Exciton dynamics of a fused ring π-conjugated nonfullerene molecule based on dithienonaphthobisthiadiazole

YS3 was synthesized according to Scheme S1 and was characterized by NMR measurements and high-resolution mass spectroscopy (HRMS). YS3 was soluble in common organic solvents such as chlorobenzene at 110 °C. The structure of YS3 was identified with ¹H NMR and HRMS (Fig. S1).

For steady-state absorption and photoluminescence (PL) spectra, time-resolved PL decay and spectra, and PL quantum yield (PLQY) measurements, chlorobenzene solution and neat films of YS3 molecules were prepared. For spectroscopic measurements, the YS3 chlorobenzene solution was prepared to be 1.2 μM and purged by Ar bubbling for 15 min before the measurements. For film preparation, YS3 was dissolved in chlorobenzene with a weight concentration of 2.0 mg ml⁻¹ and heated at 110 °C for about 2 h. The YS3 neat films were fabricated on a quartz substrate (20 × 20 mm²) by spin-coating from the chlorobenzene solution at a spin rate of 1000 rpm for 30 s in a glovebox (Korea Kiyon, KK-011AS). Finally, the neat films were encapsulated with another glass plate by using epoxy resin to prevent oxygen exposure to the film.

Absorption and PL spectra are measured with a spectrophotometer (Hitachi, U-4100) and a fluorescence spectrophotometer (Horiba Jobin Yvon, NanoLog), respectively. Time-resolved PL decay and spectra were measured with a time-correlated single-photon-counting system (Horiba, FluoroCube). The PLQY of solution samples was measured with an absolute PLQY spectrometer (Hamamatsu Photonics, Quantaurus-QY Plus C13534-01), and that of film samples was measured with an absolute PL/EL quantum yield spectrometer (Bunkoukeiki, BEL-300). The excitation wavelength was set at 560 nm for solution samples and at 530 nm for film samples to avoid scattering of the excitation light. Grazing incidence wide-angle X-ray diffraction (GIXD) measurement of the thin film was performed with a reported procedure in the beamline BL46XU at SPring-8. ²⁴⁾ The coherence length (L_C) was estimated from the simplified Scherrer's Eq. (1), where FWHM is the full-width at half-maximum of the π–π stacking diffraction peak

$$\begin{eqnarray}&&{L}_{{\rm{C}}}=\displaystyle \frac{2\pi }{{\rm{FWHM}}}.\end{eqnarray} \tag{ 1 }$$

Geometry optimization and electronic structure were calculated by using density functional theory (DFT) at the B3LYP/31-G(d) level. Calculation was performed by using the Gaussian 09 program. ²⁵⁾

The absorption and PL spectra are shown in Fig. 2. As shown in the figure, YS3 in solution exhibits a broad absorption band at around 600 nm while the neat film exhibits sharp vibrational bands originating from 0–0 and 0–1 transitions at around 740 and 680 nm, respectively. On the basis of the 0–0 bands, the Stokes shift was estimated to be 0.34 eV for the solution and to be 0.11 eV for the neat film. The red-shifted absorption suggests longer effective conjugation length in thin films than in solution. This is probably due to the interaction between the carbonyl group of the terminal indanone group and the sulfur of the adjacent thiophenyl group, which would induce a planar structure of π-conjugation. Interestingly, as shown in Fig. 2(b), the intensity of the 0–0 band was much larger than that of the 0–1 band in the absorption spectrum while the intensity of the 0–0 band was much smaller than that of the 0–1 band in the PL spectrum. This finding suggests that there are both J- and H-aggregates in the film state. ²⁶⁾ The PL originates mainly from more stable H-aggregates.

Zoom In Zoom Out Reset image size

The DFT calculation was performed for the model structure at the B3LYP/31 G. The long side chains were replaced with a methyl group to save computational cost. The π-conjugation of this molecule was almost flat in the optimized structure as shown in Fig. 3. As mentioned before, this is due to the planar structure of dithienonaphthobisthiadiazole, placement of the alkyl groups, and S···O interaction between the alkylthiophenes and the electron withdrawing end groups. ^27,28)

Zoom In Zoom Out Reset image size

Figure 4 shows the two-dimensional (2D) GIXD pattern of the YS3 neat film. Diffractions that most likely correspond to the lamellar and π–π stacking structures were observed in the ∼q_z and q_xy axes, respectively. This indicates the edge-on orientation or possibly end-on orientation of the molecules with respect to the substrate. The lamellar distance d_L and the π–π stacking distance d_π–π were evaluated to be 20.6 Å and 3.65 Å, respectively. The coherence length L_C was evaluated to be 43.4 Å for the π–π stacking structure. Such a crystalline order would originate in the highly coplanar molecular structure.

Zoom In Zoom Out Reset image size

Exciton lifetimes were evaluated by time-resolved PL decay and spectra. In solution, as shown in Fig. 5(a), the PL intensity rapidly decayed in less than 1 ns. This decay was fitted with a three-exponential function. The average lifetime was estimated to be 0.14 ns. On the other hand, as shown in Fig. 6(a), the time-resolved PL spectra show that the PL peak was gradually red shifted from 730 to 750 nm. The peak shift is probably due to structural relaxation in the excited state because YS3 has free-rotational single bonds. This is consistent with the larger Stokes shift observed for the solution samples, suggesting a large difference in molecular structures between the ground and excited states. At a later time stage, a new PL band was observed at around 750 nm, which is almost the same as that observed for neat films. Thus, it may be indicative of the formation of H-aggregated excimer-like sites even in solution.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

For neat films, as shown in Fig. 5(b), the PL intensity decayed slowly with a time constant of longer than 1 ns. Such a long exciton lifetime would be partially because of the large core unit, as reported previously. ²³⁾ As summarized in Table I, this time constant was one order of magnitude larger than that for the solution. In contrast to the solution, as shown in Fig. 6(b), no peak shift was observed in the time-resolved PL spectra of the neat film. The PL spectrum at 0 ns was the same as that observed for the steady-state PL spectrum originating from H-aggregates. Thus, this slow dynamics is ascribed to singlet excitons in H-aggregates in the neat films.

Finally, we discuss radiative and nonradiative transition rates in the solution and film states. ²⁹⁾ Table II summarizes radiative and nonradiative transition rates evaluated from the PL lifetime and PLQY. As shown in the table, the radiative transition rate in chlorobenzene solution was more than two orders of magnitude larger than that in the film. This is probably because the PL of the neat films is ascribed to forbidden transition due to H-aggregates ³⁰⁾ where π–π stacking and lamella structures are formed because of the highly coplanar structure in the film state due to S···O interaction between the sulfur of the core unit and the oxygen of the end group. There are some examples of closely packed structures by S···O interaction. ^27,28) On the other hand, as shown in the table, the nonradiative transition rate in solution was 10 times larger than that in neat films. This is partly because several single bonds can freely rotate in solution but not in a crystalline film. As mentioned before, YS3 is highly crystalline because the lamella diffraction patterns are observed up to the third order and the coherence length L_C was as large as 43.4 Å in the π–π stacking direction. As reported previously, ³¹⁾ Y6 and Y6-C2 neat films have shown clear lamella patterns. The coherence length L_C in the π–π direction was evaluated to be 34.9 Å for Y6 and 38.2 Å for Y6-C2 neat films. In addition, a long exciton lifetime of >1 ns has been reported for Y6 neat films. ^18,19) Thus, we conclude that such high crystallinity of YS3 would be also one of the key parameters to suppress nonradiative transition and therefore enhance exciton lifetime effectively.