A polymer acceptor with an optimal LUMO energy level for all-polymer solar cells

Abstract

A key parameter for polymer electron acceptors is the lowest unoccupied molecular orbital (LUMO) energy level (E_LUMO). For state-of-the-art polymer electron acceptors based on the naphthalene diimide (NDI) unit, their E_LUMO are low-lying and cannot be tuned, leading to a low open-circuit voltage (V_oc) of the resulting all-polymer solar cells (all-PSCs). We report that polymer electron acceptors based on the double B←N bridged bipyridine (BNBP) unit exhibit tunable E_LUMO because of their delocalized LUMOs over polymer backbones. The E_LUMO of the copolymer of the BNBP unit and selenophene unit (P-BNBP-Se) is lower by 0.16 eV than that of the copolymer of the BNBP unit and thiophene unit (P-BNBP-T). As a result, the energy levels of P-BNBP-Se match well with the widely-used polymer donor, poly[(ethylhexyl-thiophenyl)-benzodithiophene-(ethylhexyl)-thienothiophene] (PTB7-Th). The electron mobility of P-BNBP-Se (μ_e = 2.07 × 10⁻⁴ cm² V⁻¹ s⁻¹) is also higher than that of P-BNBP-T (μ_e = 7.16 × 10⁻⁵ cm² V⁻¹ s⁻¹). While the all-PSC device based on the PTB7-Th:P-BNBP-T blend shows a moderate power conversion efficiency (PCE) of 2.27%, the corresponding device with P-BNBP-Se as the acceptor exhibits a PCE as high as 4.26%. Moreover, owing to the suitable E_LUMO of P-BNBP-Se, the all-PSC device of P-BNBP-Se shows a V_oc up to 1.03 V, which is higher by 0.22 V than that with the conventional NDI-based polymer acceptor. These results indicate that BNBP-based polymers can give all-PSCs with high PCEs, remarkably high V_oc values and small photon energy losses.

---

Introduction

All-polymer solar cells (all-PSCs), which utilize polymers as both the electron donor and electron acceptor, have attracted much attention recently because of their great advantages over conventional polymer/fullerene PSCs.¹ These advantages include enhanced light absorption of polymer acceptors, low cost, and improved mechanical/thermal stability. Great progress in all-PSCs has been made by using absorption-complementary polymer donor/acceptors, optimizing the blend morphologies, or developing new polymer acceptors.² However, the further development of all-PSCs is severely limited by the lack of excellent polymer acceptors.³ To date, only several specific polymer acceptors based on the naphthalene diimide (NDI) unit, perylenediimide (PDI) unit and B←N bridged thienylthiazole (BNTT) unit can work as polymer acceptors for efficient all-PSCs with power conversion efficiencies (PCEs) exceeding 4%.^4,5

A key parameter for polymer acceptors is the lowest unoccupied molecular orbital (LUMO) energy level (E_LUMO). In all-PSCs, the E_LUMO difference between the acceptor and donor (ΔE_LUMO) is regarded as the driving force for the charge separation.⁶ The difference between the E_LUMO of the acceptor and the highest occupied molecular orbital (HOMO) energy level (E_HOMO) of the donor is related to the open-circuit voltage (V_oc) of all-PSCs.⁶ Therefore, to get a large ΔE_LUMO for effective charge separation and to maximize V_oc, the E_LUMO of the polymer acceptor must be carefully optimized. The state-of-the-art polymer acceptors are the NDI-based conjugated polymers.⁴ Unfortunately, the E_LUMO of these polymers are fixed at ca. −3.85 eV and cannot be effectively tuned, leading to a low V_oc of the resulting all-PSCs. According to a study by Takimiya et al.,⁷ the fixed E_LUMO of NDI-based polymers are due to the localized LUMOs on the NDI units. The E_LUMO of the NDI-based conjugated polymers are determined by the NDI unit and are not affected by the copolymerization units. Thus, it is important but challenging to develop polymer acceptors with tunable E_LUMO.

Following our strategy to develop polymer acceptors using the B←N unit,⁵ we have reported a new electron-deficient building block based on the B←N unit, double B←N bridged bipyridine (BNBP), to develop a polymer acceptor.⁸ In this manuscript, we report that BNBP-based polymer acceptors show tunable E_LUMO because of their delocalized LUMOs over the polymer backbones. The E_LUMO of the copolymer of the BNBP unit and selenophene unit (P-BNBP-Se) is lower by 0.16 eV than that of the copolymer of the BNBP unit and thiophene unit (P-BNBP-T) (Fig. 1). As a result, the energy levels of P-BNBP-Se match well with the widely-used polymer donor, poly[(ethylhexyl-thiophenyl)-benzodithiophene-(ethylhexyl)-thienothiophene] (PTB7-Th).⁹ While the all-PSC device based on the PTB7-Th:P-BNBP-T blend shows a moderate PCE of 2.27%, the corresponding device with P-BNBP-Se as the acceptor exhibits a PCE as high as 4.26% with a remarkably high V_oc of 1.03 V. These results indicate that BNBP-based polymer acceptors have different electronic structures from those of the classical NDI-based polymer acceptors and that they can give all-PSCs with remarkably high V_oc values and high PCEs.

Results and discussion

Scheme 1 shows the synthetic route of P-BNBP-Se and P-BNBP-T. The three monomers were prepared following literature methods and the two polymers were synthesized in Stille-polymerization conditions.⁸ Their chemical structures are confirmed by ¹H NMR and elemental analysis. According to gel permeation chromatography (GPC), with 1,2,4-trichlorobenzene as the eluent at 150 °C, the number-average molecular weight (M_n) and polydispersity (PDI) are 26.3 kDa and 1.93 for P-BNBP-Se and 46.2 kDa and 1.81 for P-BNBP-T, respectively. According to the thermogravimetric analysis (TGA), P-BNBP-T and P-BNBP-Se show a good thermal stability with thermal decomposition temperatures (T_d) of over 350 °C (ESI†). In addition, the two polymers show a good solubility in common organic solvents, including chlorobenzene (CB), chloroform (CHCl₃) and o-dichlorobenzene (o-DCB).

Scheme 1 Synthetic route of P-BNBP-Se and P-BNBP-T.

To elucidate the molecular orbitals of the two polymers, density functional theory (DFT) calculations at the B3LYP/6-31G* level of theory were performed with the model compounds containing six repeating units with the long alkyl chains replaced by methyl groups.¹⁰ For comparison, we also show the DFT calculation result of the state-of-the-art polymer acceptor, (poly((N,N′-bis(2-octyldodecyl)-1,4,5,8-naphthalenedicarboximide-2,6-diyl)-alt-5,5′-(2,2-bithiophene))) (N2200 or P(NDI2ODT2)) (Fig. 1a).¹¹ As shown in Fig. 2, the calculated LUMO of the model compound of N2200 is localized on the NDI units, indicating that its E_LUMO is determined by the NDI unit and cannot be effectively tuned by changing the co-monomer units. This is consistent with the DFT calculation and experimental results of NDI-based conjugated polymers in the literature.^4,11 In contrast, the calculated LUMOs of the model compounds of P-BNBP-Se/P-BNBP-T are delocalized over the BNBP units and the selenophene/thiophene units. Therefore, the LUMO levels of BNBP-based polymers are determined by both the BNBP unit and the co-monomer unit. The LUMO levels of BNBP-based polymers should be effectively tuned by changing the co-monomer units.

Fig. 1 (a) Chemical structures of P-BNBP-Se, P-BNBP-T, N2200 and PTB7-Th and (b) their LUMO energy level alignments.

Fig. 2 Kohn–Sham LUMOs of model compounds of P-BNBP-Se, P-BNBP-T and N2200, based on calculations at the B3LYP/6-31G* level.

Cyclic voltammetry was employed to estimate the LUMO/HOMO energy levels of the two polymers (ESI†).¹² As shown in Fig. 3a, P-BNBP-Se exhibits irreversible reduction and oxidation waves with onset potentials of E^red_onset = −1.14 V and E^ox_onset = +1.04 V, respectively. Accordingly, the E_LUMO/HOMO of P-BNBP-Se are estimated to be −3.66 eV/−5.84 eV (Table 1). Similarly, the E_LUMO/HOMO of P-BNBP-T are estimated to be −3.50 eV/−5.77 eV (Table 1). As reported previously, the model compound of the BNBP unit itself has an E_LUMO of −3.19 eV. The E_LUMO of the two BNBP-based polymers are much lower than that of the BNBP unit. Moreover, the E_LUMO of P-BNBP-Se is lower than that of P-BNBP-T by 0.16 eV. These results confirm that the LUMO levels of BNBP-based polymers can be effectively tuned by changing the co-monomer units. This is consistent with the delocalized LUMOs in the DFT calculation results. The lower-lying E_LUMO of P-BNBP-Se is attributed to the lower electronegativity of the Se atom (2.4) than the S atom (2.5) and the empty orbital of the Se atom.¹³

Fig. 3 (a) Cyclic voltammogram of P-BNBP-Se and P-BNBP-T in thin films using a Ag/AgCl reference electrode, Fc = ferrocene; (b) UV/Vis absorption spectra of P-BNBP-Se and P-BNBP-T in o-DCB solutions and in thin films.

Table 1 Molecular weights, photophysical and electronic properties, and electron mobilities of P-BNBP-Se and P-BNBP-T

Polymer	M n (kDa)	PDI	λ abs (nm)	λ abs (nm)	ε (cm^−1)	E ^optg (eV)	E ^oxonset (V)	E ^redonset (V)	E HOMO (eV)	E LUMO (eV)	μ e (cm^2 V^−1 s^−1)
a Measured in o-DCB solution. b Measured in thin film. c Onset potential vs. Fc/Fc^+. d E HOMO/LUMO = −(4.80 + E^oxonset/E^redonset) eV.
P-BNBP-Se	26.3	1.93	600	635	1.49 × 10^5	1.87	+1.04	−1.14	−5.84	−3.66	2.07 × 10^−4
P-BNBP-T	46.0	2.01	593	622	1.45 × 10^5	1.92	+0.97	−1.30	−5.77	−3.50	7.16 × 10^−5

---

Fig. 3b shows the absorption spectra of P-BNBP-Se and P-BNBP-T in dilute o-DCB solutions and in thin films. Both of the two polymers in solutions show broad absorption bands around λ = 580 nm. The absorption spectrum is slightly redshifted for P-BNBP-Se compared to P-BNBP-T. In thin film, P-BNBP-Se exhibits a maximum absorption at 635 nm, while P-BNBP-T shows the absorption peak at 622 nm. Both of the two films show high absorption coefficients (ε), suggesting their intense light absorption. According to the onset absorption wavelength in thin films, the optical band gaps (E_g) of P-BNBP-Se and P-BNBP-T are estimated to be 1.87 eV and 1.92 eV, respectively. The electron mobilities (μ_e) of P-BNBP-Se and P-BNBP-T were estimated using the space-charge-limited current (SCLC) method with the current density-voltage curves of the electron-only devices (device structure: ITO/PEIE/polymer/Ca/Al).¹⁴ The electron mobility of P-BNBP-Se (μ_e = 2.07 × 10⁻⁴ cm² V⁻¹ s⁻¹) is higher than that of P-BNBP-T (μ_e = 7.16 × 10⁻⁵ cm² V⁻¹ s⁻¹) (ESI†). The higher electron mobility of P-BNBP-Se is due to the stronger intermolecular interactions in Se-containing polymers because of the larger and more polarizable radii of the selenium atom than the sulfur atom. This is confirmed by the smaller π–π stacking distance of P-BNBP-Se (d_π–π = 3.77 Å) than that of P-BNBP-T (d_π–π = 3.81 Å) (ESI†). The electron mobility of P-BNBP-Se is comparable to the hole mobilities of typical polymer electron donors, which is very favourable for its application as a polymer electron acceptor in all-PSCs.

To investigate the application of P-BNBP-Se and P-BNBP-T as electron acceptors in all-PSCs, we select a widely-used polymer donor, PTB7-Th. All-PSC devices were fabricated with a configuration of ITO/PEDOT:PSS/PTB7-Th:P-BNBP-Se or P-BNBP-T/Ca/Al (ESI†). The active layer was spin-coated from the blend in o-DCB solution without any additives. Fig. 4 shows the current density–voltage (J–V) curves under AM 1.5G illumination (100 mW cm⁻²) and the external quantum efficiency (EQE) spectra of the optimal devices. The photovoltaic parameters are summarized in Table 2. The PTB7-Th : P-BNBP-T (3 : 1, w:w) device shows a PCE of 2.27% with a V_oc of 1.12 V, a short-circuit current density (J_sc) of 5.24 mA cm⁻² and a fill factor (FF) of 0.39. The device based on the PTB7-Th : P-BNBP-Se (2 : 1, w:w) blend exhibits a PCE of 4.26% with a V_oc of 1.03 V, a J_sc of 10.02 mA cm⁻² and a FF of 0.42. This PCE value is comparable to that of the reference all-PSC device based on the PTB7-Th : N2200 (1 : 1, w:w) blend from the chloroform solution (PCE = 4.57%), indicating that P-BNBP-Se is an excellent polymer acceptor. Compared with the device of P-BNBP-T, the device of P-BNBP-Se shows a slightly decreased V_oc and much increased J_sc. The slightly decreased V_oc is attributed to the lower E_LUMO of P-BNBP-Se than that of P-BNBP-T. On the other hand, the V_oc of the P-BNBP-Se device is higher than that of the N2200 device by 0.22 V (Table 2) because the E_LUMO of P-BNBP-Se is higher than that of N2200. The much increased J_sc of the P-BNBP-Se device than that of the P-BNBP-T device is in accordance with their EQE values (EQE_max = 0.47 for P-BNBP-Se and EQE_max = 0.25 for P-BNBP-T) (Fig. 4b). The J_sc calculated from the integration of the EQE spectra agrees well with the J_sc values obtained from the J–V scans within an error of 5%.

Fig. 4 (a) J–V curves and (b) EQE spectra of the all-PSC devices based on the PTB7-Th:P-BNBP-Se, PTB7-Th:P-BNBP-T and PTB7-Th:N2200 blends, respectively.

Table 2 Summary of the all-PSC device performance

Acceptor	V oc (V)	J sc (mA cm^−2)	FF	PCEmax/ave^a (%)	EQE	E loss (eV)
a The average PCE value is calculated from eight devices.
P-BNBP-Se	1.03	10.02	0.42	4.26/4.11	0.47	0.56
P-BNBP-T	1.12	5.24	0.39	2.27/2.08	0.25	0.47
N2200	0.81	10.55	0.53	4.57/4.30	0.50	0.67

---

The charge carrier mobilities of the two blends were investigated using the SCLC method with the electron-only and hole-only devices (ESI†).¹⁴ The electron mobility and hole mobility (μ_h) of the PTB7-Th:P-BNBP-Se blend are estimated to be 3.34 × 10⁻⁵ cm² V⁻¹ s⁻¹ and 2.38 × 10⁻⁴ cm² V⁻¹ s⁻¹, respectively. In comparison, the PTB7-Th:P-BNBP-T blend exhibits a μ_e = 5.96 × 10⁻⁶ cm² V⁻¹ s⁻¹ and μ_h = 7.28 × 10⁻⁴ cm² V⁻¹ s⁻¹, respectively. The higher electron mobility and the balanced electron/hole mobilities of the PTB7-Th:P-BNBP-Se blend are due to the enhanced electron mobility of P-BNBP-Se. We also investigated the bimolecular charge recombination in the all-PSC devices using the light-intensity dependence of the J–V curves (Fig. 5). The J_sc follows a power-law dependence on the illumination intensity (J_sc ∝ P_light^α), where P_light is light intensity and α is the calculated power-law exponent. The α values are 0.93 for the PTB7-Th:P-BNBP-Se device and 0.94 for the PTB7-Th:P-BNBP-T device, which are close to unity, suggesting that the bimolecular charge recombination is weak in the two devices at a short circuit condition.¹⁵ Both the weak bimolecular recombination and the high and balanced electron/hole mobilities of PTB7-Th:P-BNBP-Se can explain its excellent device performance.

Fig. 5 Short-circuit current density (J_sc) versus light intensity (P_light) data and power-law (J_sc ∝ P_light^α) fittings for the all-PSC devices.

The morphologies of the PTB7-Th:P-BNBP-Se and PTB7-Th:P-BNBP-T blends were characterized by transmission electron microscopy (TEM) and atomic force microscopy (AFM). As shown in Fig. 6, the TEM images exhibit similar nano/microstructures without large-size aggregation. The AFM images of the two blends similarly reveal smooth surface morphologies with the same root-mean-square (RMS) roughness of 1.47 nm and domain sizes of around 20–40 nm. The phase separation morphologies of the two blends are beneficial for good all-PSC devices.

Fig. 6 The TEM images and the AFM height images of ((a) and (b)) the PTB7-Th:P-BNBP-Se blend and ((c) and (d)) the PTB7-Th:P-BNBP-T blend, respectively.

In organic photovoltaics (OPVs), the ΔE_LUMO of the donor and acceptor is regarded as the driving force for charge separation. The ΔE_LUMO should be larger than a specific value for efficient charge separation. If ΔE_LUMO is too large, there is a lot of energy loss in the charge separation process, leading to a low V_oc because the V_oc of OPVs is related to the difference between the E_HOMO of the donor and E_LUMO of the acceptor.⁶ In our previous report, an all-PSC device based on the PTB7:P-BNBP-T blend (ΔE_LUMO = 0.19 eV) showed a good PCE of 3.38%.⁸ As shown in Fig. 1b, the ΔE_LUMO is only 0.06 eV for PTB7-Th:P-BNBP-T, and thus the all-PSC device shows a high V_oc but produces a low PCE due to the insufficient charge separation.^4g As the E_LUMO of P-BNBP-Se is lower than that of P-BNBP-T, the ΔE_LUMO for PTB7-Th:P-BNBP-Se is increased to 0.22 eV and ensures an efficient charge separation, resulting in higher J_sc and PCE values. Moreover, due to the suitable E_LUMO of P-BNBP-Se, the PTB7-Th:P-BNBP-Se device produces a high V_oc of 1.03 V, which is higher than that of the PTB7-Th:N2200 device by 0.22 V. These results indicate that the suitable E_LUMO of P-BNBP-Se plays an important role in enhancing the all-PSCs device performance.

It is worthy to note the remarkably low photon energy losses (E_loss) of the all-PSCs based on P-BNBP-Se and P-BNBP-T. E_loss is defined as the difference between the lowest optical bandgap of the donor/acceptor and the eV_oc of the organic photovoltaic (OPV) device (E_loss = E_g − eV_oc).¹⁶ Typically, OPVs have large E_loss values of 0.7–1.0 eV. It has been proposed that the lowest attainable E_loss of OPVs is 0.6 eV, despite several exceptional examples.¹⁷ As listed in Table 2, the E_loss for the device of PTB7-Th:P-BNBP-Se and PTB7-Th:P-BNBP-T is 0.56 eV and 0.47 eV, respectively. To our best knowledge, the E_loss of 0.47 eV is the lowest one for OPVs reported so far. A small E_loss is always observed for all-PSCs with BNBP-based polymers as electron acceptors and the exact reason is as yet unknown. We speculate that the small E_loss is related to the high-lying LUMO levels of the BNBP-based polymers.

Conclusions

In summary, we have developed a polymer acceptor based on the BNBP unit and selenophene unit with an optimal E_LUMO to simultaneously enable charge separation and maximize V_oc. BNBP-based polymers have delocalized LUMOs over the polymer backbones, so their E_LUMO can be tuned by changing the comonomer unit. The E_LUMO of P-BNBP-Se is lower by 0.16 eV than that of P-BNBP-T and consequently matches well with that of the polymer donor, PTB7-Th. While the all-PSC device based on PTB7-Th:P-BNBP-T shows a moderate PCE of 2.27%, the corresponding device with P-BNBP-Se as the acceptor exhibits a PCE as high as 4.26%. Moreover, the device of P-BNBP-Se shows a V_oc of up to 1.03 V and E_loss as small as 0.56 eV. These results indicate that BNBP-based polymer acceptors have different electronic structures from those of classical NDI-based polymer acceptors and that they can give all-PSCs with remarkably high V_oc values and high PCEs.

Experimental section

Synthesis of P-BNBP-Se

The dibromo-substituted BNBP monomer was synthesized according to the previous report.⁸ The starting materials of the dibromo-substituted BNBP monomer (220 mg, 0.248 mmol), 2,5-bis(trimethylstannyl)selenophene (114.1 mg, 0.248 mmol), Pd₂(dba)₃·CHCl₃ (4.5 mg, 0.0050 mmol) and P(o-tolyl) (12.1 mg, 0.04 mmol) were placed in a two-necked flask under argon, and then dried toluene (11 mL) was added. After the mixture was stirred at 115 °C for 48 h, an end-capping reaction was carried out by adding 2,5-bis(trimethylstannyl)selenophene (3 mg) and then bromobenzene (200 mg). After cooling, the resulting organic phase was extracted with CHCl₃ (150 mL) and washed with water. After the solvents were removed, the residue was dispersed in methanol and the precipitate was collected. The obtained dark solid was dispersed in acetonitrile, and was collected and dried in a vacuum overnight. Yield: 213 mg (95%). ¹H NMR (400 MHz, CDCl₃, 25 °C): δ 8.41 (s, 1H), 7.76 (s, 1H), 7.66 (s, 1H), 3.63 (br, 2H), 1.85 (br, 1H), 1.42–1.27 (br, 24H), 0.85 (br, 6H). GPC (TCB, polystyrene standard, 150 °C): M_n = 26 300, PDI = 1.93. Anal. calcd for C₄₆H₇₂B₂F₄N₄Se: C, 64.42; H, 8.46; B, 2.52; F, 8.86; N, 6.53; Se, 9.21. Found: C, 64.25; H, 8.58; N, 6.65; Se, 9.05.

Synthesis of P-BNBP-T

The starting materials of the dibromo-substituted BNBP monomer (150 mg, 0.166 mmol), 2,5-bis(trimethylstannyl)thiophene (68.5 mg, 0.166 mmol), Pd₂(dba)₃·CHCl₃ (3.5 mg, 0.0033 mmol) and P(o-tolyl) (8.1 mg, 0.027 mmol) were placed in a two-necked flask under argon, and then dried toluene (4 mL) was added. After the mixture was stirred at 115 °C for 50 h, an end-capping reaction was carried out by adding 2,5-bis(trimethylstannyl)thiophene (3 mg) and then bromobenzene (200 mg). After cooling, the resulting organic phase was extracted with CHCl₃ (150 mL) and washed with water. After the solvents were removed, the residue was dispersed in methanol and the precipitate was collected. The obtained dark solid was dispersed in acetonitrile, and was collected and dried in a vacuum overnight. Yield: 116.5 mg (86%). ¹H NMR (400 MHz, CDCl₃, 25 °C): δ 8.47 (s, 1H), 7.74 (s, 1H), 7.60 (s, 1H), 3.64 (br, 2H), 1.87 (br, 1H), 1.43–1.25 (br, 24H), 0.83 (br, 6H). GPC (TCB, polystyrene standard, 150 °C): M_n = 46 200, PDI = 1.81. Anal. calcd for C₄₆H₇₂B₂F₄N₄S: C, 68.14; H, 8.95; B, 2.67; F, 7.37; N, 6.91; S, 3.95. Found: C, 67.83; H, 8.82; N, 6.75; S, 4.05.

Acknowledgements

The authors are grateful for the financial support by the 973 Project (No. 2014CB643504), the Nature Science Foundation of China (No. 51373165, No. 21404099, No. 21574129), the “Thousand Talents Program” of China, the Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDB12010200), and the State Key Laboratory of Supramolecular Structure and Materials in Jilin University (No. sklssm201608).

Notes and references

1. (a) T. Kim, J.-H. Kim, T. E. Kang, C. Lee, H. Kang, M. Shin, C. Wang, B. Ma, U. Jeong, T.-S. Kim and B. J. Kim, Nat. Commun., 2015, 6, 8547; (b) A. Facchetti, Mater. Today, 2013, 16, 123; (c) G. Yu, J. Gao, J. C. Hummelen, F. Wudl and A. J. Heeger, Science, 1995, 270, 1789.
2. (a) L. Gao, Z.-G. Zhang, L. Xue, J. Min, J. Zhang, Z. Wei and Y. Li, Adv. Mater., 2016, 28, 1884; (b) Y.-J. Hwang, B. A. E. Courtright, A. S. Ferreira, S. H. Tolbert and S. A. Jenekhe, Adv. Mater., 2015, 27, 4578; (c) Y.-J. Hwang, T. Earmme, B. A. E. Courtright, F. N. Eberle and S. A. Jenekhe, J. Am. Chem. Soc., 2015, 137, 4424; (d) T. Earmme, Y.-J. Hwang, S. Subramaniyan and S. A. Jenekhe, Adv. Mater., 2014, 26, 6080; (e) T. Earmme, Y.-J. Hwang, N. M. Murari, S. Subramaniyan and S. A. Jenekhe, J. Am. Chem. Soc., 2013, 135, 14960.
3. (a) H. Benten, D. Mori, H. Ohkita and S. Ito, J. Mater. Chem. A, 2016, 4, 5340; (b) L. Lu, T. Zheng, Q. Wu, A. M. Schneider, D. Zhao and L. Yu, Chem. Rev., 2015, 115, 12666.
4. (a) N. Zhou, A. S. Dudnik, T. I. N. G. Li, E. F. Manley, T. J. Aldrich, P. Guo, H.-C. Liao, Z. Chen, L. X. Chen, R. P. H. Chang, A. Facchetti, M. O. de la Cruz and T. J. Marks, J. Am. Chem. Soc., 2016, 138, 1240; (b) C. Lee, H. Kang, W. Lee, T. Kim, K.-H. Kim, H. Y. Woo, C. Wang and B. J. Kim, Adv. Mater., 2015, 27, 2466; (c) H. Kang, M. A. Uddin, C. Lee, K.-H. Kim, T. L. Nguyen, W. Lee, Y. X. Li, C. Wang, H. Y. Woo and B. J. Kim, J. Am. Chem. Soc., 2015, 137, 2359; (d) J. W. Jung, J. W. Jo, C.-C. Chueh, F. Liu, W. H. Jo, T. P. Russell and A. K.-Y. Jen, Adv. Mater., 2015, 27, 3310; (e) L. Ye, X. Jiao, M. Zhou, S. Zhang, H. Yao, W. Zhao, A. Xia, H. Ade and J. Hou, Adv. Mater., 2015, 27, 6046; (f) P. Cheng, L. Ye, X. Zhao, J. Hou, Y. Li and X. Zhan, Energy Environ. Sci., 2014, 7, 1351; (g) Y. Zhou, T. Kurosawa, W. Ma, Y. Guo, L. Fang, K. Vandewal, Y. Diao, C. Wang, Q. Yan, J. Reinspach, J. Mei, A. L. Appleton, G. I. Koleilat, Y. Gao, S. C. B. Mannsfeld, A. Salleo, H. Ade, D. Zhao and Z. Bao, Adv. Mater., 2014, 26, 3767; (h) I. H. Jung, W.-Y. Lo, J. Jang, W. Chen, D. L. Zhao, E. S. Landry, L. Y. Lu, D. V. Talapin and L. Yu, Chem. Mater., 2014, 26, 3450; (i) C. Mu, P. Liu, W. Ma, K. Jiang, J. B. Zhao, K. Zhang, Z. H. Chen, Z. H. Wei, Y. Yi, J. N. Wang, S. H. Yang, F. Huang, A. Faccheti, H. Ade and H. Yan, Adv. Mater., 2014, 26, 7224; (j) D. Mori, H. Benten, I. Okada, H. Ohkita and S. Ito, Energy Environ. Sci., 2014, 7, 2939; (k) M. Schubert, B. A. Collins, H. Mangold, I. A. Howard, W. Schindler, K. Vandewal, S. Roland, J. Behrends, F. Kraffert, R. Steyrleuthner, Z. Chen, K. Fostiropoulos, R. Bittl, A. Salleo, A. Facchetti, F. Laquai, H. W. Ade and D. Neher, Adv. Funct. Mater., 2014, 24, 4068; (l) E. J. Zhou, J. Z. Cong, K. Hashimoto and K. Tajima, Adv. Mater., 2013, 25, 6991; (m) L. Xue, Y. Yang, Z.-G. Zhang, X. Dong, L. Gao, H. Bin, J. Zhang, Y. Yang and Y. Li, J. Mater. Chem. A, 2016, 4, 5810.
5. (a) R. Zhao, C. Dou, Z. Xie, J. Liu and L. Wang, Angew. Chem., Int. Ed., 2016, 55, 5313; (b) C. Dou, Z. Ding, Z. Zhang, Z. Xie, J. Liu and L. Wang, Angew. Chem., Int. Ed., 2015, 54, 3648.
6. R. A. Street, Adv. Mater., 2016, 28, 3814.
7. K. Takimiya, I. Osaka and M. Nakano, Chem. Mater., 2014, 26, 587.
8. (a) C. Dou, X. Long, Z. Ding, Z. Xie, J. Liu and L. Wang, Angew. Chem., Int. Ed., 2016, 55, 1436; (b) X. Long, Z. Ding, C. Dou, J. Zhang, J. Liu and L. Wang, Adv. Mater., 2016 DOI:10.1002/adam.201601205.
9. Z. He, C. Zhong, S. Su, M. Xu, H. Wu and Y. Cao, Nat. Photonics, 2012, 6, 591.
10. DFT calculations were performed using Gaussian 09 program: M. J. Frisch, et al., Gaussian 09, revision A.02, Gaussian, Inc., Wallingford, CT, 2009. For details, see ESI.†.
11. R. Steyrleuthner, M. Schubert, I. Howard, B. Klaumünzer, K. Schilling, Z. Chen, P. Saalfrank, F. Laquai, A. Facchetti and D. Neher, J. Am. Chem. Soc., 2012, 134, 18303.
12. Z. Ding, Z. Miao, Z. Xie and J. Liu, J. Mater. Chem. A, 2016, 4, 2413.
13. (a) D. Meng, D. Sun, C. Zhong, T. Liu, B. Fan, L. Huo, Y. Li, W. Jiang, H. Choi, T. Kim, J. Y. Kim, Y. Sun, Z. Wang and A. J. Heeger, J. Am. Chem. Soc., 2016, 138, 375; (b) R. S. Ashraf, I. Meager, M. Nikolka, M. Kirkus, M. Planells, B. C. Schroeder, S. Holliday, M. Hurhangee, C. B. Nielsen, H. Sirringhaus and I. McCulloch, J. Am. Chem. Soc., 2015, 137, 1314.
14. P. W. M. Blom, M. J. M. de Jong and M. G. van Munster, Phys. Rev. B: Solid State, 1997, 55, R656.
15. (a) A. K. K. Kywa, D. H. Wang, V. Gupta, W. L. Leong, L. Ke, G. C. Bazan and A. J. Heeger, ACS Nano, 2013, 7, 4569; (b) S. R. Cowan, A. Roy and A. J. Heeger, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 245207.
16. (a) R. A. J. Janssen and J. Nelson, Adv. Mater., 2013, 25, 1847; (b) D. Veldman, S. C. J. Meskers and R. A. J. Janssen, Adv. Funct. Mater., 2009, 19, 1939.
17. (a) K. Kawashima, Y. Tamai, H. Ohkita, I. Osaka and K. Takimiya, Nat. Commun., 2015, 6, 10085; (b) W. Li, K. H. Hendriks, A. Furlan, M. M. Wienk and R. A. J. Janssen, J. Am. Chem. Soc., 2015, 137, 2231; (c) M. Wang, H. Wang, T. Yokoyama, X. Liu, Y. Huang, Y. Zhang, T.-Q. Nguyen, S. Aramaki and G. C. Bazan, J. Am. Chem. Soc., 2014, 136, 12576.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental details, thermal property, theoretical calculations, as well as all-PSC device fabrications and characterizations. See DOI: 10.1039/c6sc01756h

‡ Z. Ding and X. Long contributed equally to this work.

---