Defining Cyclic–Acyclic Exciton Transition at the Single-Molecule

Letter pubs.acs.org/JPCL

Defining Cyclic−Acyclic Exciton Transition at the Single-Molecule Level: Size-Dependent Conformational Heterogeneity and Exciton Delocalization in Ethynylene-Bridged Cyclic Oligothiophenes Kyu Hyung Park,† Jae-Won Cho,† Tae-Woo Kim,† Hideyuki Shimizu,‡ Kazumi Nakao,‡ Masahiko Iyoda,*,‡ and Dongho Kim*,† †

Spectroscopy Laboratory for Functional π-Electronic Systems and Department of Chemistry, Yonsei University, Seoul 03722, Korea Department of Chemistry, Graduate School of Science and Engineering, Tokyo Metropolitan University, Hachioji, Tokyo 192-0397, Japan

‡

S Supporting Information *

ABSTRACT: Conformational disorder in π-conjugated cyclic systems plays a crucial role in controlling the extent of exciton delocalization in much the same way as that in linear counterparts. However, to date, there have been no detailed spectroscopic investigations on the nature of excitons in π-conjugated cyclic systems at the single-molecule level. Herein, we studied the effect of conformational disorder on the excitonic behaviors of cyclic oligothiophenes composed of 6, 8, 10, and 12 subunits (C-6T, C-8T, C-10T, and C-12T, respectively) by employing single-molecule fluorescence spectroscopy. We found that, due to the cyclic symmetry constraint which suppresses S1− S0 transition, small and rigid C-6T and C-8T exhibit extremely long fluorescence lifetimes, while short lifetimes typical of linear systems are dominant in large, flexible C-10T and C-12T. Two-dimensional correlation maps between fluorescence lifetimes and spectral positions show that, by torsional defects created through continued photoexcitation, fully delocalized cyclic excitons shrink to form acyclic excitons in the case of C-10T, while localized acyclic excitons from initial states are maintained in the case of C-12T. The distribution of linear dichroism values from C-6T to C-10T gradually broadens but narrows in C-12T, suggesting a cyclic-to-acyclic transition in excitonic nature between C-10T and C-12T.

O

effectively as compared to its linear counterpart, L-10T.16 We also found that in larger macrocyclic oligothiophenes, from C15T3V to C-30T6V, conformational heterogeneity increases, causing excitons to localize. Such size-dependent exciton localization has consistently been reported in CPPs and porphyrin nanorings as well.18,20,21 Our group recently reported a detailed study on molecular size dependence of exciton dynamics near the size limit of complete exciton delocalization using C-nT, macrocyclic oligothiophenes composed of 6 to 12 subunits, shown in Chart 1.22 Subtle changes in the number of subunits resulted in drastic differences in the transient fluorescence spectra. With the aid of molecular dynamics (MD) simulations, we found that complete exciton delocalization is accessible only in the systems composed of less than eight thiophene subunits, whereas those composed of more than eight thiophene subunits fail to attain excitons extended over the whole carbon backbone. While this result provides vital information on the correlation between exciton size, molecular conformation, and spectral parameters in the π-conjugated macrocyclic oligomers, a few key questions still remain: How does conformation hetero-

ver the past decades, conjugated polymers (CPs) have evolved into one of the core ingredients in organic optoelectronic devices such as light-emitting diodes, field-effect transistors, and photovoltaic cells.1−6 Performances of such devices are often greatly influenced by the degree of inter/ intrachain exciton delocalization,5,7,8 which is attained from planar geometry free of torsional defects.9−13 This suggests πconjugated macrocyclic oligomers, which have received much attention in recent years for stable and rigid π-conjugation pathways and peculiar photophysical properties originating from their highly strained structures, as candidates for effective π-conjugation extension. In π-conjugated macrocyclic oligomers, a simple end-to-end connectivity was shown to remove perturbing end-effects ubiquitous in linear counterparts and significantly lower torsional disorder of carbon backbones.14−17 In a recent study on porphyrin nanorings, it was shown that excitons initially formed upon photoexcitation are completely delocalized, as inferred from the ultrafast fluorescence anisotropy decay within 200 fs.18 Transition density matrix analysis illustrates that cycloparapheneylenes (CPPs) exhibit completely delocalized excitons while those of linear counterparts localize to the center of the molecule.19 A work by our group also demonstrated that macrocyclic oligothiophene composed of 10 subunits with two vinylene linkages, C-10T2V, acquires exciton delocalization more © XXXX American Chemical Society

Received: February 17, 2016 Accepted: March 17, 2016

1260

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

Letter

The Journal of Physical Chemistry Letters Chart 1. Molecular Structures of Cyclic Oligothiophenes

geneity in cyclic systems affect inhomogeneity in spectral signatures? Are the systems in the size-limit of exciton delocalization composed of single de/localized excitons or two subpopulations? Do rigid macrocyclic oligothiophenes exhibit less conformational fluctuations as previously reported for π-conjugated linear polymers? To answer these questions, we employed single-molecule spectroscopic tools which allow us to avoid ensemble-averaging effect and attribute observed spectral signatures of excitons to the conformational changes of individual macrocyclic molecules. In the smallest cyclic oligothiophene, C-6T, we observed a distribution of strikingly long fluorescence lifetimes associated with a forbidden transition arising from ideal cyclic structure. As the size of system increases, average lifetimes decrease inferring the lift of symmetry constraint. By analyzing two-dimenasional (2D) correlation maps between spectral peak positions and fluorescence lifetimes, we found that the ensemble of C-10T is composed of two subpopulations, delocalized and localized excitons, while C-12T exclusively generates localized excitons. Further investigation on linear dichroism (LD), which is defined as a difference to sum ratio of the two orthogonal polarization components,23 reveals that the transition from the delocalized cyclic excitons to randomly localized acyclic excitons takes place between C-10T and C-12T. Figure 1 shows representative single-molecule fluorescence data of cyclic oligothiophenes embedded in a poly(methylmethacrylate) (PMMA) matrix prepared in an anaerobic atmosphere. All cyclic oligothiophenes exhibit discrete steps which are classified by fluorescence intensity levels, lifetimes, spectra, and LDs. Each step involved in the fluorescence intensity trajectory (FIT) indicates a separate excitonic state, and spectral jumps between the steps can be associated with photoinduced structural changes of CPs, as previously reported in single chains of poly[2-methoxy-5-(2′-ethylhexyloxy)-1,4phenylenevinylene] (MEH-PPV), oligo(phenylenevinylene) (OPV), and oligothiophene.11−13 In the linear systems, spectral fluctuations are ascribed to the changes in the electronic delocalization by photoinduced torsional flips which create conjugational defects. In the case of cyclic oligothiophenes, this process requires thiophene subunits to overcome the macrocyclic ring strain to deviate from the plane cyclic backbones form. As size of the system increases, macrocyclic ring strain is alleviated24−26 and the spectral jumps become more frequent, as inferred from the increase in the average number of steps involved in a single FIT (C-6T, 1.86 steps; C-8T, 2.53 steps; C10T, 2.92 steps; C-12T, 3.75 steps). Size-dependent structural flexibility is also evident in the trend in fluorescence lifetimes and spectral positions. In Figure 1a, C-6T exhibits low yet nearly constant fluorescence intensity level and exceptionally long fluorescence lifetime of 6.4 ns. The

Figure 1. Representative fluorescence intensity trajectories (FITs) and corresponding linear dichroism (LD), fluorescence decay profiles, and spectra of (a) C-6T, (b) C-8T, (c) C-10T, and (d) C-12T. The trajectory is classified according to different intensity level, linear dichroism, fluorescence lifetime, and spectrum. The y-axis of the FIT indicates the summation of photon counts of two polarization channels per 20 ms. Fluorescence decay profile is obtained from each classified state. The exponential fitting line and the fitted fluorescence lifetime of each decay profile are shown. The fluorescence decay profiles are shown in semilogarithmic scale. For C-8T, C-10T, and C-12T, multiple single-molecule spectra, measured for 1 s for each, were averaged over one molecular state to produce each spectrum shown. However, for C-6T, because of low intensity, a spectrum was not obtained. For C-10T and C-12T, the Gaussian fitting line which is composed of three Gaussian curves is drawn as solid lines over the spectra. The 0−0 vibronic peak positions and peak intensity ratios (I0−1/I0−0) are obtained from decomposed Gaussian curves. The number of molecules obtained for statistical analysis was 193, 183, 173, and 165 for C-6T, C-8T, C-10T, and C-12T, respectively.

low fluorescence intensity, as well as extremely long fluorescence lifetime, seems to arise from an ideal cyclic symmetry. In fully conjugated cyclic structures, exciton states rearrange to produce a forbidden S1 state and pairwise1261

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

Letter

The Journal of Physical Chemistry Letters degenerate higher states, where only the S2 and S3 pair carries appreciable oscillator strength.18,22,27 Fluorescence from the S1 state in ideally cyclic structure is thus weak and slow, as is the case of C-6T. The cyclic symmetry constraint is gradually lifted as structural disorder increases with increasing the size of system. As structural disorder, such as inter-ring torsion and bending in cyclic structure, is added, pairwise degeneracy is lifted and the oscillator strengths focused on the S2 and S3 pair are redistributed to other states, including the S1 state. Short lifetimes in the first steps (C-8T, 780 ps; C-10T, 640 ps; C12T, 590 ps) of representative FITs indicate that conformational disorder, and consequent symmetry lowering, of C-8T, C-10T, and C-12T is initially much greater than that of C-6T. The blue-shift of the 0−0 vibronic peak positions in the initial steps from C-10T to C-12T coincides with this trend, although vibronic features in the fluorescence spectra of C-6T and C-8T are indiscernible because of their low fluorescence intensity. Reduction in the 0−1 to 0−0 vibronic peak ratio (I0−1/I0−0) from 1.29 in C-10T to 0.85 in C-12T, which reflects the degree of exciton coherence in the cyclic π-conjugated systems, also supports this argument.28 In going from the initial to the following steps, fluorescence lifetimes of C-8T and C-10T drastically decrease and the corresponding LDs increase in magnitude. In cyclic systems, photoinduced torsional defects can perturb the cyclic symmetry of a molecule and allow the oscillator strength of S1 to grow.22,29,30 Reduction of fluorescence lifetime from 780 to 600 ps, as well as increase of LD from 0 to 0.25 in the representative case of C-8T, in this regard reflects a change in the excitonic manifold as a result of perturbation in cyclic symmetry. Changes in the fluorescence lifetime from 640 to 550 ps, 0−0 vibronic peak position from 522 to 517 nm, and LD from −0.15 to −0.42 in C-10T also imply the similar changes in the excitonic state. However, from the significantly shorter lifetime and nonzero LD in the initial state of C-10T, we expect that excitons in this representative case originate from disordered cyclic structure with or without torsional defects. Furthermore, spectral jumps in C-12T, unlike C-8T and C-10T, manifest no clear changes in the spectral signatures, which are commonly related to the extent of exciton delocalization. An absence of correlations between spectral parameters in C-12T is reminiscent of excitonic signatures in linear CPs.11 It was shown that correlation between fluorescence intensity, spectral width, and position is weak or absent depending on the chains one observes. Therefore, clarifying an ambiguous nature of C10T in its initial state, as well as linear-like behaviors in C-12T, requires a statistical analysis of fluorescence parameters. To illustrate the effect of size-dependent static disorder on the extent of exciton delocalization, we performed statistical analyses on the fluorescence lifetimes as shown in Figure 2. In accordance with the trend from ensemble measurements in solution phase,22 average fluorescence lifetimes of single cyclic oligothiophenes decrease as the size of the molecule increases (C-6T, 5.6 ns; C-8T, 1.1 ns; C-10T, 590 ps; C-12T, 540 ps). The most prominent features are a broad lifetime distribution of C-6T, which is centered at around 5 ns, and an appearance of subpopulation of C-8T possessing lifetimes ranging from 2 to 9 ns. These extraordinarily long lifetimes are uncommon in π-conjugated linear oligo-/polymers, whose lifetimes are typically in the range of hundreds of picoseconds.10,13,31,32 C6T and C-8T are, due to their large ring strains, conformationally rigid and retain planar cyclic geometry. Accordingly, an extremely long fluorescence lifetime originating from the

Figure 2. Fluorescence lifetime histograms of (a) C-6T, (b) C-8T, (c) C-10T, and (d) C-12T. The histograms consist of the lifetime values obtained from each molecular state of all molecules measured. Inset in panel b is an extended view of a region between 2 and 9 ns for C-8T. For C-10T and C-12T, extended views of a region between 300 and 1000 ps of each original histogram are shown as insets. For C-10T and C-12T, histograms are fitted with Gaussian functions and fitted curves are drawn in solid lines in the insets of panels c and d, respectively. Average lifetimes are obtained by averaging all lifetimes in the case of C-6T and C-8T, while those of C-10T and C-12T are obtained from the Gaussian curve fitting.

forbidden S1−S0 transition in an ideal cyclic geometry is more often observed. It is worth noting that the breadth of the lifetime distribution is immense. This situation is reminiscent of sensitive modulation of excimer lifetimes by small changes in the interchromophoric spacing between two linear oligomers in close proximity, as recently investigated by Stangl et al.31 It seems that the oscillator strength of an emitting state of C-6T in an ideal cyclic conformation can greatly vary by a small conformational inhomogeneity which lifts symmetric constraints. Despite the structural rigidity of C-6T, there is a minor subpopulation (9.5%) with fluorescence lifetimes shorter than 2 ns. This implies that the cyclic geometry of C-6T can also be perturbed by a static disorder. On the other hand, in the case of C-8T, the population possessing short lifetimes is dominant (84% under 2 ns) and those with extremely long fluorescence lifetimes are sparsely 1262

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

Letter

The Journal of Physical Chemistry Letters

Figure 3. Two-dimensional contour plots between fluorescence lifetimes and peak positions of 0−1 vibronic transition of initial (panel a for C-10T and panel d for C-12T) and all molecular states (panel b for C-10T and panel e for C-12T). Each contour plot was obtained from a Gaussian mixture model. Color scale bar in the bottom left of each panel indicates population density. The corresponding fluorescence lifetime histogram and spectral distribution are depicted at the top and right side of each panel, respectively. The solid lines correspond to the fitted Gaussian curves for each histogram, and the shaded area indicates the subpopulation created by photoinduced conformational fluctuations. Panels c and f are the schematic illustrations of the change in excitonic states from the first to the next steps of C-10T and C-12T, respectively.

2D correlation maps between the positions of 0−1 vibronic peaks and fluorescence lifetimes as shown in Figure 3. In the initial steps of C-10T, a negative correlation between the peak positions and lifetimes is pronounced. In this step, both peak positions and lifetimes show monomodal distributions centered at 2.23 eV and 620 ps, respectively. In the transition from the initial to next steps, the distribution of peak positions becomes bimodal with an increase in the population of the higher-energy state at 2.31 eV and the distribution of fluorescence lifetimes shifts to 590 ps. In our previous work, we demonstrated that the bimodality in the 2D correlation map of C-10T2V is a manifestation of two different types of excitons: cyclic and acyclic excitons. In the same way, a domain which appears in the initial step and persists in the next steps represents a population of fully delocalized cyclic excitons. An apparent negative correlation between the fluorescence peak position and lifetime indicates that the energy and oscillator strength of cyclic excitons are modulated by conformational disorder in conjugation pathways. Continued illumination creates a subpopulation located at higher fluorescence energy and shorter lifetime. In this region, change in the fluorescence lifetimes in response to the change in peak energies is very small. Considering that conformational fluctuations induced by continued photoexcitation sporadically creates torsional defects, the domain populated after the first step can be regarded as a signature of acyclic excitons. It was demonstrated that in a cyclic system, torsional disorder over π-conjugation cutoff angle induces exciton localization and blue-shift of fluorescence, while leaving radiative rate essentially unchanged.33 Such a disorderinsensitive fluorescence lifetime is also reflected in the new subpopulation of C-10T, suggesting the formation of acyclic

distributed from 2 to 9 ns, implying that the static disorder of C-8T on average is greater than that in C-6T. C-10T and C12T do not show any population in the region above 2 ns. They both display simple bell-shaped distributions of lifetimes centered at 590 and 540 ps for C-10T and C-12T, respectively. The Gaussian curve fit shows that the full width at halfmaximum (fwhm) of the lifetime distribution of C-12T is 190 ps, whereas that of C-10T is 230 ps. In the case of C-12T, essentially all fluorescence lifetimes fall into the range which those of typical π-conjugated linear oligomers belong to. In conjugation with previous ensemble study, we can expect that severe conformational disorder in C-12T obstructs πconjugation extension over the whole molecule. Excitons created in such carbon frameworks localize into the conjugational segments and behave much the same way as linear excitons. On the other hand, the lifetime distribution of C-10T contains a portion of lifetimes which are too long to be accounted for as linear excitons (20% over 700 ps). In C-10T2V, a system similar to C-10T in size and shape, we have identified the existence of fully delocalized excitons with small conformational disorders, which differ from torsional defects associated with acyclic excitons.33 In the same way, we can expect C-10T to form fully delocalized excitons, although their shapes are not so symmetric as those formed in C-6T or C-8T. However, whether the lifetime distribution of C-10T consists solely of delocalized excitons or two subpopulations differing in the extent of excitonic delocalization has to be analyzed by evaluating the correlation of two or more parameters. To investigate the development of excitons by photoinduced conformational change and to identify the excitonic characters of the subpopulations created by such changes, we studied the 1263

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

Letter

The Journal of Physical Chemistry Letters excitons from cyclic excitons in the initial molecular conformations. It is interesting to note that, as compared to the previously studied system, C-10T2V, the distribution of peak positions is located at higher energy irrespective of steps (C-10T, 2.23 and 2.31 eV; C-10T2V, 2.14 and 2.21 eV).33 This difference can be accounted for by the backbone stiffening induced by vinylene linkers, which causes a red-shift in the fluorescence spectra.17 Nonetheless, a similarity in the overall behaviors in the two 2D correlation maps assures a correspondence in photoinduced dynamics. In the initial step of C-12T, both peak positions and fluorescence lifetimes show monomodal distributions centered at 2.27 eV and 540 ps, respectively. In the transition from the initial to next steps, the distribution of peak positions becomes bimodal with an increase in the population of the higher-energy state at 2.42 eV. Surprisingly, unlike C-10T, the fluorescence lifetime is preserved during the transition. It is very unlikely that the same photoinduced conformational change is absent in the larger cyclic system, because the energy barrier along the torsional angle change in C-12T is much smaller than that in the smaller cyclic systems which have larger ring strains.24−26 The stagnant lifetime distribution despite the torsional fluctuation is thus attributed to the disorder-insensitive nature of localized excitons. This is further supported by the lifetime saturation observed in ensemble measurements with πconjugated linear and cyclic oligomers.16,18,34 If a correlation between exciton delocalization and fluorescence lifetime is strong for localized excitons, the larger conformational heterogeneity in C-12T should have given a lifetime distribution broader than that of C-10T. On the contrary, the fwhm values from the Gaussian curve fits of lifetime distributions show that the distribution of C-12T (160 and 190 ps for initial and total steps, respectively) is indeed much narrower than that of C-10T (240 and 230 ps for initial and total steps, respectively), which can further support the formation of localized excitons in C-12T. It is obvious that photoinduced conformational fluctuations in two samples all result in a disruption of exciton delocalization. However, in C10T, this induces a transition from cyclic to acyclic exciton, whereas in C-12T, the exciton merely shrinks in size but retains its acyclic character. These processes are illustrated in Figure 3c and Figure 3f for C-10T and C-12T, respectively, as schematic drawings. To investigate the nature of delocalized and localized excitons and the effect of conformational disorder on the emitting behaviors of cyclic oligothiophenes, we have analyzed their LD distributions as shown in Figure 4. All four histograms show bell-shaped distribution centered at 0. The Gaussian curve fit reveals that the fwhm of the distribution increases when going from C-6T to C-10T but decreases in C-12T (0.70 for C-6T, 0.75 for C-8T, 0.96 for C-10T, and 0.78 for C-12T), which can be correlated with the transition of excitonic nature in the largest cyclic oligothiophene. In C-6T, the dominant mechanism of electronic transition is Herzberg−Teller (HT) coupling, which breaks the symmetry of electronic wave function by nontotally symmetric vibrational motions.35 HT coupling was found to provide excellent explanations on the absorption and emission properties of not only the long-studied small cyclic systems, such as benzene, fullerene, and porphyrin,36−38 but also recently synthesized porphyrin nanorings.29,35 As previously described by Sprafke et al., HT coupling to nontotally symmetric vibrational modes momentarily deforms cyclic structure to produce degenerate x-

Figure 4. Linear dichroism histograms of (a) C-6T, (b) C-8T, (c) C10T, and (d) C-12T. The histograms consist of linear dichroism values obtained from each excitonic state of all molecules measured. Gaussian fitting curves are drawn as solid lines.

and y-polarized transitions oscillating in the frequency of the mode coupled to the transition.35,39,40 In a binning time of typical single-molecule spectroscopy, this ultrafast polarization change in the delocalized exciton appears as a sum of two degenerate transition dipoles, which gives a value of 0 in LD. The distribution centered at 0 in Figure 4a indicates that C-6T maintains an ideal cyclic structure which retains strict cyclic symmetry. It is worth mentioning that ΔLD of C-6T is also very sharp, in contrast to broad distributions of other cyclic oligothiophenes, indicating its conformational ideality against photoinduced conformational fluctuations (see Figure S6 in the Supporting Information). As the size of the system increases, so does the innate static disorder of the system. Based on the lifetime distribution in Figure 2b and MD simulations in our previous work, C-8T on average maintains planarity devoid of torsional defects, but its symmetry is lowered because of increased flexibility of the structure.22 Deviation from circularity greatly alters the excitonic manifold as demonstrated in the modeling of exciton states of light-harvesting 2 (LH2) complexes possessing elliptical deformation.30,41 This type of disorder was shown to lift the degeneracy of doubly degenerate transitions and transfers the oscillator strength of k = ± 1 states to k = ± 3 states. In a similar context, Yong et al. suggested that a 1264

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

The Journal of Physical Chemistry Letters

■

deviation from circularity has a strong correlation to the increase of radiative rate in porphyrin nanorings.29 We expect that, due to a combination of the defect-free π-conjugation pathway and static disorder which lifts the cyclic symmetry, a degeneracy between two mutually orthogonal transition dipole moments is perturbed to create one dominant and the other subordinate transition dipole moments. As a sum they appear as a polarized emission,42 which contributes to the broadening of the LD distribution of C-8T in Figure 4b. In C-10T, where conformational disorder is much more severe than that of C8T, this effect further broadens the distribution of LD. Thus, an increase in the fwhm of the LD distribution going from C-6T to C-10T is a contribution from a deviation of circularity which lifts the degeneracy of transition dipoles of the emitting states. Surprisingly, the LD distribution of C-12T becomes narrower than that of C-10T, displaying the fwhm value of 0.78. Unlike other cyclic oligothiophenes, C-12T exclusively generates acyclic excitons due to the largest static disorder involving torsional defects, as evidenced by the distribution of short lifetimes (Figure 3d,e). Assuming that the cyclic structure does not collapse, excitons generated in C-12T can localize in the indefinite segments on the rim of cyclic conjugation pathway. From a narrow distribution of LD values centered at 0, which is similar to the case reported by Aggarwal et al.43 and Thiessen et al.,44 we suggest that nondeterministic exciton localization takes place in the largest cyclic oligothiophene C12T. In summary, we have investigated the exciton delocalization−localization transition in a series of cyclic oligothiophenes at the single-molecule level. By analyzing their lifetime distributions, we found that in C-6T and C-8T, extremely long fluorescence lifetimes arising from ideal cyclic conformation appear, while those are absent in C-10T and C-12T. The 2D correlation plots between fluorescence lifetimes and spectral peak positions revealed that the transition from cyclic to acyclic excitons takes place between C-10T and C-12T. Investigation on LD distribution has shown that acyclic excitons of C-12T nondeterministically relaxes to a part of the cyclic rim, while cyclic excitons in C-6T relaxes via HT coupling to nontotally symmetric vibrational modes. We suggest that the lift of degeneracy by increasing static disorder in C-8T and C-10T is an origin of the broadening of LD distributions. The singlemolecule study we present on the morphology of excitons in these model cyclic systems, as far as we know, is the first singlemolecule level observation of molecular size-limit in the formation of cyclic excitons retaining complete delocalization. The effect of conformational heterogeneity on the anisotropic emission of delocalized excitons, as well as characterization of cyclic and acyclic excitons by multiple spectral parameters will provide meaningful insight into the dynamic processes in a novel class of materials under active research and is believed to help developing systems with photophysical properties apt for device applications.

■

Letter

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The work at Yonsei University was supported by Global Research Laboratory through the National Research Foundation of Korea (NRF) (NRF-2013K1A1A2A02050183) funded by the Ministry of Science, ICT (Information and Communication Technologies), and Future Planning (D.K.). The work at Tokyo Metropolitan University was supported by a Grant-inAid for Scientific Research from JSPS and by CREST of JST (Japan Science and Technology Corporation) (M.I.).

■

REFERENCES

(1) Ho, P. K. H.; Thomas, D. S.; Friend, R. H.; Tessler, N. AllPolymer Optoelectronic Devices. Science 1999, 285, 233−236. (2) Günes, S.; Neugebauer, H.; Sariciftci, N. S. Conjugated PolymerBased Organic Solar Cells. Chem. Rev. 2007, 107, 1324−1338. (3) Heeger, A. J. Semiconducting Polymers: the Third Generation. Chem. Soc. Rev. 2010, 39, 2354−2371. (4) Facchetti, A. Organic Semiconductors: Made to Order. Nat. Mater. 2013, 12, 598−600. (5) Falke, S. M.; Rozzi, C. A.; Brida, D.; Maiuri, M.; Amato, M.; Sommer, E.; De Sio, A.; Rubio, A.; Cerullo, G.; Molinari, E.; Lienau, C. Coherent Ultrafast Charge Transfer in an Organic Photovoltaic Blend. Science 2014, 344, 1001−1005. (6) Ye, L.; Zhang, S.; Huo, L.; Zhang, M.; Hou, J. Molecular Design toward Highly Efficient Photovoltaic Polymers Based on TwoDimensional Conjugated Benzodithiophene. Acc. Chem. Res. 2014, 47, 1595−1603. (7) Köhler, A.; dos Santos, D. A.; Beljonne, D.; Shuai, Z.; Brédas, J.L.; Holmes, A. B.; Kraus, A.; Müllen, K.; Friend, R. H. Charge Separation in Localized and Delocalized Electronic States in Polymeric Semiconductors. Nature 1998, 392, 903−906. (8) Dang, M. T.; Hirsch, L.; Wantz, G.; Wuest, J. D. Controlling the Morphology and Performance of Bulk Heterojunctions in Solar Cells. Lessons Learned from the Benchmark Poly(3-hexylthiophene):[6,6]Phenyl-C61-butyric Acid Methyl Ester System. Chem. Rev. 2013, 113, 3734−3765. (9) Chen, P.-Y.; Rassamesard, A.; Chen, H.-L.; Chen, S.-A. Conformation and Fluorescence Property of Poly(3-hexylthiophene) Isolated Chains Studied by Single Molecule Spectroscopy: Effects of Solvent Quality and Regioregularity. Macromolecules 2013, 46, 5657− 5663. (10) Adachi, T.; Vogelsang, J.; Lupton, J. M. Chromophore Bending Controls Fluorescence Lifetime in Single Conjugated Polymer Chains. J. Phys. Chem. Lett. 2014, 5, 2165−2170. (11) Pullerits, T.; Mirzov, O.; Scheblykin, I. G. Conformational Fluctuations and Large Fluorescence Spectral Diffusion in Conjugated Polymer Single Chains at Low Temperatures. J. Phys. Chem. B 2005, 109, 19099−19107. (12) Kobayashi, H.; Tsuchiya, K.; Ogino, K.; Vacha, M. Spectral Multitude and Spectral Dynamics Reflect Changing Conjugation Length in Single Molecules of Oligophenylenevinylenes. Phys. Chem. Chem. Phys. 2012, 14, 10114−10118. (13) Kim, T.-W.; Kim, W.; Park, K. H.; Kim, P.; Cho, J.-W.; Shimizu, H.; Iyoda, M.; Kim, D. Chain-Length-Dependent Exciton Dynamics in Linear Oligothiophenes Probed Using Ensemble and Single-Molecule Spectroscopy. J. Phys. Chem. Lett. 2016, 7, 452−458. (14) Williams-Harry, M.; Bhaskar, A.; Ramakrishna, G.; Goodson, T., III; Imamura, M.; Mawatari, A.; Nakao, K.; Enozawa, H.; Nishinaga, T.; Iyoda, M. Giant Thienylene-Acetylene-Ethylene Macrocycles with Large Two-Photon Absorption Cross Section and SemishapePersistence. J. Am. Chem. Soc. 2008, 130, 3252−3253.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.6b00360. Experimental details, steady-state absorption and fluorescence spectra, additional FITs, 2D correlation plots, distribution of vibronic peak ratios, and distribution of ΔLDs (PDF) 1265

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266

Letter

The Journal of Physical Chemistry Letters

Disorder of a Conjugated Macrocycle. J. Phys. Chem. B 2015, 119, 4116−4126. (34) Schumacher, S.; Ruseckas, A.; Montgomery, N. A.; Skabara, P. J.; Kanibolotsky, A. L.; Paterson, M. J.; Galbraith, I.; Turnbull, G. A.; Samuel, I. D. W. Effect of Exciton Self-trapping and Molecular Conformation on Photophysical Properties of Oligofluorenes. J. Chem. Phys. 2009, 131, 154906. (35) Sprafke, J. K.; Kondratuk, D. V.; Wykes, M.; Thompson, A. L.; Hoffmann, M.; Drevinskas, R.; Chen, W.-H.; Yong, C. K.; Kärnbratt, J.; Bullock, J. E.; Malfois, M.; Wasielewski, M. R.; Albinsson, B.; Herz, L. M.; Zigmantas, D.; Beljonne, D.; Anderson, H. L. Belt-Shaped πSystems: Relating Geometry to Electronic Structure in a Six-Porphyrin Nanoring. J. Am. Chem. Soc. 2011, 133, 17262−17273. (36) Murrell, J. N.; Pople, J. A. The Intensities of the Symmetryforbidden Electronic Bands of Benzene. Proc. Phys. Soc., London, Sect. A 1956, 69, 245−252. (37) Leach, S.; Vervloet, M.; Desprès, A.; Bréheret, E.; Hare, J. P.; Dennis, T. J.; Kroto, H. W.; Taylor, R.; Walton, D. R. M. Electronic Spectra and Transitions of the Fullerene C60. Chem. Phys. 1992, 160, 451−466. (38) Santoro, F.; Lami, A.; Improta, R.; Bloino, J.; Barone, V. Effective Method for the Computation of Optical Spectra of Large Molecules at Finite Temperature Including the Duschinsky and Herzberg−Teller Effect: The Qx Band of Porphyrin as a Case Study. J. Chem. Phys. 2008, 128, 224311. (39) Kano, H.; Saito, T.; Kobayashi, T. Dynamic Intensity Borrowing in Porphyrin J-Aggregates Revealed by Sub-5-fs Spectroscopy. J. Phys. Chem. B 2001, 105, 413−419. (40) Kano, H.; Saito, T.; Kobayashi, T. Observation of Herzberg− Teller-type Wave Packet Motion in Porphyrin J-Aggregates Studied by Sub-5-fs Spectroscopy. J. Phys. Chem. A 2002, 106, 3445−3453. (41) Matsushita, M.; Ketelaars, M.; van Oijen, A. M.; Köhler, J.; Aartsma, T. J.; Schmidt, J. Spectroscopy on the B850 Band of Individual Light-Harvesting 2 Complexes of Rhodopseudomonas acidophila II. Exciton States of an Elliptically Deformed Ring Aggregate. Biophys. J. 2001, 80, 1604−1614. (42) Becker, K.; Fritzsche, M.; Höger, S.; Lupton, J. M. Phenylene− Ethynylene Macrocycles as Model Systems of Interchromophoric Interactions in π-Conjugated Macromolecules. J. Phys. Chem. B 2008, 112, 4849−4853. (43) Aggarwal, A. V.; Thiessen, A.; Idelson, A.; Kalle, D.; Würsch, D.; Stangl, T.; Steiner, F.; Jester, S.-S.; Vogelsang, J.; Höger, S.; Lupton, J. M. Fluctuating Exciton Localization in Giant π-Conjugated Spokedwheel Macrocycles. Nat. Chem. 2013, 5, 964−970. (44) Thiessen, A.; Würsch, D.; Jester, S.-S.; Aggarwal, A. V.; Idelson, A.; Bange, S.; Vogelsang, J.; Höger, S.; Lupton, J. M. Exciton Localization in Extended π-Electron Systems: Comparison of Linear and Cyclic Structures. J. Phys. Chem. B 2015, 119, 9949−9958.

(15) Donehue, J. E.; Varnavski, O. P.; Cemborski, R.; Iyoda, M.; Goodson, T., III Probing Coherence in Synthetic Cyclic LightHarvesting Pigments. J. Am. Chem. Soc. 2011, 133, 4819−4828. (16) Kim, P.; Park, K. H.; Kim, W.; Tamachi, T.; Iyoda, M.; Kim, D. Relationship between Dynamic Planarization Processes and Exciton Delocalization in Cyclic Oligothiophenes. J. Phys. Chem. Lett. 2015, 6, 451−456. (17) Kim, W.; Sung, J.; Park, K. H.; Shimizu, H.; Imamura, M.; Han, M.; Sim, E.; Iyoda, M.; Kim, D. The Role of Linkers in the ExcitedState Dynamic Planarization Processes of Macrocyclic Oligothiophene 12-Mers. J. Phys. Chem. Lett. 2015, 6, 4444−4450. (18) Parkinson, P.; Kondratuk, D. V.; Menelaou, C.; Gong, J. Q.; Anderson, H. L.; Herz, L. M. Chromophores in Molecular Nanorings: When is a Ring a Ring? J. Phys. Chem. Lett. 2014, 5, 4356−4361. (19) Wong, B. M. Optoelectronic Properties of Carbon Nanorings: Excitonic Effects from Time-Dependent Density Functional Theory. J. Phys. Chem. C 2009, 113, 21921−21927. (20) Adamska, L.; Nayyar, I.; Chen, H.; Swan, A. K.; Oldani, N.; Fernandez-Alberti, S.; Golder, M. R.; Jasti, R.; Doorn, S. K.; Tretiak, S. Self-Trapping of Excitons, Violation of Condon Approximation, and Efficient Fluorescence in Conjugated Cycloparaphenylenes. Nano Lett. 2014, 14, 6539−6546. (21) Liu, J.; Adamska, L.; Doorn, S. K.; Tretiak, S. Singlet and Triplet Excitons and Charge Polarons in Cycloparaphenylenes: a Density Functional Theory Study. Phys. Chem. Chem. Phys. 2015, 17, 14613− 14622. (22) Park, K. H.; Kim, P.; Kim, W.; Shimizu, H.; Han, M.; Sim, E.; Iyoda, M.; Kim, D. Excited-State Dynamic Planarization of Cyclic Oligothiophenes in the Vicinity of a Ring-to-Linear Excitonic Behavioral Turning Point. Angew. Chem., Int. Ed. 2015, 54, 12711− 12715. (23) Steiner, F.; Bange, S.; Vogelsang, J.; Lupton, J. M. Spontaneous Fluctuations of Transition Dipole Moment Orientation in OLED Triplet Emitters. J. Phys. Chem. Lett. 2015, 6, 999−1004. (24) Ali, M. A.; Krishnan, M. S. Computational Studies on Cyclic [n]Paraphenyleneacetylenes Using Homodesmotic Reactions. Mol. Phys. 2009, 107, 2149−2158. (25) Segawa, Y.; Omachi, H.; Itami, K. Theoretical Studies on the Structures and Strain Energies of Cycloparaphenylenes. Org. Lett. 2010, 12, 2262−2265. (26) Esser, B. Theoretical Analysis of [5.5.6]Cyclacenes: Electronic Properties, Strain Energies and Substituent Effects. Phys. Chem. Chem. Phys. 2015, 17, 7366−7372. (27) Hoffmann, M.; Kärnbratt, J.; Chang, M.-H.; Herz, L. M.; Albinsson, B.; Anderson, H. L. Enhanced π Conjugation around a Porphyrin[6] Nanoring. Angew. Chem. 2008, 120, 5071−5074. (28) Hestand, N. J.; Spano, F. C. The Effect of Chain Bending on the Photophysical Properties of Conjugated Polymers. J. Phys. Chem. B 2014, 118, 8352−8363. (29) Yong, C.-K.; Parkinson, P.; Kondratuk, D. V.; Chen, W.-H.; Stannard, A.; Summerfield, A.; Sprafke, J. K.; O’Sullivan, M. C.; Beton, P. H.; Anderson, H. L.; Herz, L. M. Ultrafast Delocalization of Excitation in Synthetic Light-Harvesting Nanorings. Chem. Sci. 2015, 6, 181−189. (30) Cogdell, R. J.; Köhler, J. Use of Single-Molecule Spectroscopy to Tackle Fundamental Problems in Biochemistry: Using Studies on Purple Bacterial Antenna Complexes as an Example. Biochem. J. 2009, 422, 193−205. (31) Stangl, T.; Wilhelm, P.; Schmitz, D.; Remmerssen, K.; Henzel, S.; Jester, S.-S.; Höger, S.; Vogelsang, J.; Lupton, J. M. Temporal Fluctuations in Excimer-Like Interactions between π-Conjugated Chromophores. J. Phys. Chem. Lett. 2015, 6, 1321−1326. (32) Stangl, T.; Wilhelm, P.; Remmerssen, K.; Höger, S.; Vogelsang, J.; Lupton, J. M. Mesoscopic Quantum Emitters from Deterministic Aggregates of Conjugated Polymers. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, E5560−E5566. (33) Yang, J.; Ham, S.; Kim, T.-W.; Park, K. H.; Nakao, K.; Shimizu, H.; Iyoda, M.; Kim, D. Inhomogeneity in the Excited-State Torsional 1266

DOI: 10.1021/acs.jpclett.6b00360 J. Phys. Chem. Lett. 2016, 7, 1260−1266