Dialkoxyphenyldithiophene-based small molecules with enhanced absorption for solution processed organic solar cells

Abstract

Compared to the popular benzodithiophene (BDT) unit, dialkoxyphenyldithiophene (PDT) has the advantages of increased planarity along the molecular backbone (decreased dihedral angles between PDT and π bridges) and stronger donating ability, which may lead to better absorption, higher hole mobility and even higher power conversion efficiency (PCE). In this paper, two small molecules (CNO2TPDT and RD2TPDT) with PDT unit as a donor, bithiophene (2T) as π bridges, and oxoalkylated nitrile (CNO) or 3-ethylrhodanine (RD) as end-capped acceptors were synthesized and characterized. As expected, CNO2TPDT and RD2TPDT showed excellent absorption with edges of 720 and 745 nm, respectively, which are red-shifted by ca. 30 nm compared with their BDT analogues. A CNO2TPDT-based device exhibited relatively high V_oc of 0.87 V but a low PCE of 4.16% due to a large phase separation. In contrast, a RD2TPDT-based device exhibited higher PCE of 6.64% with a J_sc of 12.74 mA cm⁻² due to its red-shift absorption, better miscibility with [6,6]-phenyl-C₇₁ butyric acid methylester (PC₇₁BM) and higher hole mobility. This study demonstrated rational molecule design and effective morphology control can lead to good absorption ability and hole mobility, indicating the possibility for obtaining higher J_sc and PCE.

---

Introduction

Organic solar cells have attracted intense interest due to their advantages of low cost, solution process and flexibility.^1–3 In past decades, power conversion efficiencies (PCEs) of ca. 10% have been achieved for small-molecule organic solar cells (SM OSCs).^4–6 SM OSCs have the advantages of easier band structure control and less batch-to-batch variations in contrast to polymer organic solar cells (PSCs).⁷ Acceptor–donor–acceptor (A–D–A) type small molecular donors have exhibited the highest performance in SM OSCs, owing to suitable energy level, wide absorption and high mobility.^6,8 Common donor units including benzodithiophene (BDT),^8–10 dithienosilole (DTS),^11–13 etc. and acceptor units including oxoalkylated nitrile,¹⁴ rhodanine,⁸ indanedione,¹⁵ etc. are widely used in A–D–A type small-molecule donors. Although BDT-based small-molecule donors have shown promising performance in OSCs, higher performance devices are limited by their absorption spectra due to low lying highest occupied molecular orbital (HOMO) energy level. High-performance small-molecule donors should be delicately designed by considering the balance of short-circuit current density (J_sc), open-circuit voltage (V_oc) and fill factor (FF). Introduction of stronger acceptor units in BDT-based small molecules for lowering the bandgap to obtain better absorption is effective, but it is difficult to develop novel strong acceptor units.^16–18 Another rational strategy for achieving good absorption and higher J_sc is by introducing simple and easily accessible stronger donor units to form relatively planar backbone in molecular design, which can supply various facile methods. Although an appropriate stronger donor may slightly sacrifice V_oc value, subtle compromise can be achieved through increased absorption ability and hole mobility.

Given the advantages of a simple structure and easier synthesis, the dialkoxyphenyldithiophene (PDT) unit has been used as donor in high-performance PSCs.^19,20 Its stronger electron donating ability than the BDT unit and maintaining good conjugation by O⋯S weak noncovalent interaction between phenyl oxygen atom and adjacent thiophene sulphur atom in PDT are supposed to reduce optical bandgaps and improve absorption ability. Moreover, the O⋯S weak interaction makes PDT units approach a planar configuration and simultaneously leads to smaller torsional angles between PDT and adjacent thiophene π bridge (Scheme 1a), which benefits the coplanarity of molecular backbones and intermolecular π–π stacking, resulting in improved charge transport characteristics.^21,22 Therefore, PDT-based small molecules are expected to have high J_sc and FF values. Besides, compared with the rigid backbone of donor units such as BDT and DTS, PDT units may be beneficial in terms of the excellent solubility of PDT-based small-molecule donors. Therefore, the PDT unit might be a promising building block for high-performance solution-processed SM OSCs.

Scheme 1 (a) Breaking of BDT to form PDT (average dihedral angles were calculated by DFT calculation). (b) Chemical structures of CNO2TPDT and RD2TPDT (with calculated dihedral angles). (c) Geometries of optimized structures of CNO2TPDT and RD2TPDT.

In this study, we designed and synthesized two small molecules (CNO2TPDT and RD2TPDT) with simple PDT as donor, 3,4′-dihexyl-2,2′-bithiophene (2T) as π bridge and two common end-capped groups, oxoalkylated nitrile (CNO) and 3-ethylrhodanine (RD) units, as acceptors. Our previous work reported a small-molecule donor end-caped with CNO with relatively low HOMO energy level,¹⁴ indicating CNO as acceptor units may compensate the decreased V_oc of PDT donor-based devices. Besides, with RD as acceptor units, small molecules can obtain good film quality.²³ To further ensure good solubility, hexyl side chains of PDT units for CNO2TPDT are replaced by 2-ethylhexyl groups for RD2TPDT. The chemical structures of CNO2TPDT and RD2TPDT are shown in Scheme 1 and the synthetic routes are shown in Scheme 2. The planarity between PDT and π bridges (with decreased dihedral angles for CNO2TPDT and RD2TPDT) was improved compared with that of BDT molecules in density functional theory (DFT) calculation. The two molecules showed excellent absorption with red-shift of ca. 30 nm compared with BDT analogues. In addition, conspicuous diffraction peaks of CNO2TPDT and RD2TPDT in grazing incidence wide-angle X-ray scattering (GIWAXS) images indicated high crystallinity due to improved ordering. These results indicate that the introduction of a simple planar PDT unit increased the planarity and ordering of the molecular backbone, leading to good absorption ability and high crystallinity.

Scheme 2 Synthetic routes to CNO2TPDT and RD2TPDT.

Results and discussion

Synthesis

The detailed synthetic routes to CNO2TPDT and RD2TPDT are illustrated in ESI.† Two intermediates, ((2,5-dialkoxy-1,4-phenylene)bis(thiophene-5,2-diyl))bis(trimethylstannane) (PDT, compound 1 in Scheme 2)²⁰ and 3,4′-dihexyl-2,2′-bithiophene (Br2TCHO, compound 2 in Scheme 2),²⁴ were easily prepared according to the literature. CHO2TPDT (compound 3 in Scheme 2) was synthesized by using Stille coupling reaction with PDT and Br2TCHO. CNO2TPDT and RD2TPDT were prepared using Knoevenagel condensation with CHO2TPDT and corresponding acceptor units. CNO2TPDT and RD2TPDT both have suitable solubility in common organic solvents, such as chloroform, chlorobenzene, tetrahydrofuran, etc. The thermal stabilities of CNO2TPDT and RD2TPDT were investigated by thermogravimetric analysis (TGA; see Fig. S1 in ESI†) and the decomposition temperatures of CNO2TPDT and RD2TPDT were 355 °C and 394 °C, respectively, which meet the requirement of application in OSCs. The melting points of CNO2TPDT and RD2TPDT were 164 °C and 205 °C, respectively (see Fig. S1 in ESI†).

Theoretical analysis

Optimized geometries and electronic structures were obtained in DFT with B3LYP/6-31G(d,p). Optimized geometries of CNO2TPDT and RD2TPDT showed small torsion angles (15.8° for CNO2TPDT and 12.7° for RD2TPDT, see Scheme 1b) in PDT, which indicates O⋯S weak noncovalent interaction effectively planarized PDT. Moreover, the introduction of PDT enhances planarity and conjugation along the molecular backbone. Although BDT is a rather planar donor itself, the rigidity of BDT induces larger torsion angle between BDT and adjacent conjugation units. After breaking BDT to form PDT, the average dihedral angles between donor and thiophene bridges decreased from 23.5° to 18.2° compared with BDT analogues (detail chemical structures seen Fig. S2 in ESI†). With these advantages, optimized geometries of CNO2TPDT and RD2TPDT indeed showed relatively planar molecular backbones. Electronic structures of CNO2TPDT and RD2TPDT are shown in Fig. 1a. The calculated HOMO energy levels for CNO2TPDT and RD2TPDT are −5.06 and −4.97 eV, respectively. The calculated lowest unoccupied molecular orbital (LUMO) energy levels for CNO2TPDT and RD2TPDT are −2.83 and −2.77 eV, respectively. The calculated energy levels indicated CNO units have stronger electron withdrawing ability than RD units.

Fig. 1 (a) Electronic structures of CNO2TPDT and RD2TPDT. (b) Absorption spectra of CNO2TPDT and RD2TPDT in chloroform solution and as thin films. (c) Cyclic voltammetry of CNO2TPDT and RD2TPDT films in 0.1 mol L⁻¹ Bu₄NPF₆ solution of CH₃CN.

Optical and electrochemical properties

The UV-visible absorption spectra of CNO2TPDT and RD2TPDT in chloroform solution and as thin films are shown in Fig. 1b. CNO2TPDT and RD2TPDT in diluted chloroform solution show similar absorption spectra with the same absorption peak at 513 nm and the same absorption edge at 630 nm. The CNO2TPDT film shows a red-shifted absorption peak at 588 nm while RD2TPDT film shows a broader absorption spectrum with a more red-shifted absorption peak at 614 nm. A vibronic shoulder at 667 nm for RD2TPDT is observed, indicating an effective π–π packing between molecule backbones in the solid state. The optical band gaps calculated from absorption onset of CNO2TPDT and RD2TPDT are estimated to be 1.71 and 1.66 eV, respectively. The absorption of RD2TPDT was red-shifted by 30 nm compared with BDT-based molecules (DR₃TBDT, 713 nm).²³ Moreover, absorption coefficients of CNO2TPDT and RD2TPDT in solution are 8.6 × 10⁴ and 8.2 × 10⁴ M⁻¹ cm⁻¹, respectively. CNO2TPDT and RD2TPDT showed higher absorption coefficients as thin films: 1.03 × 10⁵ and 1.04 × 10⁵ cm⁻¹, respectively. Compared with BDT-based molecules (DR₃TBDT, ε_solution: 8.1 × 10⁴ M⁻¹ cm⁻¹; ε_film: 6.3 × 10⁴ cm⁻¹),²³ CNO2TPDT and RD2TPDT showed higher absorption coefficients both in solution and as films. The detailed data are summarized in Table 1.

Table 1 Optical and electrochemical properties of compounds

Compounds	λ^solutionmax (nm)	ε^solution (M^−1 cm^−1)	λ^filmmax (nm)	ε^film (cm^−1)	λ^filmonset (nm)	E^optg^a (eV)	HOMO^CV (eV)	LUMO^CV (eV)
a Estimated from 1240/λ^filmonset.
CNO2TPDT	513	8.6 × 10^4	588	1.03 × 10^5	720	1.72	−5.03	−3.43
RD2TPDT	513	8.2 × 10^4	614	1.04 × 10^5	745	1.66	−4.98	−3.41

---

The electrochemical properties of CNO2TPDT and RD2TPDT were investigated by cyclic voltammetry (Fig. 2c). HOMO and LUMO levels are calculated from the onset of oxidation potential and reduction potential, respectively. HOMO levels for CNO2TPDT and RD2TPDT are −5.03 and −4.98 eV, respectively. Consistent with the DFT calculation, the HOMO level of CNO2TPDT is slightly lower than that of RD2TPDT, which indicates CNO2TPDT-based devices may achieve higher V_oc. The measured HOMO levels are slightly higher than those of BDT-based molecules (−5.1 to −5.3 eV).^25,26 The corresponding LUMO levels are −3.43 and −3.41 eV. The details of the electrochemical properties are summarized in Table 1.

Fig. 2 (a) J–V curves of CNO2TPDT/PC₇₁BM (1.5 : 1, w/w) and RD2TPDT/PC₇₁BM (1 : 0.8, w/w) with DIO (0.8%, v/v). (b) Corresponding EQE spectra.

Photovoltaic properties

A device architecture of ITO/PEDOT:PSS/donors:PC₇₁BM/Ca/Al was applied to investigate the photovoltaic properties of CNO2TPDT and RD2TPDT. The blends of molecular donors and PC₇₁BM were processed using chloroform as solvent. The device performance was optimized by tuning donor and acceptor (D : A) ratio and solvent addition. The current density–voltage (J–V) curves of optimized devices under optimal conditions at AM 1.5G (100 mW cm⁻²) illumination are shown in Fig. 2a, and the corresponding photovoltaic performances are summarized in Table 2.

Table 2 OSC device performance and hole mobility

Compounds	D : A ratio	Additives	PCE^a (%)	Voc (V)	Jsc (mA cm^−2)	FF (%)	Thickness (nm)	Hole mobility (cm^2 V^−1 s^−1)
a Average performance from 8 devices.
CNO2TPDT	1.5 : 1	No	4.16 (4.07)	0.87	7.09	67.6	68 (±4)	3.36 × 10^−5
RD2TPDT	1 : 1	No	3.35 (3.16)	0.84	9.63	40.7	88 (±4)	4.31 × 10^−5
RD2TPDT	1 : 0.8	0.8% DIO	6.64 (6.59)	0.77	12.74	67.6	90 (±3)	1.18 × 10^−4

---

The device based on CNO2TPDT exhibited a highest PCE of 4.16% with a moderate V_oc of 0.87 V, a J_sc of 7.09 mA cm⁻² and a high FF of 67.6% without any treatment. Thermal annealing and adding solvent additives were used to improve film quality of CNO2TPDT/PC₇₁BM blends but no improvement in PCE was obtained (detailed condition and performance shown in Table S1 in ESI†). In contrast, the device based on RD2TPDT with a D : A weight ratio of 1 : 1 exhibited a PCE of 3.35% with a V_oc of 0.84 V, a J_sc of 9.63 mA cm⁻² and a FF of 40.7% without any treatment. However, the RD2TPDT-based device with D : A weight ratio of 1 : 0.8 and 1,8-diiodooctane (DIO) volume ratio of 0.8% exhibited a highest PCE of 6.64% with a decreased V_oc of 0.77 V, a significantly enhanced J_sc of 12.74 mA cm⁻² and an improved FF of 67.6%. Compared with BDT molecules (V_oc of 0.94 V, J_sc of 12.56 mA cm⁻² and a FF of 70%),²⁷ PDT molecules showed slightly decreased V_oc and maintained high FF values but higher J_sc. The slightly decreased V_oc for the DIO-processed device relative to the device without DIO was probably due to the relative lifting up of the HOMO level caused by improved crystallinity of blends.^28,29

The external quantum efficiency (EQE) of optimized devices based on CNO2TPDT and RD2TPDT is shown in Fig. 2b. The EQE curve of RD2TPDT/PC₇₁BM (1 : 0.8, w/w) with DIO (0.8%, v/v) showed a red-shift edge of 745 nm with the highest EQE value reaching 63% at 500 nm. In comparison, the EQE curve of CNO2TPDT/PC₇₁BM (1 : 1, w/w) exhibited an edge of 720 nm with the highest EQE value of 40.7%. EQE spectra are consistent with the UV absorption spectra. The calculated J_sc agreed with that from the EQE within 5% mismatch compared with the J_sc from J–V measurement.

Morphology

Transmission electron microscopy (TEM) and atomic force microscopy (AFM) were used to investigate the morphology of active layers. TEM images showed the domain size of CNO2TPDT/PC₇₁BM blends (Fig. 3a) was as large as 150 nm with root-mean-square (RMS) value of 1.51 nm calculated from AFM images, leading to less efficient charge separation and thus low J_sc. In contrast, TEM images demonstrated that RD2TPDT/PC₇₁BM blends exhibited uniform and small domain size with RMS value of 0.61 nm calculated from AFM images. However, original RD2TPDT/PC₇₁BM blends showed no obvious phase separation, which was unfavorable for effective charge transport, leading to low J_sc and FF. When DIO was added into the blend, it induced larger phase separation with increased RMS value of 4.72 nm, which benefited the charge transport, resulting in higher J_sc and FF values. The significantly distinct morphology for CNO2TPDT and RD2TPDT may be also due to the increased miscibility of RD2TPDT with PC₇₁BM after shortening the length of terminal units from hexyl groups to ethyl groups.^23,30

Fig. 3 AFM height images (2 × 2 μm)/corresponding AFM phase images (2 × 2 μm)/TEM images of the active layers: (a, d and g) CNO2TPDT/PC₇₁BM (1.5 : 1, w/w); (b, e and h) RD2TPDT/PC₇₁BM (1 : 1, w/w); (c, f and i) RD2TPDT/PC₇₁BM (1 : 0.8, w/w) with DIO (0.8%, v/v).

GIWAXS

GIWAXS analyses (Fig. 4) were used to investigate the relationship among crystallinity, molecular stacking and device performance of CNO2TPDT and RD2TPDT. CNO2TPDT/PC₇₁BM blend films showed an obvious first-order diffraction peak (100, q_z ≈ 0.36 Å⁻¹), second-order diffraction peak (200, q_z ≈ 0.72 Å⁻¹) and third-order diffraction peak (300, q_z ≈ 1.08 Å⁻¹) in the out-of-plane direction, indicating a highly ordered structure and excellent crystallinity in the blends, accounting for high FF values in the CNO2TPDT-based device. In contrast, without DIO, RD2TPDT/PC₇₁BM blend films just showed a visible diffraction peak (100) in the out-of-plane direction, indicating poor crystallinity of RD2TPDT with PC₇₁BM. With DIO, RD2TPDT/PC₇₁BM blend films showed a series of obvious diffraction peaks (100, q_z ≈ 0.37 Å⁻¹; 200, q_z ≈ 0.74 Å⁻¹; 300, q_z ≈ 1.11 Å⁻¹) with larger q_z value in the out-of-plane direction, indicating improved crystallinity and more compact microstructures in the blends, accounting for decreased V_oc and improved FF values in the device. The diffraction peaks of in-plane (010) corresponding to π–π packing indicated two molecules preferentially occupied the edge-on orientation in the blends. The calculated π–π packing (ca. 3.8 Å) of two molecule distances is slightly larger than that of high-performance BDT-based small-molecule donor (ca. 3.7 Å),^4,31 mainly due to the flexibility of PDT units. The relatively low PCE of PDT molecules may be due to the large π–π packing distance. These results indicate the delicate balance of miscibility and crystallinity should be considered in molecular design.

Fig. 4 2D GIWAXS images: (a) CNO2TPDT/PC₇₁BM (1.5 : 1, w/w); (b) RD2TPDT/PC₇₁BM (1 : 1, w/w); (c) RD2TPDT/PC₇₁BM (1 : 0.8, w/w) with DIO (0.8%, v/v). The corresponding (d) in-plane patterns and (e) out-of-plane patterns for 2D GIWAXS.

Mobility

The mobilities of the optimized CNO2TPDT/PC₇₁BM and RD2TPDT/PC₇₁BM blend films were investigated by hole-only space charge limited current (SCLC) measurement with a device architecture of ITO/PEDOT:PSS/active layers/MoO₃/Ag (Fig. S3, see the ESI†). The CNO2TPDT/PC₇₁BM blends showed a hole mobility of 3.36 × 10⁻⁵ cm² V⁻¹ s⁻¹ due to poor morphology. In contrast, RD2TPDT/PC₇₁BM blends without DIO presented a low hole mobility of 4.31 × 10⁻⁵ cm² V⁻¹ s⁻¹. With DIO, RD2TPDT/PC₇₁BM blends showed an improved hole mobility of 1.18 × 10⁻⁴ cm² V⁻¹ s⁻¹, which is comparable with that of a BDT molecule-based device (3.10 × 10⁻⁴ cm² V⁻¹ s⁻¹).²⁷ The enhancement of hole mobility can be ascribed to improved crystallinity in the blends as predicted from GIWAXS images, which is beneficial to exciton separation and charge transport.

Conclusions

In conclusion, two small molecules, CNO2TPDT and RD2TPDT, with simple strongly electron donating PDT units as core were synthesized and characterized. PDT has the advantages of easy accessibility, good solubility and miscibility. Compared with common rigid BDT molecules, PDT molecules ensure better planarity along the molecular backbone, exhibiting excellent light absorption with edges of more than 720 nm and the potential of a high J_sc and high FF. An RD2TPDT-based device showed a highest PCE of 6.64% with a high J_sc of 12.74 mA cm⁻² with DIO as additive due to red-shifted absorption spectra and better morphology. Although these PDT molecules did not show higher performance compared with BDT analogues, with further molecule design, such as fine tuning of side chains, π bridges and conjugation length, followed by fabrication optimization like morphology control, the potential of red-shifted absorption for PDT molecules will be exploited fully, leading to higher J_sc and even higher PCE.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (no. 21125420, 21534004, 91427302, 21474022) and the Chinese Academy of Sciences.

Notes and references

1. F. C. Krebs, M. Jørgensen, K. Norrman, O. Hagemann, J. Alstrup, T. D. Nielsen, J. Fyenbo, K. Larsen and J. Kristensen, Sol. Energy Mater. Sol. Cells, 2009, 93, 422–441.
2. S. Günes, H. Neugebauer and N. S. Sariciftci, Chem. Rev., 2007, 107, 1324–1338.
3. G. Yu, J. Gao, J. C. Hummelen, F. Wudl and A. J. Heeger, Science, 1995, 270, 1789–1791.
4. K. Sun, Z. Xiao, S. Lu, W. Zajaczkowski, W. Pisula, E. Hanssen, J. M. White, R. M. Williamson, J. Subbiah, J. Ouyang, A. B. Holmes, W. W. Wong and D. J. Jones, Nat. Commun., 2015, 6, 6013.
5. B. Kan, M. Li, Q. Zhang, F. Liu, X. Wan, Y. Wang, W. Ni, G. Long, X. Yang and H. Feng, J. Am. Chem. Soc., 2015, 137, 3886–3893.
6. C. Cui, X. Guo, J. Min, B. Guo, X. Cheng, M. Zhang, C. J. Brabec and Y. Li, Adv. Mater., 2015, 27, 7469–7475.
7. Y. Lin, Y. Li and X. Zhan, Chem. Soc. Rev., 2012, 41, 4245–4272.
8. B. Kan, Q. Zhang, M. Li, X. Wan, W. Ni, G. Long, Y. Wang, X. Yang, H. Feng and Y. Chen, J. Am. Chem. Soc., 2014, 136, 15529–15532.
9. D. Deng, Y. J. Zhang, L. Yuan, C. He, K. Lu and Z. X. Wei, Adv. Energy Mater., 2014, 4, 1400538.
10. J. Hou, M.-H. Park, S. Zhang, Y. Yao, L.-M. Chen, J.-H. Li and Y. Yang, Macromolecules, 2008, 41, 6012–6018.
11. W. Ni, M. Li, F. Liu, X. Wan, H. Feng, B. Kan, Q. Zhang, H. Zhang and Y. Chen, Chem. Mater., 2015, 27, 6077–6084.
12. Y. Sun, G. C. Welch, W. L. Leong, C. J. Takacs, G. C. Bazan and A. J. Heeger, Nat. Mater., 2012, 11, 44–48.
13. A. J. Moulé, A. Tsami, T. W. Bünnagel, M. Forster, N. M. Kronenberg, M. Scharber, M. Koppe, M. Morana, C. J. Brabec and K. Meerholz, Chem. Mater., 2008, 20, 4045–4050.
14. D. Deng, Y. Zhang, L. Zhu, J. Zhang, K. Lu and Z. Wei, Phys. Chem. Chem. Phys., 2015, 17, 8894–8900.
15. X. Guo, M. Zhang, W. Ma, L. Ye, S. Zhang, S. Liu, H. Ade, F. Huang and J. Hou, Adv. Mater., 2014, 26, 4043–4049.
16. J. Huang, C. Zhan, X. Zhang, Y. Zhao, Z. Lu, H. Jia, B. Jiang, J. Ye, S. Zhang and A. Tang, ACS Appl. Mater. Interfaces, 2013, 5, 2033–2039.
17. A. Tang, C. Zhan and J. Yao, Chem. Mater., 2015, 27, 4719–4730.
18. L. Ye, S. Zhang, W. Zhao, H. Yao and J. Hou, Chem. Mater., 2014, 26, 3603–3605.
19. T. L. Nguyen, H. Choi, S.-J. Ko, M. A. Uddin, B. Walker, S. Yum, J.-E. Jeong, M. H. Yun, T. Shin and S. Hwang, Energy Environ. Sci., 2014, 7, 3040–3051.
20. J. E. Carlé, J. W. Andreasen, M. Jørgensen and F. C. Krebs, Sol. Energy Mater. Sol. Cells, 2010, 94, 774–780.
21. B. Z. Xia, K. Lu, L. Yuan, J. Q. Zhang, L. Y. Zhu, X. W. Zhu, D. Deng, H. Li and Z. X. Wei, Polym. Chem., 2016, 7, 1323–1329.
22. X. Guo, F. S. Kim, S. A. Jenekhe and M. D. Watson, J. Am. Chem. Soc., 2009, 131, 7206–7207.
23. J. Zhou, X. Wan, Y. Liu, Y. Zuo, Z. Li, G. He, G. Long, W. Ni, C. Li and X. Su, J. Am. Chem. Soc., 2012, 134, 16345–16351.
24. Y. Liu, X. Wan, F. Wang, J. Zhou, G. Long, J. Tian and Y. Chen, Adv. Mater., 2011, 23, 5387–5391.
25. S. Shen, P. Jiang, C. He, J. Zhang, P. Shen, Y. Zhang, Y. Yi, Z. Zhang, Z. Li and Y. Li, Chem. Mater., 2013, 25, 2274–2281.
26. C. Cui, J. Min, C.-L. Ho, T. Ameri, P. Yang, J. Zhao, C. J. Brabec and W.-Y. Wong, Chem. Commun., 2013, 49, 4409–4411.
27. Y. Chen, X. Wan and G. Long, Acc. Chem. Res., 2013, 46, 2645–2655.
28. W. C. Tsoi, S. J. Spencer, L. Yang, A. M. Ballantyne, P. G. Nicholson, A. Turnbull, A. G. Shard, C. E. Murphy, D. D. Bradley and J. Nelson, Macromolecules, 2011, 44, 2944–2952.
29. Z. J. Hu, S. Tang, A. Ahlvers, S. I. Khondaker and A. J. Gesquiere, Appl. Phys. Lett., 2012, 101, 053308.
30. Y. Liang, D. Feng, Y. Wu, S.-T. Tsai, G. Li, C. Ray and L. Yu, J. Am. Chem. Soc., 2009, 131, 7792–7799.
31. Y. Wang, Q. Zhang, F. Liu, X. Wan, B. Kan, H. Feng, X. Yang, T. P. Russell and Y. Chen, Org. Electron., 2016, 28, 263–268.

---

Footnotes

† Electronic supplementary information (ESI) available: Synthesis of CNO2TPDT and RD2TPDT, SCLC test for hole-only devices, MS, ¹H NMR and ¹³C-NMR spectra of CNO2TPDT and RD2TPDT, and a summary of the photovoltaic data of devices on other conditions. See DOI: 10.1039/c6ra09417a

‡ Junjue Zhao and Benzheng Xia contributed equally.

---