Variation of the Ultrafast Fluorescence Quenching in 2,6-Sulfanyl-Core

J. Phys. Chem. A 2010, 114, 12555–12560

12555

Variation of the Ultrafast Fluorescence Quenching in 2,6-Sulfanyl-Core-Substituted Naphthalenediimides by Electron Transfer Igor Pugliesi,† Patrizia Krok,†,‡ Stefan Lochbrunner,§ Alfred Błaszczyk,| Carsten von Ha¨nisch,| Marcel Mayor,*,|,⊥ and Eberhard Riedle*,† Lehrstuhl fu¨r BioMolekulare Optik, Ludwig-Maximilians-UniVersita¨t Mu¨nchen, Germany, Institut fu¨r Physik, UniVersita¨t Rostock, Germany, Institute of Nanotechnology, Karlsruhe Institute of Technology (KIT), Germany, and Department fu¨r Chemie, UniVersita¨t Basel, Switzerland ReceiVed: August 16, 2010; ReVised Manuscript ReceiVed: October 1, 2010

The ultrafast fluorescence quenching of 2,6-sulfanyl-core-substituted naphthalenediimides was investigated by transient spectroscopy. We find a strong dependence of the relaxation on the chemical structure of the substituent. Direct linking of an aryl rest to the sulfur atom leads to a strong red shift of the fluorescence in 1 ps and the disappearance of the emission in 5-7 ps depending on the polarity and viscosity of the solvent. This complex behavior is interpreted with the help of quantum chemical calculations. The calculations suggest that the initial relaxation corresponds to a planarization of the substituents and an associated partial electron transfer. This is followed by a twisting of the phenylsulfanyl substituents out of the molecular plane that allows a complete localization of the electron-donating orbital on the aryl group. Finally the back transfer happens in another 5-7 ps. For an additional methylene spacer group between the sulfur and the aryl, this sequence of relaxation steps is not possible and a simple exponential decay, slower by about 1 order of magnitude, is found. I. Introduction Naphthalenediimides (NDIs) are versatile building blocks in supramolecular structures and organic materials1,2 with potential applications as molecular sensors,3 photomolecular switches, photosynthetic cascades,4,5 conducting polymers, and p-dopants.6-8 They are stable chromophores with intense optical transitions and potentially high fluorescence quantum yields for selected derivatives. At the same time they are good electron acceptors and widely used in molecular arrays.9-12 Various functionalizations of the NDI molecules are possible, as two different sites for substitution are offered. Substitution on the N-termini of the dicarboxy imide group allows for the control of physicochemical and structural properties like solubility and aggregation. Due to the large spin density present on the naphthyl group, core-substitution allows for the tuning of the HOMO-LUMO gap. This is of particular interest to molecular engineers as it introduces control of the optical properties and the energetics relevant for energy and electron transfer processes.13,14 This development was fueled by new efficient strategies for the synthesis of core-substituted NDIs13-15 and led for example to NDI-based rigid rod π-stack architectures in supramolecular heterojunctions.16,17 Efficient charge separation in the picosecond range and length-dependent charge recombination make these multichromophoric NDI systems interesting building blocks for complex photoactive devices.18 Some of the core-substituted NDIs exhibit extremely low fluorescence yields well below 1%, * Authors to whom correspondence should be addressed: Eberhard Riedle, [email protected]; Marcel Mayor, marcel.mayor@ unibas.ch. † Ludwig-Maximilians-Universita¨t Mu¨nchen. ‡ Present address: Laser Research Institute, University of Stellenbosch, South Africa. § Universita¨t Rostock. | Karlsruhe Institute of Technology. ⊥ Universita¨t Basel.

while others emit with yields above 50%,14,19 which has strong implications for the intended applications. It was suggested that conformational changes and electron transfer processes might be responsible for the varying fluorescence yields.13,14,19 A detailed notion and deeper understanding are lacking and therefore chemical design strategies for the control of the quantum yield are still missing. In this paper we compare the ultrafast photoinduced dynamics of a set of 2,6-sulfanyl-core-substituted NDIs and obtain a clear picture of the mechanism responsible for the fluorescence quenching. With the help of ab initio calculations, we characterize the properties of the relevant molecular orbitals. The theoretical results support our notion of the mechanism and reveal the impact of the shape and spatial extension of the orbitals. It turns out that the associated conformation dependent localization and delocalization of the electron density is the key to understand the underlying mechanism. The paper is organized as follows. In the second section we introduce briefly the experimental and theoretical methods and specify the compounds. In the third section the observed ultrafast dynamics of the various compounds is described and interpreted and the results of the calculations are presented. In the final section we discuss the implication of the quenching mechanism for the design of functional NDIs. II. Methods and Materials The broadband transient spectrometer has been described in full detail in ref 20. For the reported experimental data, pump pulses at 520 nm with 100 to 150 nJ energy were focused to a diameter of about 150 µm for molecular excitation. A continuum spanning from below 300 to 720 nm and polarized at the magic angle was used as probe light. The typical temporal resolution was 100 fs, well below all observed decay rates. The samples were contained in 1 mm path length fused silica cells at a

10.1021/jp107742x  2010 American Chemical Society Published on Web 11/10/2010

12556

J. Phys. Chem. A, Vol. 114, No. 48, 2010

Pugliesi et al.

CHART 1: Structure of the Investigated Naphthalenediimide Model Compounds: Ph-NDI-Ph (N,N′-Bis(p-tert-butylphenyl)-2,6-bis(p-tertbutylphenylsulfanyl)-1,4,5,8-tetracarboxylic Acid Naphthalenediimide), Bn-NDI-Bn (N,N′-Bis(p-tert-butylphenyl)-2,6-bis(p-tertbutylbenzylsulfanyl)-1,4,5,8-tetracarboxylic Acid Naphthalenediimide), A-NDI-Ph (N,N′-Bis(p-tertbutylphenyl)-2-(p-tert-butylphenylsulfanyl)-6(isopentylsulfanyl)-1,4,5,8-tetracarboxylic Acid Naphthalenediimide), and A-NDI-A (N,N′-Bis(p-tert-butylphenyl)-2,6-bis(isopentlysulfanyl)1,4,5,8-tetracarboxylic Acid Naphthalenediimide)

Figure 1. Solid-state structure of Ph-NDI-Ph, ellipsoids at the 60% probability level.

concentration of 0.2 mM for an optical density of about 0.5 at the excitation wavelength. The four molecules investigated are shown in Chart 1 and their synthesis is described in detail in the Supporting Information. Substitution of the chlorines of the common precursor Cl-NDI-Cl afforded the desired dyes. To provide the required solubility of these differently core-substituted planar π-systems for spectroscopic investigations, substituents comprising tertbutyl groups have been chosen. While in the case of PhNDI-Ph, the p-tert-butylphenyl group is directly linked to the sulfur atom at the NDI core, an additional methylene group (changing the phenyl to a benzyl moiety) separates these subunits in the case of Bn-NDI-Bn (compare Chart 1). The asymmetrically substituted dye A-NDI-Ph has been assembled in two subsequent substitution steps and comprises only one directly linked p-tert-butylphenyl group and an aliphatic chain at the opposite side which is expected to be passive as far as quenching mechanisms are concerned. To test this assumption also the A-NDI-A with only aliphatic chains is considered. All new compounds have been fully characterized by 1H and 13 C NMR spectroscopy, MALDI-TOF-mass spectrometry, and elemental analysis. In addition, single crystals suitable for X-ray analysis have been obtained for Ph-NDI-Ph by slow evaporation of a chloroform solution. Ph-NDI-Ph crystallizes in the triclinic space-group P-1 with four solvent molecules per formula unit.21 In the solid-state structure of Ph-NDI-Ph (Figure 1), all four p-tert-butylphenyl substituents are found with torsion angles above 70° rather perpendicular to the plane of the central NDI core. For reasons that will become clear later, ab initio calculations were only performed on Ph-NDI-Ph. To reduce computational time, a suitable analogue of Ph-NDI-Ph was generated. The tert-butyl groups of the various core-substituents and the tertbutylphenyl substituents on the N-termini have been replaced by hydrogens. Full neutral ground (S0) state geometry optimizations have been carried out using the DFT methodology and the B3-LYP functional. To ensure that true minima were found, additional frequency calculations were performed. For the first

singlet excited (S1) state optimizations, the TDDFT methodology and the B3-LYP functional were used. Structures were optimized following analytic gradients. Here, due to computational limitations, no frequencies were computed. The electronic excitations were calculated in the framework of linear response theory from the ground state Kohn-Sham (KS) orbitals using the adiabatic local density approximation (ALDA) for the functional derivatives of the exchange-correlation potential. In the calculations presented here only singlet excitations have been considered. The Karlsruhe split valence basis set22,23 augmented with a polarization function (SVP) was used in all calculations, which have been performed using the DFT and TDDFT24 routines implemented in the TURBOMOLE 5.9.1 ab initio suite.25,26 III. Results and Discussion To investigate the electronic relaxation mechanism of coresubstituted NDIs, we compare the photoinduced dynamics of Ph-NDI-Ph, which shows negligible fluorescence, with the weakly but significantly fluorescing Bn-NDI-Bn. In both molecules phenyl groups are attached to the sulfur atoms of the naphthalene core, however in Bn-NDI-Bn with an additional methylene spacer (Chart 1). Figure 2 shows the evolution of the transient absorption spectrum of Ph-NDI-Ph in chloroform after excitation at 520 nm. We observe ground-state bleach (GSB) between 480 and 550 nm and between 350 and 380 nm. The rest of the transient absorption spectrum is dominated by excited-state absorption (ESA). The ESA bands below 500 nm and the GSB appear within the cross-correlation time and decay synchronously but nonexponentially within 40 ps. The ESA signatures above 550 nm show a more complex behavior at short times. In the region from 550 to 665 nm a delayed rise occurs, while the excitation induced band between 665 and 720 nm decays within the first few picoseconds. In addition to these obvious features, a careful inspection of the data reveals small spectral shifts of the bands on the 10 ps time scale. These effects can be ascribed to internal vibrational redistribution (IVR) and cooling processes and result in a variation of time constants for fits at single probe wavelengths. Therefore global fitting procedures are applied in the following to extract the population dynamics. In Figure 3, the evolution of the transient spectra of PhNDI-Ph is compared to Bn-NDI-Bn. The transient absorption of Bn-NDI-Bn shows in contrast to Ph-NDI-Ph negligible spectral reshaping and a smooth decay within 100 ps. The

Variations in Ultrafast Fluorescence Quenching

J. Phys. Chem. A, Vol. 114, No. 48, 2010 12557 TABLE 1: Fit Results from Transient Absorption Data molecule

solvent

τe/τ1 (ps)

τkk/τ2 (ps)

Ph-NDI-Ph

chloroform THF chloroform

1.0 ( 0.2 1.4 ( 0.3 1.2 ( 0.3

6.5 ( 0.5 4.8 ( 0.7 7.2 ( 1.3

A-NDI-Ph

Figure 2. Photoinduced absorption changes of Ph-NDI-Ph in chloroform after 520 nm excitation.

molecule

solvent

τVR (ps)

τS1 (ps)

Bn-NDI-Bn A-NDI-A

chloroform chloroform

2.8 ( 1.1 2.1

41 ( 11 38

strong red contribution. Further support is given by timeresolved Kerr-gating experiments, which show a red shift of the emission on the same time scale as the transient absorbance changes (see Figure S2 in the Supporting Information). To quantify the relaxation dynamics, we analyze the temporal evolution of the absorbance changes using a rate model. The global fit procedure employed is described in ref 27. The fit results are listed in Table 1, and fits at selected wavelengths are shown in Figure 4 superimposed on the corresponding experimental data. The relaxation dynamics of Bn-NDI-Bn is dominated (90%) by a single exponential decay with a time constant of 41 ps. This is the excited-state lifetime. There is also a small contribution of 2.8 ps that can be ascribed to vibrational redistribution. For Ph-NDI-Ph a two-exponential model gives unsatisfactory agreement with the data. The addition of a third time constant makes the fit unstable. In particular, the two longer decay times obtained are always very similar and associated with big amplitudes of opposite sign compensating each other to a large extent. Such an instability is expected if two rates are very similar. Stable fits with a minimum set of parameters for the measured signals s(t) are obtained with the model function shown in eq 1

( )

s(t) ) Ae exp -

( )

( )

t t t + Ak exp + Ak,tt exp τe τkk τkk

(1)

Figure 3. Comparison of the transient spectra of (a) Ph-NDI-Ph and (b) Bn-NDI-Bn. For reference the stationary absorption and fluorescence spectra are shown in addition.

Ph-NDI-Ph spectra before 0.5 ps are very similar to the Bn-NDI-Bn spectra indicating that the originally populated electronic states are comparable. The steady-state fluorescence spectrum of Bn-NDI-Bn (Figure 3b) indicates that stimulated emission (SE) contributes to the transient spectra in the region from 540 to 650 nm but is mostly compensated by ESA. The Strickler-Berg symmetry with respect to the steady-state absorption spectrum shows that the fluorescence and SE result from the optically populated state. In the case of Ph-NDI-Ph the rise of the transient absorption in the region from 550 to 665 nm after excitation indicates that the original SE is quenched within the first picoseconds and the ESA becomes dominant. This implies that the population of the optically excited state is transferred to an intermediate state. The fast quenching dynamics is in agreement with the low fluorescence quantum yield of Ph-NDI-Ph of about 10-4. Furthermore, the absorption decrease above 665 nm indicates a red shift of the competing SE after optical excitation. This is reflected in the steady-state fluorescence spectrum of Ph-NDI-Ph (Figure 3a), which shows no Strickler-Berg symmetry but a

Figure 4. Temporal evolution of the absorbance changes after 520 nm excitation of (a) Ph-NDI-Ph and (b) Bn-NDI-Bn. For each detection wavelength the fitted model is shown as a gray line.

12558

J. Phys. Chem. A, Vol. 114, No. 48, 2010

Pugliesi et al.

This function is the solution of a rate model with one exponential decay with rate ke and a sequential double decay with two equal rates k ke ) 1/τe

k ) 1/τkk

k ) 1/τkk

n1 98 n2 98 n3 98 n0

(2)

Ae and τe in eq 1 are the amplitude and time constant of the first exponential decay, Ak and Ak,t the first and second amplitude of the double decay with time constant τkk. In contrast to a standard three-exponential decay model, the last pre-exponential factor in eq 1 contains the product of an amplitude Ak,t and the time variable itself. We obtain the time constants τe ) 1.0 and τkk ) 6.5 ps. The experimental results clearly show that direct bonding of the phenyl ring to the sulfanyl core substituents in Ph-NDI-Ph leads to complex excited-state dynamics and the initially excited molecule goes through two intermediate states before it returns to the ground state, as summarized in eq 2. To exclude that the complex dynamics is due to the two symmetric substituents, we have also measured the transient absorption of A-NDI-Ph (Figure S3 in the Supporting Information) which has only one phenylsulfanyl substituent (see Chart 1). We find the same dynamics with very similar time constants (see Table 1). The aliphatic chain (isopentyl) has no influence on the quenching dynamics as revealed by the transient measurements on A-NDI-A (Figure S3 in the Supporting Information), which shows time constants that are very similar to Bn-NDI-Bn. To gain further insight into the complex mechanism, we invoke the aid of ab initio calculations. DFT and TDDFT calculations have been performed in order to characterize the states involved in the relaxation process and to understand what happens with the molecule from an electronic and geometric point of view. The ground-state geometry optimization was carried out using a starting geometry, where the phenyl groups are twisted slightly out of the plane of the central NDI chromophore. As can be seen in Figure 5a, in the optimized structure of the S0 state the phenyl rings are twisted by 90° with respect to the NDI chromophore, as in a planar or any quasi-planar conformation there would be a strong steric repulsion between the hydrogens of the NDI and the phenyl substituent. The theoretical findings are confirmed by the crystallographic data in Figure 1, which reveal torsional angles larger than 70° between the phenyl rings and the NDI plane. The optimization of the first singlet excited state has been carried out using the S0 equilibrium structure as starting geometry. The geometry obtained (S1 locally excited ) LE in Figure 5a, left side) is very similar to that of the neutral ground state structure indicating that only minor changes are induced by the S0-S1 electronic excitation. The calculated fluorescence from this S1 state has a vertical transition wavelength of 529 nm (2.34 eV) and a strong oscillator strength (f ) 0.24). The values compare well with the initial fluorescence component at 580 nm (2.13 eV) measured in the Kerr-gating experiment (Figure S2 in the Supporting Information) and is a strong indication that this locally excited ππ* state is the initially optically populated state of Ph-NDI-Ph. The ππ*-S1-LE state is not the only minimum of the S1 potential energy surface. There are two further minima, a local one and the S1 global minimum as shown in Figure 5a. The S1 partial-CT is a local minimum that is almost isoenergetic to the initial S1 LE minimum (Figure 5c). It corre-

Figure 5. (a) S0 and S1 equilibrium geometries for a model of Ph-NDI-Ph without terminal substituents (LE, locally excited state; CT, charge transfer state). (b) Molecular orbitals associated with the three conformations. The electronic occupation is indicated by black arrows on the left of each orbital; the character is written in gray. The large gray arrows point out the energetic order of the orbitals evolving during the geometric relaxation process. (c) Scheme of the state energies involved in the CT process. The solid arrows indicate optical transitions, the dotted ones transitions between the different states along with the measured time constants.

sponds to a quasi-planar structure, where one of the phenyl substituents has rotated into the NDI plane by 50°. From an orbital perspective (see Figure 5b), the S0-S1 transition in the quasi-planar conformation corresponds to a transition from the Ph-π orbital that is mainly localized on the phenyl substituent but has also strong contributions on the naphtyl core, to the antibonding NDI-π*, which is totally localized on the NDI chromophore. It should be noted that the planarization leads to a change of the energy ordering of the NDI-π and Ph-π orbitals. The quasi-planar S1 state is thus of partial CT character. The calculated fluorescence (734 nm, 1.69 eV) from this state into the ground state is red-shifted with respect to the S1-LE fluorescence at 580 nm due to the rise of the ground-state energy in the partial-CT geometry and has a strong oscillator strength (f ) 0.13). This is in line with the red shift of the SE observed in the transient data and the Kerr-gating experiment indicating that the S1 partial-CT is the first intermediate state accessed by the molecule after electronic excitation. Furthermore, the redshifted fluorescence component dominating the steady-state emission can be assigned to the partial CT state. The S1-CT state is the global S1 minimum. It lies 1 eV below the S1 LE state and corresponds to a structure where the whole phenylsulfanyl substituent is twisted by 90° out of the NDI chromophore plane. In this structure the Ph-π orbital is localized on the phenyl ring and the calculated fluorescence oscillator strength to the S0 state (f ) 0.54 × 10-4) is 4 orders of magnitude smaller compared to the values for the S1 LE and S1 partial-CT states, as the Ph-π orbital is conformationally decoupled from the NDI-π-system. Recent theoretical studies on TDDFT have shown that the functionals available in present day ab initio suites can severely

Variations in Ultrafast Fluorescence Quenching underestimate the energies of CT states.28,29 The nature of this failure is due to the locality approximation contained in the adiabatic exchange functionals. We have therefore performed single point energy calculations on the three S1 minima using the RI-CC2/SVP level of theory.30,31 The coupled cluster method reproduces both the relative energy ordering of the states and their electronic character. We can therefore safely assume that the results of the TDDFT calculations are correct also at the other geometries. From the good agreement between the theoretical results and the experimental data, the quenching mechanism emerging for Ph-NDI-Ph is a photoinduced charge transfer from the directly bound phenyl donor to the NDI chromophore upon an asymmetric conformational change (Figure 5c): After electronic excitation into the S1 LE state the molecule accesses the first intermediate state, the quasi-planar S1 partial-CT, with 1 ps. The nominal 0.002 eV energy increase inherent to this torsional deformation is overcome thermally as the experiments have been carried out at room temperature and some vibrational excess energy is supplied by the optical excitation. In the ground state such a conformational deformation raises the energy much more (0.73 eV) compared to the S1 state and therefore the quasiplanar structure is not thermally populated. This indicates that the electronic excitation considerably flattens the potential energy surface along the benzene torsional coordinate. In the quasi-planar conformation the benzene localized orbital Ph-π can overlap with the π-system of the NDI chromophore and due to the torsion induced energy increase it donates an electron to the initially depopulated NDI-π orbital. From the partial-CT state the molecule accesses the second, full CT state within 6.5 ps by twisting the whole phenylsulfanyl substituent out of the plane of the NDI chromophore and thereby lowers its energy by at least 1 eV. With a further 6.5 ps time constant the molecule returns to the ground state by back charge transfer. The combination of the strong red shift of the radiative transition to the ground state, the low oscillator strength and the rapid nonradiative decay result in the observed lack of fluorescence. The above mechanism is further supported by spectroelectrochemical measurements on amino core-substituted NDIs.19 Compared to sulfanyl core-substituted NDIs, the S0-S1 absorption of amino core-substituted NDIs is red-shifted from 530 to 602 nm. The NDI monoanion shows a strong absorption band to the red (687 nm) of the first absorption band of the neutral (602 nm). The relative energy of 0.26 eV between these two bands compares well to energy differences of 0.3 eV between the GSB and the 550-665 nm ESA band of Ph-NDI-Ph observed after 1 ps (Figure 3a). This is a strong indication that the ESA feature can be assigned to the spectral absorption of the sulfanyl core-substituted NDI anion generated by photoinduced electron transfer. From the strength of the fluorescence signal obtained in chloroform, we estimate a fluorescence quantum yield (FQY) in the range of 1% for Bn-NDI-Bn and a factor of 100 less for Ph-NDI-Ph. This is in line with the FQYs that result from the comparison of the measured lifetimes with the radiative lifetimes of 11 ns for Bn-NDI-Bn and 10 ns for Ph-NDI-Ph calculated from the stationary absorption and fluorescence spectra via the Strickler-Berg approach.32 As the major component of the Ph-NDI-Ph fluorescence comes from the S1 partial-CT state, a polarity-dependent Stokes shift might be expected. To test this thesis we carried out further steady-state fluorescence measurements in solvents other than chloroform. Due to the extremely low FQYs, measuring the fluorescence of Ph-NDI-Ph is very difficult but possible. The results are

J. Phys. Chem. A, Vol. 114, No. 48, 2010 12559 shown in Figure S4 in the Supporting Information. Although there is a net increase in solvent polarity when going from chloroform (ε ) 4.81) via THF (ε ) 7.52) to dichloroethane (ε ) 10.5), no consistent shift pattern of the fluorescence can be observed. Furthermore the Stokes shift is polarity independent and has a value of about 2900 cm-1. Predicting the Stokes shift via the Lippert-Mataga equations33 using ab initio electric dipole moments for the S1 partial-CT state (9.3 D) shows that the decisive factor suppressing variations due to the solvent polarity is the large Onsager radius of 6 Å. Transient absorption measurements on Ph-NDI-Ph in THF (Figure S5 in the Supporting Information) reveal that the time constant for the last two steps decreases from 6.5 to 4.8 ps compared to the measurements in chloroform (see Table 1). This can be explained with the lower viscosity of THF of 0.46 mPa · s at 20 °C compared to chloroform (0.54 mPa · s at 20 °C) and is therefore a clear indication that the electron transfer is associated with large conformational changes of the molecule. IV. Conclusions The reported time-resolved studies and ab initio calculations on arylsulfanyl core-substituted NDIs show that if an electron donor is directly attached to the core substituent, a complex electronic relaxation path involving charge transfer processes results in ultrafast and efficient fluorescence quenching. After optical excitation an electron is transferred in two steps first partially and then completely from the phenyl ring to the naphthyl core. The ab inito calculations allow an interpretation of the fluorescence quenching at the level of the molecular orbitals. The reoccupation of the optically depopulated NDI-π orbital (see Figure 5) by the CT blocks the radiative transfer from the NDI-π* orbital even though the system is still electronically excited. By a subsequent back transfer the molecule returns finally to the electronic ground state nonradiatively. All transfer steps are associated with large conformational changes and reorientations of the phenyl ring. If the electron donor is linked to the sulfur via a spacer group like in the case of Bn-NDI-Bn, the charge transfer is significantly slower than for Ph-NDI-Ph and also slower than the intrinsic internal conversion. Accordingly, significant emission from the optically excited S1 LE state is observed and the contribution from possible intermediates is negligible. Many of the previously reported time-resolved studies on NDIs concentrate on charge transfer processes from substituents attached to the N-terminus of the dicarboxy imide group. These processes can range from several 100 ps9,34 down to 0.5 ps35 depending on the electrochemical potential of the substituent. To the best of our knowledge, this work is the first time-resolved study reporting ultrafast rates for charge transfer processes occurring from the core substituents to the NDI unit, which involve conformational changes. Chaignon et al.15 reported a two-step process with rates of 52 ps for the forward electron transfer and 14 ps for the back electron transfer in a Ru(bpy)3-NDI dyad. However the bridge and core substituent between the Ru(bpy)3 and the naphthyl core is a rigid phenylacetlylene spacer and does therefore not involve any reorientation of the subsituents in order for the charge transfer to occur. In combination with our results this indicates that the lifetime of the S1 state can also be tuned by the rigidity of the linker and the electron-donating group. Acknowledgment. This work was supported by the Austrian Science Fund within the framework of the Special Research Program F16 (Advanced Light Sources), by the SFB749, and

12560

J. Phys. Chem. A, Vol. 114, No. 48, 2010

by the DFG-Cluster of Excellence: Munich-Centre for Advanced Photonics. The Alexander von Humboldt Stiftung is gratefully acknowledged for a fellowship (I.P.), and the Leibniz-Rechenzentrum LRZ Munich is gratefully acknowledged for computing time and access to TURBOMOLE. We thank Bjo¨rn Heinz and Peter Gilch for kindly recording the transient fluorescence spectra and for the discussions. Dorit Shemesh from the group of Professor Dr. Wolfgang Domcke is greatly acknowledged for carrying out the computer intensive RI-CC2 calculations on the Ph-NDI-Ph molecule. Supporting Information Available: A full chemical and spectroscopic characterization of the investigated naphthalenediimides. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Sakai, N.; Mareda, J.; Vauthey, E.; Matile, S. Chem. Commun. 2010, 46, 4225. (2) Bhosale, S. V.; Jani, C. H.; Langford, S. J. Chem. Soc. ReV. 2008, 37, 331. (3) Licchelli, M.; Biroli, A. O.; Poggi, A. Org. Lett. 2006, 8, 915. (4) Levanon, H.; Galili, T.; Regev, A.; Wiederrecht, G. P.; Svec, W. A.; Wasielewski, M. R. J. Am. Chem. Soc. 1998, 120, 6366. (5) Wiederrecht, G. P.; Niemczyk, M. P.; Svec, W. A.; Wasielewski, M. R. J. Am. Chem. Soc. 1996, 118, 81. (6) Angadi, M. A.; Gosztola, D.; Wasielewski, M. R. J. Appl. Phys. 1998, 83, 6187. (7) Katz, H. E.; Lovinger, A. J.; Johnson, J.; Kloc, C.; Slegrist, T.; Li, W.; Lin, Y.-Y.; Dodabalapur, A. Nature 2000, 404, 478. (8) Miller, L. L.; Mann, K. R. Acc. Chem. Res. 1996, 29, 417. (9) Debreczeny, M. P.; Svec, W. A.; Wasielewski, M. R. Science 1996, 274, 584. (10) Wasielewski, M. R. J. Org. Chem. 2006, 71, 5051. (11) Shiratori, H.; Ohno, T.; Nozaki, K.; Yamazaki, I.; Nishimura, Y.; Osuka, A. J. Org. Chem. 2000, 65, 8747. (12) Osuka, A.; Yoneshima, R.; Shiratori, H.; Okada, T.; Taniguchi, S.; Mataga, N. Chem. Commun. 1998, 1567. (13) Błaszczyk, A.; Fischer, M.; von Hänisch, C.; Mayor, M. HelV. Chim. Acta 2006, 89, 1986. (14) Thalacker, C.; Röger, C.; Würthner, F. J. Org. Chem. 2006, 71, 8098.

Pugliesi et al. (15) Chaignon, F.; Falkenström, M.; Karlsson, S.; Blart, E.; Odobel, F.; Hammarström, L. Chem. Commun. 2007, 64. (16) Sakai, N.; Sisson, A. L.; Bu¨rgi, T.; Matile, S. J. Am. Chem. Soc. 2007, 129, 15758. (17) Sisson, A. L.; Sakai, N.; Banerji, N.; Fu¨rstenberg, A.; Vauthey, E.; Matile, S. Angew. Chem., Int. Ed. 2008, 47, 3727. (18) Banerji, N.; Fürstenberg, A.; Bhosale, S.; Sisson, A. L.; Sakai, N.; Matile, S.; Vauthey, E. J. Phys. Chem. B 2008, 112, 8912. (19) Chopin, S.; Chaigon, F.; Blart, E.; Odobel, F. J. Mater. Chem. 2007, 17, 4139. (20) Megerle, U.; Pugliesi, I.; Schriever, C.; Sailer, C. F.; Riedle, E. Appl. Phys. B: Laser Opt. 2009, 96, 215. (21) Ph-NDI-Ph (C54H53N2O4S2 · 4CHCl3): a ) 1344.2(3), b ) 1499.6(3), c ) 1713.9(3) pm, R ) 66.96(3), β ) 81.68(3), γ ) 87.85(3)°, V ) 3145.1(11) × 106 pm3; triclinic P-1, Z ) 2, Fcalcd ) 1.410 g cm-1, µ(Mo KR) ) 0.640 mm-1, STOE IPDS2, Mo KR radiation, λ ) 0.71073 Å, T ) 200 K, 2θmax ) 45°; 15582 reflections measured, 7928 independent reflections (Rint ) 0.0453), 6317 independent reflections with Fo > 4σ(Fo). The structure was solved by direct methods and refined, by full-matrix least square techniques against F2, 697 parameters (Cl, S, O, N, C refined anisotropically, H atoms were calculated at ideal positions). R1 ) 0.0764; wR2 ) 0.2262 (all data); GOF: 1.073; maximum peak 1.007 e Å-3. CCDC-667499 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from Cambridge Crystallographic Data Centre, 12 Union Rd, Cambridge CB21EZ, UK; fax, (+44)1223-336-033; or [email protected]). (22) Scha¨fer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571. (23) Scha¨fer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829. (24) Furche, F.; Ahlrichs, R. J. Chem. Phys. 2002, 117, 7433. (25) Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995, 102, 346. (26) Ahlrichs, R.; Ba¨r, M.; Ha¨ser, M.; Horn, H.; Ko¨lmel, C. Chem. Phys. Lett. 1989, 162, 165. (27) Fita, P.; Luzina, E.; Dziembowska, T.; Radzewicz, Cz.; Grabowska, A. J. Chem. Phys. 2006, 125, 184508. (28) Dreuw, A.; Weisman, J. L.; Head-Gordon, M. J. Chem. Phys. 2003, 119, 2943. (29) Dreuw, A.; Head-Gordon, M. J. Am. Chem. Soc. 2004, 126, 4007. (30) Christiansen, O.; Koch, H.; Jørgensen, P. Chem. Phys. Lett. 1995, 243, 409. (31) Ha¨ttig, C.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 2000, 113, 5154. (32) Strickler, S. J.; Berg, R. A. J. Chem. Phys. 1962, 37, 814. (33) Kawski, A. Z. Naturforsch. 2002, 57a, 255. (34) Chaignon, F.; Buchet, F.; Blart, E.; Falkenström, M.; Hammarström, L.; Odobel, F. New. J. Chem. 2009, 33, 408. (35) Ganesan, P.; Baggerman, J.; Zhang, H.; Sudhölter, E. J. R.; Zuilhof, H. J. Phys. Chem. A 2007, 111, 6151.

JP107742X