2 type orbital. While an optimized dx2−y2 MC could not be realized, the slight dips in the PES at (2.20, 2.20) and (2.26, 2.26) are this type of MC state, all of which point to many possible unstable triplets with some MC character. The excited-state lifetime of a complex actually depends on a complex mix of a lot of structural and electronic structure factors.55 The calculated 3MLCT/GS energy gaps in both [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ are ∼1.8−2 eV (Table 1), matching well with the measured first transition energies (E0−0 ≈ 1.8 eV).26 This type of energy gap with the small geometric reorganization between the GS and 3MLCT indicates low radiative loss, as measured by the emission yield of each complex, 2 × 10−2 and 10−3%, respectively.26 In addition to this radiative pathway, the 3MLCT can decay nonradiatively via an activated transition to the MC state or directly to the ground state. Direct nonradiative decay is normally expected for low energy gaps and large reorganization between the 3 MLCT and GS. Both the complexes have equivalent energy gaps, within measurement and computational error, and the 3 MLCT structure of [Ru(DQzP)2]2+ only has bond increases ∼0.02 Å and ΔO and ΔD < 1° more than that for [Ru(DQP)2]2+. This suggests equivalent amounts of direct MLCT → GS nonradiative decay. Thus, the shorter excitedstate lifetime of [Ru(DQzP)2]2+ must arise for some other reason than the energy-gap law or reorganization energy. It is clear from the spin surfaces that [Ru(DQzP)2]2+’s MLCT surface area is ∼25% (a 75% difference) that of [Ru(DQP)2]2+’s, with 0.0268 and 0.1052 Å2, respectively (Figure 2B,E). This reduction in the accessible 3MLCT area for [Ru(DQzP)2]2+ is related to the smaller Ru−ligand expansion required to reach the 3MC minimum for [Ru(DQzP)2]2+ compared to [Ru(DQP)2]2+ as mentioned above for the 1D q1 scan and the fully relaxed minima on the PPES. This indicates that there is a larger set of degrees of freedom which define 3MLCT structures for [Ru(DQP)2]2+ than [Ru(DQzP)2]2+, correlating to a longer lifetime.24,39 In other words, a larger MLCT surface area correlates to more MLCT configurations and thus a longer MLCT lifetime. It is worth noting that a full quantitative treatment of the accessible MLCT states would be a multidimensional hypervolume over the complete PES; however, for these surfaces, it is clear that the spin surface area is a decent proxy for the full hypervolume. The recently measured kinetic properties of DQP-like complexes showed ten-fold reduction in the radiative and nonradiative decay directly to the ground state from the 3 MLCT, compared to those of typical complexes with similar

captures the overall structural variations. Changes to the Ru− ligand bonds in tridentate ligands are constrained by ligand back-bone twisting and ligand twisting around the Ru to accommodate the new Ru bonding without ligand loss; thus R, O, D, and q1 and q2 are correlated. On the PPES, points of similar spin type (as determined by the Mulliken spin density on Ru) are represented by schematic curves (gray Figure 1), where the crossing points of the triplet surfaces are drawn based on calculations along the q1 coordinate. A comparison of the calculated PPES’s for the two complexes shows that the [Ru(DQzP)2]2+ 3MLCT and 3 MC minima are slightly lower in energy and that the 3MC minimum is calculated to be closer to the 3MLCT geometry along the q1 axis than for [Ru(DQP)2]2+. This makes it hard to predict which complex will undergo 3MLCT → 3MC conversion more easily without more comprehensive consideration of the energy surface connecting these points. Constrained triplet geometry calculations, where the Ru−Q bond distances on one ligand are frozen at various lengths between the 1.9 and ∼2.5 Å offers a picture of the transition between 3MLCT and 3MC and thus provides important insight into the structure of the triplet surfaces and associated nonradiative decay pathways from the initially excited MLCT state. Previous scans of triplet PESs of [Ru(DQP)2]2+ using both uDFT24 and time-dependent density functional theory (TD-DFT)25 show very similar shapes, with slightly lower energies for the TD-DFT. Similarly, these slices along the PESs calculated either with uDFT or TD-DFT (Figure S4) reveal that there is virtually no barrier between the 3MLCT and 3MC for [Ru(DQzP)2]2+ and a very low barrier for [Ru(DQP)2]2+, pointing to a different MLCT → MC transition for these two complexes. Because excited-state surfaces are flatter and the triplet surfaces here include multiple spin states, each of these points is independently optimized from a guess geometry with the various Ru−Q bond lengths. This methodology provides potential energy slices where each point can be characterized as MLCT or MC based on its calculated spin while avoiding stability issues, often observed scanning a surface from either the MLCT or the MC with these starting guesses. In order to map the lowest triplet energy surface between the 3MLCT and 3MC geometries of [Ru(DQP)2]2+ and [Ru(DQzP)2]2+, we have performed extensive constrained (the only constraints are the lengths of the four Ru−Q bonds) excited-state optimizations for q1 and q2 between 1.96 and 2.50 Å (Figure 2). This set of coordinates allows us to directly compare to previous triplet surface calculations,24 including direct comparison with a previously calculated triplet surface of [Ru(DQP)2]2+ using the B3LYP functional.53,54 Each point on the PESs is optimized from a new starting guess geometry and the MLCT or MC nature of the structure is determined after the optimization based on the spin density map and Mulliken spin on the Ru. The PESs for [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ clearly show that while the freely optimized geometries of the ground and excited states and the PPESs of the two complexes are qualitatively quite similar, indicating at least one similar deactivation pathway, the broader triplet behavior of the two complexes is quite different. The triplet [Ru(DQP)2]2+ surface is fairly flat with isoenergic 3MLCT and 3MC regions and a small activated barrier (∼0.33 eV) between them (Figure 2A). This barrier matches very well with the experimental activation barrier of 0.322 eV55 and is on the same order of magnitude as the B3LYP calculated barrier of ∼0.13 eV.24 The spin map of 5296



Article

■

emission energies.55 This matches well with the fact that beyond the small low energy region, [Ru(DQP)2]2+ MLCT geometries are more likely to decay through activation to the MC state because of the relatively low MLCT/MC energy difference (0.1 eV) and the low (0.33 eV) activation barrier between the two states. In addition, the minimum energy crossing point (MECP) optimized using easyMECP56 and Gaussian 0950 (Table S2 and Figure S5) for [Ru(DQP)2]2+ is between the 3MC and singlet surface with further expanded Ru−ligand bonds. The initial absorption in Ru-complexes is an excitation to the 1MLCT state. Because emission is observed from the 3 MLCT, it makes sense that complexes with larger 1 MLCT/3MLCT mixing, and thus large extinction coefficients, have faster radiative decay back to the ground state. Abrahamsson et al. observed a decrease in both radiative and nonradiative decay from the MLCT directly to the ground state in DQP derivatives, suggesting that they have unusually small singlet−triplet mixing.55 While [Ru(DQzP)2]2+ was designed as a DQP derivative, it is clear that the change to the ligand backbone, as opposed to just R-group modification, has changed the Ru−ligand interaction significantly, and thus we would expect the 1 MLCT/3MLCT to be more like that of polypyridyl complexes. In particular, the large mixed MLCT/MC region on the [Ru(DQzP)2]2+ triplet surface, along with a destabilized MC (no true minimum) suggests that nonradiative decay from a mixed MLCT/MC triplet is its majority deactivation path. This is supported by a minimum energy crossing between the triplet and singlet surfaces close to the 3MLCT minimum for [Ru(DQzP)2]2+ (Figure S5 and Table S2). Thus, the degree of MLCT/MC mixing in the triplet that undergoes nonradiative decay seems to promote more nonradiative decay, and activation to a pure MC state correlates to a longer lifetime. Additionally, multidimensional surfaces have so far only been computed for a limited number of metal-centered complexes,39,40 largely precluding studies directly targeting the effect of functional until now. By recalculating the [Ru(DQP)2]2+ surface in order to make it directly comparable to [Ru(DQzP)2]2+, we can in turn directly compare the PES structure and features across two hybrid functionals (Figure S7). We also reoptimized the minima with dispersion corrections (D3-PBE0) but did not see any significant changes to the geometry, electronic structure, or energies (Table S7) to warrant the significant increase in computational cost. Hybrid functionals have been seen to have varying effects on the energetics with only small differences in the geometries of fully relaxed excited states.38 The most obvious difference in their structures is the double MLCT and MC minima of the B3LYP and PBE0 surfaces, respectively. While quantitatively the MLCT spin area for B3LYP is about 50% that of PBE0, 0.052 and 0.105 Å2, respectively, the low energy MLCT areas are very similar, 0.023 and 0.034 Å2, respectively. Both lowest triplet surfaces have a large area MLCT region which includes many asymmetrical MLCT geometries, a near isoenergic MC minimum, with the lowest barrier path between them being highly asymmetric and the more symmetric path having a higher activated transition barrier, indicating little effect from different hybrid functionals on the understanding into deactivation pathways provided by multidimensional PESs of photoexcited transition-metal complexes.

CONCLUSIONS A better understanding of how subtle structural changes associated with ligand modifications and substitutions affect the electronic structure and excited states of Ru-sensitizers is needed to move past the mostly trial-and-error methods still employed to find better dyes. While increasing the octahedricity has improved excited-state lifetimes over traditional polypyridyl complexes, it is still hard to predict why complexes with extremely similar Ru−ligand binding, electronic structures, and relaxed excited states have very different measured excited-state lifetimes and properties. The calculated 3 MLCT/GS energy gaps in both [Ru(DQP)2]2+ and [Ru(DQzP)2]2+, and their relatively small differences in 3MLCT/ GS geometry are in line with their very low radiative emission. However, the lowest triplet excited-state surfaces of [Ru(DQP)2]2+ and [Ru(DQzP)2]2+ help to explain their order of magnitude difference in lifetime beyond what can be suggested by simply considering either vertical excitations or the energy balance between optimized 3MLCT and 3MC structures alone. In particular, the clearest differences in calculated excited-state properties for the two complexes are only fully evident in the extensive 2D surfaces. The calculations therefore suggest that the significant experimentally observed difference in excited lifetime between the two complexes investigated here is best explained by the qualitatively and quantitatively different shapes and spin characteristics of the lowest triplet energy surfaces, where the shorter excited state lifetime is associated with a smaller set of available 3MLCT conformations and vice versa. In addition, a comparison of PESs calculated with two common hybrid functionals show for the first time how the relative structures of calculated multidimensional surfaces vary with hybrid functionals. Overall, calculating multidimensional excited-state surfaces is critical to elucidate key aspects of excited-state cascades in photoactive transition-metal complexes and predict differences in excited-state lifetimes.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.9b02927. Additional computational details including the spin densities, electronic structure diagrams, redox energies, points optimized on the restricted PES, and the complete list of TD-DFT calculated transitions (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Lisa A. Fredin: 0000-0002-4091-0899 Petter Persson: 0000-0001-7600-3230 Present Address †

Department of Chemistry, Lehigh University, 6 East Packer Ave, Bethlehem, PA 18015, USA. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Prof. Maria Abrahamsson, Chalmers University, for invaluable discussions. We acknowledge the Crafoord Foundation, the Swedish Research Council, and the Knut 5297


Article


(16) Hecker, C. R.; Gushurst, A. K. I.; McMillin, D. R. Phenyl Substituents and Excited-State Lifetimes in Ruthenium(II) Terpyridyls. Inorg. Chem. 1991, 30, 538−541. (17) Halcrow, M. A. Recent Advances in the Synthesis and Applications of 2,6-Dipyrazolylpyridine Derivatives and Their Complexes. New J. Chem. 2014, 38, 1868−1882. (18) Hedberg Wallenstein, J.; Fredin, L. A.; Jarenmark, M.; Abrahamsson, M.; Persson, P. Chemical consequences of pyrazole orientation in RuII complexes of unsymmetric quinoline-pyrazole ligands. Dalton Trans. 2016, 45, 11723−11732. (19) Metherell, A. J.; Cullen, W.; Stephenson, A.; Hunter, C. A.; Ward, M. D. Fac and mer isomers of Ru(ii) tris(pyrazolyl-pyridine) complexes as models for the vertices of coordination cages: structural characterisation and hydrogen-bonding characteristics. Dalton Trans. 2014, 43, 71−84. (20) Abrahamsson, M.; Jäger, M.; Kumar, R. J.; Ö sterman, T.; Persson, P.; Becker, H.-C.; Johansson, O.; Hammarströ m, L. Bistridentate Ruthenium(II)polypyridyl-Type Complexes with Microsecond3MLCT State Lifetimes: Sensitizers for Rod-Like Molecular Arrays. J. Am. Chem. Soc. 2008, 130, 15533−15542. (21) Abrahamsson, M.; Lundqvist, M. J.; Wolpher, H.; Johansson, O.; Eriksson, L.; Bergquist, J.; Rasmussen, T.; Becker, H.-C.; Hammarström, L.; Norrby, P.-O.; et al. Steric Influence on the Excited-State Lifetimes of Ruthenium Complexes with Bipyridyl− Alkanylene−Pyridyl Ligands. Inorg. Chem. 2008, 47, 3540−3548. (22) Borg, O. A.; Godinho, S. S. M. C.; Lundqvist, M. J.; Lunell, S.; Persson, P. Computational Study of the Lowest Triplet State of Ruthenium Polypyridyl Complexes Used in Artificial Photosynthesis. J. Phys. Chem. A 2008, 112, 4470−4476. (23) Dixon, I. M.; Alary, F.; Heully, J.-L. Electronic Peculiarities of the Excited States of [RuN5C]+vs. [RuN6]2+ Polypyridine Complexes: Insight from Theory. Dalton Trans. 2010, 39, 10959− 10966. (24) Ö sterman, T.; Abrahamsson, M.; Becker, H.-C.; Hammarström, L.; Persson, P. Influence of Triplet State Multidimensionality on Excited State Lifetimes of Bis-tridentate RuII Complexes: A Computational Study. J. Phys. Chem. A 2012, 116, 1041−1050. (25) Ö sterman, T.; Persson, P. Excited state potential energy surfaces of bistridentate RuII complexes - A TD-DFT study. Chem. Phys. 2012, 407, 76−82. (26) Parada, G. A.; Fredin, L. A.; Santoni, M.-P.; Jäger, M.; Lomoth, R.; Hammarström, L.; Johansson, O.; Persson, P.; Ott, S. Tuning the Electronics of Bis(tridentate)ruthenium(II) Complexes with LongLived Excited States: Modifications to the Ligand Skeleton Beyond Classical Electron Donor or Electron Withdrawing Group Decorations. Inorg. Chem. 2013, 52, 5128−5137. (27) Vlček, A., Jr.; Záliš, S. Modeling of Charge-Transfer Transitions and Excited States in d6 Transition Metal Complexes by DFT Techniques. Coord. Chem. Rev. 2007, 251, 258−287. (28) Fantacci, S.; De Angelis, F.; Selloni, A. Absorption Spectrum and Solvatochromism of the [Ru(4,4’-COOH-2,2’-bpy)2(NCS)2] Molecular Dye by Time Dependent Density Functional Theory. J. Am. Chem. Soc. 2003, 125, 4381−4387. (29) Guillemoles, J.-F.; Barone, V.; Joubert, L.; Adamo, C. A Theoretical Investigation of the Ground and Excited States of Selected Ru and Os Polypyridyl Molecular Dyes. J. Phys. Chem. A 2002, 106, 11354−11360. (30) Lundqvist, M. J.; Galoppini, E.; Meyer, G. J.; Persson, P. Calculated Optoelectronic Properties of Ruthenium Tris-bipyridine Dyes Containing Oligophenyleneethynylene Rigid Rod Linkers in Different Chemical Environments. J. Phys. Chem. A 2007, 111, 1487− 1497. (31) Meylemans, H. A.; Damrauer, N. H. Controlling Electron Transfer through the Manipulation of Structure and Ligand-Based Torsional Motions: A Computational Exploration of Ruthenium Donor−Acceptor Systems using Density Functional Theory. Inorg. Chem. 2009, 48, 11161−11175.

and Alice Wallenberg Foundation, the Swedish Energy Agency (Energimyndigheten), as well as the Swedish Strategic Research Foundation (SSF) for funding as well as supercomputing facilities support from the NSC and LUNARC. L.A.F. acknowledges research computing resources provided by Lehigh University.

■

REFERENCES

(1) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; von Zelewsky, A. Ru(II) Polypyridine Complexes: Photophysics, Photochemistry, Eletrochemistry, and Chemiluminescence. Coord. Chem. Rev. 1988, 84, 85−277. (2) O’Regan, B.; Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal TiO2 Films. Nature 1991, 353, 737−740. (3) Nazeeruddin, M. K.; Klein, C.; Liska, P.; Grätzel, M. Synthesis of Novel Ruthenium Sensitizers and Their Application in Dye-Sensitized Solar Cells. Coord. Chem. Rev. 2005, 249, 1460−1467. (4) Bomben, P. G.; Robson, K. C. D.; Koivisto, B. D.; Berlinguette, C. P. Cyclometalated Ruthenium Chromophores for the DyeSensitized Solar Cell. Coord. Chem. Rev. 2012, 256, 1438−1450. (5) Friedman, A. E.; Chambron, J. C.; Sauvage, J. P.; Turro, N. J.; Barton, J. K. A Molecular Light Switch for DNA: Ru(bpy)2(dppz)2+. J. Am. Chem. Soc. 1990, 112, 4960−4962. (6) Gill, M. R.; Thomas, J. A. Ruthenium(ii) polypyridyl complexes and DNA-from structural probes to cellular imaging and therapeutics. Chem. Soc. Rev. 2012, 41, 3179−3192. (7) King, A. W.; Wang, L.; Rack, J. J. Excited State Dynamics and Isomerization in Ruthenium Sulfoxide Complexes. Acc. Chem. Res. 2015, 48, 1115−1122. (8) Campagna, S.; Puntoriero, F.; Nastasi, F.; Bergamini, G.; Balzani, V. Photochemistry and Photophysics of Coordination Compounds: Ruthenium. In Photochemistry and Photophysics of Coordination Compounds I; Balzani, V., Campagna, S., Eds.; Springer Berlin Heidelberg: Berlin, 2007; pp 117−214. (9) McCusker, J. K. Femtosecond Absorption Spectroscopy of Transition Metal Charge-Transfer Complexes. Acc. Chem. Res. 2003, 36, 876−887. (10) Ponseca, C. S.; Chábera, P.; Uhlig, J.; Persson, P.; Sundström, V. Ultrafast Electron Dynamics in Solar Energy Conversion. Chem. Rev. 2017, 117, 10940−11024. (11) Abrahamsson, M.; Becker, H.-C.; Hammarströ m , L.; Bonnefous, C.; Chamchoumis, C.; Thummel, R. P. Six-Membered Ring Chelate Complexes of Ru(II): Structural and Photophysical Effects. Inorg. Chem. 2007, 46, 10354−10364. (12) Abrahamsson, M.; Jäger, M.; Ö sterman, T.; Eriksson, L.; Persson, P.; Becker, H.-C.; Johansson, O.; Hammarström, L. A 3.0 μs Room Temperature Excited State Lifetime of a Bistridentate RuII− Polypyridine Complex for Rod-like Molecular Arrays. J. Am. Chem. Soc. 2006, 128, 12616−12617. (13) Abrahamsson, M.; Wolpher, H.; Johansson, O.; Larsson, J.; Kritikos, M.; Eriksson, L.; Norrby, P.-O.; Bergquist, J.; Sun, L.; Åkermark, B.; et al. A New Strategy for the Improvement of Photophysical Properties in Ruthenium(II) Polypyridyl Complexes. Synthesis and Photophysical and Electrochemical Characterization of Six Mononuclear Ruthenium(II) Bisterpyridine-Type Complexes. Inorg. Chem. 2005, 44, 3215−3225. (14) Maestri, M.; Armaroli, N.; Balzani, V.; Constable, E. C.; Thompson, A. M. W. C. Complexes of the Ruthenium(II)-2,2’:6’,2’’Terpyridine Family. Effect of Electron-Accepting and -Donating Substituents on the Photophysical and Electrochemical Properties. Inorg. Chem. 1995, 34, 2759−2767. (15) Hammarström, L.; Barigelletti, F.; Flamigni, L.; Indelli, M. T.; Armaroli, N.; Calogero, G.; Guardigli, M.; Sour, A.; Collin, J.-P.; Sauvage, J.-P. A Study on Delocalization of MLCT Excited States by Rigid Bridging Ligands in Homometallic Dinuclear Complexes of Ruthenium(II). J. Phys. Chem. A 1997, 101, 9061−9069. 5298


Article

The Journal of Physical Chemistry A (32) Persson, P.; Lundqvist, M. J. Calculated Structural and Electronic Interactions of the Ruthenium Dye N3 with a Titanium Dioxide Nanocrystal. J. Phys. Chem. B 2005, 109, 11918−11924. (33) Alary, F.; Boggio-Pasqua, M.; Heully, J.-L.; Marsden, C. J.; Vicendo, P. Theoretical Characterization of the Lowest Triplet Excited States of the Tris-(1,4,5,8-tetraazaphenanthrene) Ruthenium Dication Complex. Inorg. Chem. 2008, 47, 5259−5266. (34) Alary, F.; Heully, J.-L.; Bijeire, L.; Vicendo, P. Is the3MLCT the Only Photoreactive State of Polypyridyl Complexes? Inorg. Chem. 2007, 46, 3154−3165. (35) Heully, J.-L.; Alary, F.; Boggio-Pasqua, M. Spin-Orbit Effects on the Photophysical Properties of Ru(bpy)32+. J. Chem. Phys. 2009, 131, 184308. (36) Salassa, L.; Garino, C.; Salassa, G.; Gobetto, R.; Nervi, C. Mechanism of Ligand Photodissociation in Photoactivable [Ru(bpy)2L2]2+Complexes: A Density Functional Theory Study. J. Am. Chem. Soc. 2008, 130, 9590−9597. (37) Salassa, L.; Garino, C.; Salassa, G.; Nervi, C.; Gobetto, R.; Lamberti, C.; Gianolio, D.; Bizzarri, R.; Sadler, P. J. Ligand-Selective Photodissociation from [Ru(bpy)(4AP)4]2+: a Spectroscopic and Computational Study. Inorg. Chem. 2009, 48, 1469−1481. (38) Leshchev, D.; Harlang, T. C. B.; Fredin, L. A.; Khakhulin, D.; Liu, Y.; Biasin, E.; Laursen, M. G.; Newby, G. E.; Haldrup, K.; Nielsen, M. M.; et al. Tracking the Picosecond Deactivation Dynamics of a Photoexcited Iron Carbene Complex by TimeResolved X-ray Scattering. Chem. Sci. 2018, 9, 405−414. (39) Fredin, L. A.; Wärnmark, K.; Sundström, V.; Persson, P. Molecular and Interfacial Calculations of Iron(II) Light Harvesters. ChemSusChem 2016, 9, 667−675. (40) Soupart, A.; Alary, F.; Heully, J.-L.; Elliott, P. I. P.; Dixon, I. M. Exploration of Uncharted 3PES Territory for [Ru(bpy)3]2+: A New 3MC Minimum Prone to Ligand Loss Photochemistry. Inorg. Chem. 2018, 57, 3192−3196. (41) Graetzel, M.; Janssen, R. A. J.; Mitzi, D. B.; Sargent, E. H. Materials Interface Engineering for Solution-Processed Photovoltaics. Nature 2012, 488, 304−312. (42) Magnuson, A.; Anderlund, M.; Johansson, O.; Lindblad, P.; Lomoth, R.; Polivka, T.; Ott, S.; Stensjö, K.; Styring, S.; Sundström, V.; et al. Biomimetic and Microbial Approaches to Solar Fuel Generation. Acc. Chem. Res. 2009, 42, 1899−1909. (43) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (44) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple [Phys. Rev. Lett. 77, 3865 (1996)]. Phys. Rev. Lett. 1997, 78, 1396. (45) Adamo, C.; Barone, V. Toward Reliable Density Functional Methods Without Adjustable Parameters: The PBE0 Model. J. Chem. Phys. 1999, 110, 6158−6170. (46) Lundqvist, M. J.; Galoppini, E.; Meyer, G. J.; Persson, P. Calculated Optoelectronic Properties of Ruthenium Tris-bipyridine Dyes Containing Oligophenyleneethynylene Rigid Rod Linkers in Different Chemical Environments. J. Phys. Chem. A 2007, 111, 1487− 1497. (47) Persson, P.; Knitter, M.; Galoppini, E. Light-Harvesting and Electronic Contacting Capabilities of Ru(II) Ipa Rod and Star Complexes-First Principles Predictions. RSC Adv. 2012, 2, 7868− 7874. (48) Jarenmark, M.; Fredin, L. A.; Hedberg, J. H. J.; Doverbratt, I.; Persson, P.; Abrahamsson, M. A Homoleptic Trisbidentate Ru(II) Complex of a Novel Bidentate Biheteroaromatic Ligand Based on Quinoline and Pyrazole Groups: Structural, Electrochemical, Photophysical, and Computational Characterization. Inorg. Chem. 2014, 53, 12778−12790. (49) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. Energy-adjustedabinitiopseudopotentials for the first row transition elements. J. Chem. Phys. 1987, 86, 866−872. (50) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, 2009.

(51) Tomasi, J.; Mennucci, B.; Cammi, R. Quantum Mechanical Continuum Solvation Models. Chem. Rev. 2005, 105, 2999−3094. (52) Liu, Y.; Harlang, T.; Canton, S. E.; Chábera, P.; SuárezAlcántara, K.; Fleckhaus, A.; Vithanage, D. A.; Göransson, E.; Corani, A.; Lomoth, R.; et al. Towards Longer-Lived Metal-to-Ligand Charge Transfer States of Iron(ii) Complexes: an N-heterocyclic Carbene Approach. Chem. Commun. 2013, 49, 6412−6414. (53) Perdew, J. P. Density-Functional Approximation for the Correlation Energy of the Inhomogeneous Electron Gas. Phys. Rev. B 1986, 33, 8822−8824. (54) Becke, A. D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652. (55) Abrahamsson, M.; Becker, H.-C.; Hammarström, L. Microsecond 3MLCT Excited State Lifetimes in Bis-Tridentate Ru(II)complexes: Significant Reductions of Non-Radiative Rate Constants. Dalton Trans. 2017, 46, 13314−13321. (56) Rodríguez-Guerra, J. easyMECP, GitHub repository, 2019. https://github.com/jaimergp/easymecp (accessed May 9, 2019).

5299


Influence of Triplet Surface Properties on Excited-State Deactivation of

Recommend Documents