Direct S0→T Excitation of a Conjugated Polymer Repeat Unit

Letter pubs.acs.org/JPCL

Direct S0→T Excitation of a Conjugated Polymer Repeat Unit: Unusual Spin-Forbidden Transitions Probed by Time-Resolved Electron Paramagnetic Resonance Spectroscopy Deborah L. Meyer,† Florian Lombeck,‡,§ Sven Huettner,∥ Michael Sommer,‡ and Till Biskup*,† †

Institut für Physikalische Chemie and ‡Institut für Makromolekulare Chemie, Albert-Ludwigs-Universität Freiburg, 79104 Freiburg, Germany § Optoelectronics Group, University of Cambridge, Cavendish Laboratory, Cambridge CB3 0HE, United Kingdom ∥ Organic and Hybrid Electronics, Macromolecular Chemistry I, Universität Bayreuth, 95440 Bayreuth, Germany S Supporting Information *

ABSTRACT: A detailed understanding of the electronic structure of semiconducting polymers and their building blocks is essential to develop efficient materials for organic electronics. (Time-resolved) electron paramagnetic resonance (EPR) is particularly suited to address these questions, allowing one to directly detect paramagnetic states and to reveal their spin-multiplicity, besides its clearly superior resolution compared to optical methods. We present here evidence for a direct S0→T optical excitation of distinct triplet states in the repeat unit of a conjugated polymer used in organic photovoltaics. These states differ in their electronic structure from those populated via intersystem crossing from excited singlet states. This is an additional and so far unconsidered route to triplet states with potentially high impact on efficiency of organic electronic devices.

O

unrivalled sensitivity and selectivity for paramagnetic states allowing for unambiguous assignment of the spin multiplicity and outperforming optical spectroscopy by far in terms of resolution. To gain insight into the electronic structure of polymers, starting with their building blocks has proven to be valuable.17 Here, we present results on Cbz-TBT (Figure 1), the repeat

rganic semiconductors have been widely studied over the last two decades and are currently used in a large range of applications such as light-emitting diodes,1 field-effect transistors,2 light detectors,3 and solar cells.4 Compared to conventional, inorganic, and mainly silicon-based electronics and photovoltaics, using organic polymers has a number of advantages, making it a fascinating and very promising endeavor with great potential for application. Perhaps most prominent is the capability of nearly endlessly and systematically tailoring molecules for the desired purpose using welldeveloped protocols of organic synthetic chemistry.5 A detailed understanding of the electronic structure of polymers and their building blocks is essential to develop efficient materials for organic electronics.6 Energy level matching has been regarded as one of the two key aspects for this task, besides controlling morphology.7 Furthermore, the role of triplets in organic electronics is still highly debated,8,9 ranging from major source of losses to boosting the efficiency due to processes such as triplet−triplet annihilation10,11 and singlet fission.12 Electron paramagnetic resonance (EPR) spectroscopy is particularly well-suited to probe the electronic structure of molecules on a microscopic level. Questions that can be addressed range from the delocalization length and geometry of excited states13,14 to investigation of paramagnetic defects in polymers forming traps for charge-carriers.15 Recently, we showed the power of time-resolved EPR (TREPR) for probing both the orientation and the degree of ordering of a polymer film on a substrate making use of the intrinsic orientation dependence of the signals of triplet excitons.16 The key is its © XXXX American Chemical Society

Figure 1. Chemical structure of Cbz-TBT. Cbz is colored in blue, TBT in red.

unit of the efficient and air- and temperature-stable18 polymer PCDTBT (poly[N-9′-heptadecanyl-2,7-carbazole-alt-5,5-(4′,7′di-2-thienyl-2′,1′,3′-benzothiadiazole)]).19,20 This molecule is a push−pull system comprising an electron-rich carbazole (Cbz) moiety and an electron-deficient dithiophene-benzothiadiazole (TBT) unit, the latter in itself a push−pull system. The push− pull character gives rise to a prominent optical absorption band Received: March 16, 2017 Accepted: March 27, 2017 Published: March 27, 2017 1677

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682

Letter

The Journal of Physical Chemistry Letters

the polymer, as apparent from the values for the parameter |D| of the ZFS tensor. [However, the overall spectral shape seems to be dominated by TBT rather than Cbz, as apparent from preliminary data on TBT alone, clearly resembling those from Cbz-TBT and PCDTBT.] The shape (and intensity) of the triplet spectra does not change for excitation wavelengths throughout the whole CT band. Quite to our surprise, excitation red-shifted from the CT band, where no absorption can be detected in the optical spectrum, gives rise to signals of triplet species in TREPR spectroscopy that are up to twice more intense than those from excitation in the CT band. Even more, these signals change their shape in a consistent manner toward larger delocalization lengths of the triplet exciton, as obvious from the decreasing |D| values (cf. Table 1). Both, the

in the visible range, centered at 492 nm with a width (fwhm) of about 100 nm,17 usually termed charge-transfer (CT) band. However, there is no visible absorption of this molecule beyond 600 nm toward the long-wavelength range. If excited in its CT band, Cbz-TBT readily forms short-lived triplet states that manifest themselves as characteristic spectra in TREPR spectroscopy (Figure 2; for details of experiments

Table 1. Parameters (D, E) of the ZFS Tensor Together with the Populations (p1, p2, p3) of the Triplet Sublevels and the Lorentzian Lineshape Γ As Obtained from Spectral Simulations of the TREPR Spectra Shown in Figure 2a λ/nm 492 630 650 680

|D|/MHz 1361.6 1344.7 1317.2 1288.5

± ± ± ±

3.0 1.5 1.4 1.3

|E|/MHz

p1

p2

p3

Γ/mT

± ± ± ±

0 0 0 0

0.138 0.098 0.068 0.043

0.862 0.902 0.932 0.957

3.42 2.08 1.83 1.54

75.9 77.7 75.0 73.7

1.4 0.7 0.7 0.6

λ is the laser wavelength used for optical excitation. For further details, see the Supporting Information.

a

clear trend in shape of the TREPR spectra as well as the ZFS parameters obtained from spectral simulations that excellently resemble the experimental data leaves us with no doubt that the observed effects are real and provide insight into the electronic structure of the underlying triplet states. Quantitative Analysis of TREPR Spectra. Normally, TREPR spectra cannot be analyzed in a quantitative fashion, as signal intensity strongly depends on a series of experimental and device parameters. Even more fundamentally, these spectra consist of overlapping contributions of differently polarized (emissive or absorptive) lines that partially compensate each other, making the resulting signals particularly sensitive to even slight changes in the electronic structure of the investigated system. Here, we have taken special precautions and measured the identical sample under entirely identical conditions with only the excitation wavelength changing. All other parameters have been kept identical, and the sample has not been touched between measurements. Furthermore, as evident from the simulation parameters (Table 1) and the experimental data (Figure 2), the polarization pattern does not change dramatically. Thus, for the series of TREPR measurements whose signal intensity has been plotted versus the optical absorption spectrum (Figure 3), a direct comparison is possible. [Note that only the six measurements recorded with excitation wavelengths between 430 and 680 nm in 50 nm steps can be compared quantitatively. Additional spectra recorded at other wavelengths (492 nm, 650 nm) fit perfectly with respect to their spectral shape, but cannot be compared quantitatively due to changes of experimental parameters. For further details about the quantitative analysis, see the Supporting Information.] However, no quantification of the triplet yield in terms of quantum efficiencies can be drawn from the TREPR data. Obviously, the channel for forming the triplet states observed at excitation wavelengths red-shifted with respect to the CT band is in its maximum about twice as effective as the one in

Figure 2. TREPR spectra of Cbz-TBT excited at different wavelengths. All spectra are averages over 200 ns centered about 500 ns after optical excitation. Whereas excitation in the CT band of Cbz-TBT always results in spectra of identical shape (and intensity), red-shifted excitation leads to spectrally distinct states with increasing delocalization length and correspondingly narrower spectra, as indicated by gray arrows. Spectra have been normalized to the same area.

and simulations, see the Supporting Information). These spectra are entirely dominated by the zero-field splitting (ZFS) interaction resulting from the dipolar coupling of the two unpaired electron spins of the triplet state,21,22 and the absolute values for the two characteristic parameters D and E of the corresponding interaction tensor can be directly extracted from the signal. As the light-induced triplet states are initially created with their three sublevels populated far from thermal equilibrium, this gives rise to a huge signal enhancement of the corresponding TREPR spectra that show signals in enhanced absorption (A) and emission (E), respectively. This signal enhancement allows one to omit the lock-in detection scheme used in conventional cw-EPR and therefore to increase time resolution up to about 10 ns.23,24 The overall shape and width of the TREPR signals of CbzTBT resembles those of the light-induced triplet states in the corresponding polymer PCDTBT.16 Not surprisingly, CbzTBT shows a slightly smaller delocalization length compared to 1678

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682

Letter

The Journal of Physical Chemistry Letters

TBT in the CT band leads always to triplet states with identical electronic structure and intensity. Nevertheless, the overall electronic structure of all triplet states and particularly the geometry of the transition dipole moments involved are generally rather similar, as the populations of the three triplet sublevels, and in turn the overall spectral shape, do not change substantially. Nature of the Dif ferent Triplet States. Whereas excitation within the CT band always results in a triplet state with identical electronic structure, red-shifted excitation leads to distinct triplet states as evidenced by the different ZFS parameters of their TREPR spectra. Whether this different electronic structure is due to separate electronic triplet states Tm or reflects different vibronic states of the same electronic triplet state cannot be decided unequivocally from the TREPR data themselves. In a naive picture based on the anharmonic oscillator, if it were different vibronic states of the same electronic triplet state, one would expect the delocalization to be larger for higher-lying vibronic states. As the opposite is true for the triplet states observed here using TREPR spectroscopy, this may be a hint for distinct electronic states. [Further insight might be gained by obtaining optical excitation spectra of the S0→T transition at very low temperatures. This is, however, clearly out of scope of the present study.] Additionally, vibronic states that cannot interconvert would violate Kasha’s rule.26 The pure Lorentzian line broadening is an additional argument in favor of distinct electronic states, as excited vibronic states should generally lead to additional Gaussian broadening. The same would be true for aggregates, as their inhomogeneous distribution would also cause Gaussian line broadening. Furthermore, the decrease in Lorentzian line width with increasing excitation wavelength points clearly toward increasing lifetime of the respective triplet state, consistent with its progressively lower energy. Aggregates and Multiphoton Processes. Alternative explanations for the unusual formation of triplet states when exciting the molecule red-shifted from its optical absorption include aggregates formed between close-lying Cbz-TBT molecules or multiphoton processes during excitation that would allow one to bridge the band gap between ground and excited state within the singlet manifold. Both can be excluded for the data presented in this contribution. Aggregates of Cbz-TBT molecules are in stark contrast to the exclusive Lorentzian (homogeneous) line broadening of the observed TREPR data, as aggregation would in nearly all cases lead to a certain degree of inhomogeneity in the sample reflected in a substantial contribution of Gaussian (inhomogeneous) line broadening. To further exclude aggregation of Cbz-TBT under the conditions used for the TREPR spectroscopy, we recorded absorption spectra at lower temperatures (cf. Figure S5). The solvent used for the TREPR spectroscopy (o-DCB) does not form a glass at cryogenic temperatures, and no changes in the absorption of Cbz-TBT could be detected above the melting point of the solvent (about −18 °C). Additional measurements in a different, glass-forming solvent (2-methyltetrahydrofuran) gave clear indications of aggregation at low temperatures from optical spectroscopy that were reflected in complicated TREPR spectra different from those obtained in o-DCB that cannot be described with a single triplet species (data not shown). Taken together, we therefore exclude aggregate formation as a possible explanation for the triplet formation when exciting the sample red-shifted from its absorption band (i.e., the S0→S1 band gap).

Figure 3. Comparison of the absorption spectrum of Cbz-TBT and the triplet yield as determined by TREPR signal intensity. Not only can triplet states be observed for excitation beyond the absorption band of Cbz-TBT, but their signal intensity is even stronger, with a maximum at about 650 nm. This is evidence for a (spin-forbidden) direct S0→T optical excitation with higher triplet yield compared to the (conventional) triplet pathway via ISC from S1.

operation when exciting within the CT band. Furthermore, despite the dramatic differences in oscillator strength for the S0→S1 optical transition in the CT band when exciting in its maximum at 492 nm or in the outer wings (430 nm or even 580 nm), the triplet yield of the underlying process appears to be unaffected. This could be explained by processes competing with the intersystem crossing (ISC) from S1→T that are proportional to the probability of the S0→S1 transition. As TREPR directly probes and is highly sensitive to the triplet state of Cbz-TBT, the signals obtained for excitation red-shifted from the CT band (and any optical absorption) are clear evidence for a spin-forbidden direct S0→T optical excitation. From these observations, a number of questions arise that shall be addressed hereafter. Probability of Spin-Forbidden Optical Transitions. As excitation red-shifted from the optical absorption gives rise to even stronger triplet signals, the most obvious explanation would be a spin-forbidden direct S0→T transition. Generally, processes of such kind have been discussed and observed early on for rather simple organic molecules25−28 and even for flavins in biological context.29 Whereas the theoretical limit of the probability of this transition is zero, the experimentally observed minimum extinction coefficients εmin are about 10−4, and the corresponding maximum values εmax ≈ 100 for molecules comprising heavy atoms.31 With an estimated extinction coefficient for the spinallowed transition in the CT band of Cbz-TBT of εCT ≈ 103− 104, the difference in oscillator strength is at least on the order of 104 if not larger, rendering the S0→T transition invisible in optical spectroscopy under ambient conditions. EPR, by contrast, is selectively sensitive to paramagnetic states such as triplets that can therefore be detected, unequivocally assigned, and characterized. Dif ferences in the Electronic Structure of the Observed Triplet States. The increasing delocalization (decreasing D value; cf. Table 1) with increasing wavelength of the optical excitation beyond the CT band is consistent with exciting into energetically progressively lower-lying triplet states. Furthermore, these triplet states appear to be distinct states that cannot interconvert into each other under the given experimental conditions. Otherwise, one could not observe these states with their distinct ZFS parameters. Quite in contrast, exciting Cbz1679

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682

Letter

The Journal of Physical Chemistry Letters

Figure 4. Term schemes depicting the two distinct pathways for triplet formation. When excited in the CT band, triplet formation proceeds via ISC from S1 (left), while direct S0→Tm transitions to distinct triplet states occur for red-shifted excitation wavelengths (right). The different triplet states Tm+x do not interconvert under the given experimental conditions (80 K, frozen solution).

of the observed transitions. Furthermore, supporting quantumchemical calculations are much easier for small molecules such as building blocks and generally restricted to small oligomers. Recently, it has been demonstrated that computational results obtained for monomers and small oligomers can easily be extrapolated to explain the behavior observed experimentally in polymers.35 This makes us confident that our results obtained for the repeat unit Cbz-TBT of the polymer PCDTBT can readily be extrapolated to the polymer. Potential Impact for Polymers and Organic Electronic Devices. The role of triplet states in organic electronics is still highly debated.36 At least in the context of organic photovoltaics, they are usually regarded to be detrimental for the overall device efficiency.37 The energy of the low-lying triplet states of the donor is too far below the LUMO level of the acceptor for these triplet states to contribute substantially to charge separation. Low-lying triplet states that cannot be populated from the excited singlet state, but directly via optical (spinforbidden) transition from the singlet ground state, as proposed here for the Cbz-TBT molecule, extend and complete the existing picture of pathways leading to triplet states. Interestingly, direct optical excitation of these low-lying triplet states seems to be a more efficient route in the material investigated here than via ISC from the excited singlet state. In polymers, long-wavelength radiation (e.g., near-IR) as available from ambient sunlight could well lead to forming unproductive, energetically low-lying triplet states that are inaccessible for charge separation, thus contributing to the overall loss of efficiency of organic photovoltaic devices. However, using acceptors with orbitals energetically matching those of these low-lying triplet states could help boosting the quantum efficiency of the devices by using long-wavelength irradiation for productive charge separation. Additionally, one could make use of pathways to repopulate the singlet state, either thermally38,39 or by triplet−triplet annihilation.10 These processes lead to delayed fluorescence in organic light-emitting diodes, therefore greatly enhancing their efficiency. Taken together, we could show that TREPR spectroscopy is well-suited to characterize excited states in polymer semiconductors and their building blocks and provides unique

Multiphoton processes, on the other hand, could explain how to excite the molecule from the singlet ground state to an excited singlet state with radiation of less energy than the band gap. However, multiphoton processes only take place with a sufficient probability above a certain photon density, a condition not fulfilled with our setup. Furthermore, varying the excitation power does not alter the spectral shape of the resulting triplet spectra, only their intensity (and in turn, their signal-to-noise ratio). Two Distinct Routes toward Triplet States in Cbz-TBT. In CbzTBT, there appear to be two different routes leading to triplet states with distinct electronic properties: one via ISC (S1→Tn) from the optically excited singlet state S1, and the other via direct population of triplet states (S0→Tm) from the singlet ground state (Figure 4). The striking overall similarity of the observed triplet states putatively populated via these two distinct mechanisms could be explained by the high similarity of the involved singlet states S1 and S0, respectively. This is supported by absorption and emission spectra of Cbz-TBT being close to mirror-images of each other.17 The experimental data clearly show that the probabilities 7 of the two routes are different, the latter surprisingly being more likely than the former, thus 7(S1 → Tn) < 7(S0 → Tm)

(1)

This means that exciting Cbz-TBT, and potentially the polymer built from it, with long-wavelength electromagnetic radiation selectively populates low-lying and potentially long-lived triplet states. The Importance of Investigating Building Blocks. Although usually polymers are used in organic electronic devices, thoroughly investigating the respective building blocks is required for gaining a fundamental understanding of the electronic structure and its impact on the characteristics of the polymer. In the case of the present study, only with using building blocks instead of polymers can one have a chance to observe direct S0→T transitions. In a polymer, tail states lead to a fade-out of the absorption bands toward longer wavelengths,32−34 therefore preventing an unambiguous assignment 1680

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682

Letter

The Journal of Physical Chemistry Letters

(7) Jackson, N. E.; Savoie, B. M.; Marks, T. J.; Chen, L. X.; Ratner, M. A. The next breakthrough for organic photovoltaics? J. Phys. Chem. Lett. 2015, 6, 77−84. (8) Yost, S. R.; Hontz, E.; Yeganeh, S.; Van Voorhis, T. Triplet vs singlet energy transfer in organic semiconductors: the tortoise and the hare. J. Phys. Chem. C 2012, 116, 17369−17377. (9) Veldman, D.; Meskers, S. C. J.; Janssen, R. A. J. The energy of charge-transfer states in electron donor−acceptor blends: insight into the energy losses in organic solar cells. Adv. Funct. Mater. 2009, 19, 1939−1948. (10) Gehrig, D. W.; Howard, I. A.; Laquai, F. Charge carrier generation followed by triplet state formation, annihilation, and carrier recreation in PDBTTT-C/PC60BM photovoltaic blends. J. Phys. Chem. C 2015, 119, 13509−13515. (11) Yanai, N.; Kimizuka, N. Recent emergence of photon upconversion based on triplet energy migration in molecular assemblies. Chem. Commun. 2016, 52, 5354−5370. (12) Smith, M. B.; Michl, J. Singlet fission. Chem. Rev. 2010, 110, 6891−6936. (13) Niklas, J.; Mardis, K. L.; Banks, B. P.; Grooms, G. M.; Sperlich, A.; Dyakonov, V.; Beaupré, S.; Leclerc, M.; Xu, T.; Yu, L.; et al. Highly-efficient charge separation and polaron delocalization in polymer−fullerene bulk-heterojunctions: a comparative multi-frequency EPR and DFT study. Phys. Chem. Chem. Phys. 2013, 15, 9562−9574. (14) Kobori, Y.; Miura, T. Overcoming coulombic traps: geometry and electronic characterizations of light-induced separated spins at the bulk heterojunction interface. J. Phys. Chem. Lett. 2015, 6, 113−123. (15) Susarova, D. K.; Piven, N. P.; Akkuratov, A. V.; Frolova, L. A.; Polinskaya, M. S.; Ponomarenko, S. A.; Babenko, S. D.; Troshin, P. A. ESR spectroscopy as a powerful tool for probing the quality of conjugated polymers designed for photovoltaic applications. Chem. Commun. 2015, 51, 2239−2241. (16) Biskup, T.; Sommer, M.; Rein, S.; Meyer, D. L.; Kohlstädt, M.; Würfel, U.; Weber, S. Ordering of PCDTBT revealed by time-resolved electron paramagnetic resonance spectroscopy of its triplet excitons. Angew. Chem., Int. Ed. 2015, 54, 7707−7710. (17) Banerji, N.; Gagnon, E.; Morgantini, P.-Y.; Valouch, S.; Mohebbi, A. R.; Seo, J.-H.; Leclerc, M.; Heeger, A. J. Breaking down the problem: optical transitions, electronic structure, and photoconductivity in conjugated polymer PCDTBT and in its separate building blocks. J. Phys. Chem. C 2012, 116, 11456−11469. (18) Cho, S.; Seo, J. H.; Park, S. H.; Beaupré, S.; Leclerc, M.; Heeger, A. J. A thermally stable semiconducting polymer. Adv. Mater. 2010, 22, 1253−1257. (19) Park, S. H.; Roy, A.; Beaupré, S.; Cho, S.; Coates, N.; Moon, J. S.; Moses, D.; Leclerc, M.; Lee, K.; Heeger, A. J. Bulk heterojunction solar cells with internal quantum efficiency approaching 100%. Nat. Photonics 2009, 3, 297−303. (20) Lombeck, F.; Komber, H.; Fazzi, D.; Nava, D.; Kuhlmann, J.; Stegerer, D.; Strassel, K.; Brandt, J.; de Zerio Mendaza, A. D.; Müller, C.; et al. On the effect of prevalent carbazole homocoupling defects on the photovoltaic performance of PCDTBT:PC71BM solar cells. Adv. Energy Mater. 2016, 6, 1601232. (21) Stevens, K. W. H. The spin-Hamiltonian and line widths in nickel Tutton salts. Proc. R. Soc. London, Ser. A 1952, 214, 237−246. (22) Hutchison, C. A., Jr.; Mangum, B. W. Paramagnetic resonance absorption in naphthalene in its phosphorescent state. J. Chem. Phys. 1961, 34, 908−922. (23) Biskup, T. Time-resolved EPR of radical pair intermediates in cryptochromes. Mol. Phys. 2013, 111, 3698−3703. (24) Forbes, M. D.; Jarocha, L. E.; Sim, S.; Tarasov, V. F. Timeresolved electron paramagnetic resonance spectroscopy: history, technique, and application to supramolecular and macromolecular chemistry. Adv. Phys. Org. Chem. 2013, 47, 1−83. (25) Lewis, G. N.; Kasha, M. Phosphorescence in fluid media and the reverse process of singlet−triplet absorption. J. Am. Chem. Soc. 1945, 67, 994−1003.

insight that cannot be obtained by other spectroscopic methods. TREPR spectroscopy is particularly superior to optical methods for investigating triplet states, as it allows one to directly detect triplet states and to identify the observed spectra with a chemical species, in this case the Cbz-TBT molecule. By contrast, optical spectroscopy, particularly with long path lengths, suffers strongly from potential impurities in the solvent and the lack of direct assignment of the absorption bands to a distinct molecule, besides the fact that it cannot detect the spin multiplicity of an observed species. Additionally, the experimental evidence presented here for spin-forbidden direct optical S0→T transitions further advances our picture of possible triplet-forming processes in these materials. This direct triplet excitation is probably much more widespread than previously anticipated and may have a substantial impact on photophysical and photochemical processes in semiconducting polymers.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.7b00644.

■

Details of experiments and simulations (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Michael Sommer: 0000-0002-2377-5998 Till Biskup: 0000-0003-2913-0004 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

■

REFERENCES

The authors thank the German Research Foundation (DFG, Grant BI-1249/3-1 to T.B., SPP1355 to M.S.) and AlbertLudwigs-Universität Freiburg (Innovationsfonds Forschung, to T.B. and M.S.) for financial support, G. Kothe for helpful comments, and S. Weber for continuous support. S.H. is thankful to the Bavarian framework program Soltech for financial support.

(1) Minaev, B.; Baryshnikov, G.; Agren, H. Principles of phosphorescent organic light emitting devices. Phys. Chem. Chem. Phys. 2014, 16, 1719−1758. (2) Torsi, L.; Magliulo, M.; Manoli, K.; Palazzo, G. Organic fieldeffect transistor sensors: a tutorial review. Chem. Soc. Rev. 2013, 42, 8612−8628. (3) Baeg, K.-J.; Binda, M.; Natali, D.; Caironi, M.; Noh, Y.-Y. Organic light detectors: photodiodes and phototransistors. Adv. Mater. 2013, 25, 4267−4295. (4) Mazzio, K. A.; Luscombe, C. K. The future of organic photovoltaics. Chem. Soc. Rev. 2015, 44, 78−90. (5) Guo, X.; Baumgarten, M.; Müllen, K. Designing π-conjugated polymers for organic electronics. Prog. Polym. Sci. 2013, 38, 1832− 1908. (6) Scharber, M. C.; Mühlbacher, D.; Koppe, M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. J. Design rules for donors in bulkheterojunction solar cellstowards 10% energy-conversion efficiency. Adv. Mater. 2006, 18, 789−794. 1681

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682

Letter

The Journal of Physical Chemistry Letters (26) Kasha, M. Characterization of electronic transitions in complex molecules. Discuss. Faraday Soc. 1950, 9, 14−19. (27) Evans, D. F. Perturbation of singlet−triplet transitions of aromatic molecules by oxygen under pressure. J. Chem. Soc. 1957, 1351−1357. (28) Goodman, L.; Laurenzi, B. J. Probability of singlet−triplet transitions. Adv. Quantum Chem. 1968, 4, 153−169. (29) Delbrück, M.; Katzir, A.; Presti, D. Responses of Phycomyces indicating optical excitation of the lowest triplet state of riboflavin. Proc. Natl. Acad. Sci. U. S. A. 1976, 73, 1969−1973. (30) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular Photochemistry of Organic Molecules; University Science Books: Sausalito, CA, 2010. (31) Reference 30, p. 200. (32) Bässler, H.; Köhler, A. Charge transport in organic semiconductors. Top. Curr. Chem. 2011, 312, 1−66. (33) Bässler, H.; Köhler, A. “Hot or cold”: how do charge transfer states at the donor−acceptor interface of an organic solar cell dissociate? Phys. Chem. Chem. Phys. 2015, 17, 28451−28462. (34) Rivnay, J.; Noriega, R.; Northrup, J. E.; Kline, R. J.; Toney, M. F.; Salleo, A. Structural origin of gap states in semicrystalline polymers and the implications for charge transport. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 121306. (35) Larsen, R. E. Simple extrapolation method to predict the electronic structure of conjugated polymers from calculations on oligomers. J. Phys. Chem. C 2016, 120, 9650−9660. (36) Köhler, A.; Bässler, H. Triplet states in organic semiconductors. Mater. Sci. Eng., R 2009, 66, 71−109. (37) Rao, A.; Chow, P. C. Y.; Gélinas, S.; Schlenker, C. W.; Li, C.-Z.; Yip, H.-L.; Jen, A. K.-Y.; Ginger, D. S.; Friend, R. H. The role of spin in the kinetic control of recombination in organic photovoltaics. Nature 2013, 500, 435−439. (38) Uoyama, H.; Goushi, K.; Shizu, K.; Nomura, H.; Adachi, C. Highly efficient organic light-emitting diodes from delayed fluorescence. Nature 2012, 492, 234−238. (39) Lee, S. Y.; Yasuda, T.; Komiyama, H.; Lee, J.; Adachi, C. Thermally activated delayed fluorescence polymers for efficient solution-processed organic light-emitting diodes. Adv. Mater. 2016, 28, 4019−4024.

1682

DOI: 10.1021/acs.jpclett.7b00644 J. Phys. Chem. Lett. 2017, 8, 1677−1682