Design strategies of n-type conjugated polymers for organic thin-film transistors

Abstract

Significant progress in n-type conjugated polymers (CPs) for organic thin-film transistors (OTFTs), which are required in making complementary metal oxide semiconductor (CMOS)-like logic circuits together with p-type CPs, has been made in the last two decades by developing novel building blocks and optimizing device structures as well as processing conditions. However, the device performance, in terms of mobility and air stability, of n-type CPs is still much lower than that of their p-type counterparts. CPs play a key role in determining the device performance. The properties of CPs, such as molecular packing structures and frontier molecular orbital energy levels, can be appropriately adjusted by molecular design. In the current review, we summarize the progress of n-type CPs from the aspect of molecular design, and four molecular design strategies, i.e., donor–acceptor (D–A) CPs with strong A units, D–A CPs with weak D units, D–A CPs with dual-A and A–A type CPs are discussed. It has been demonstrated that D–A CPs with weak D units and A–A type CPs are highly desirable in the construction of unipolar n-type CPs because they generally have low-lying highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels.

---

Ying Sui

Ying Sui received her BS degree in Materials Science and Engineering from Tianjin University in 2015. Now she is a PhD candidate at the School of Materials Science and Engineering, Tianjin University. She mainly researches on the synthesis of conjugated polymers by direct arylation polycondensatioin.

Yunfeng Deng

Yunfeng Deng received his PhD degree in Polymer Chemistry and Physics from Changchun Institute of Applied Chemistry, Chinese Academy Of Sciences in 2015. He then joined the University of Waterloo and Technion-Israel Institute of Technology as a postdoctoral research fellow. Currently, he is an associate professor in the School of Materials Science and Engineering, Tianjin University, China. His research mainly focuses on the design and synthesis of conjugated polymers and their related electronic devices.

Tian Du

Tian Du received his BS degree in Material Chemistry from Zhengzhou University in 2017. He then moved to Tianjin University to pursue his PhD degree. His research interests include molecular design and synthesis of π-conjugated materials and their applications in thin-film transistors.

Yibo Shi

Yibo Shi received his BS degree in Materials Science and Engineering from Tianjin University in 2018. He is currently pursuing his PhD degree at the School of Materials Science and Engineering, Tianjin University. His research interests mainly focus on organic semiconducting polymers.

Yanhou Geng

Yanhou Geng received his BS degree from Shanghai Jiaotong University in 1991 and PhD degree from Changchun Institute of Applied Chemistry (CIAC), Chinese Academy of Sciences (CAS) in 1996. From 1998 to 2003, he worked in the Max-Planck Institute for Polymer Research in Mainz (Germany) as an AvH stipendiat and the University of Rochester (USA) as a postdoctoral researcher. From 2003 to 2015, he was a professor in the State Key Laboratory of Polymer Physics and Chemistry, CIAC. Currently, he is a full professor at the School of Materials Science and Engineering, Tianjin University, China. He has published over 200 papers in peer-reviewed journals and more than 20 patents. His research interests include (1) polymerization methods for conjugated polymers; (2) organic and polymeric semiconductors for solar cells and thin-film transistors.

---

Introduction

Organic thin-film transistors (OTFTs) based on conjugated polymers (CPs) have attracted considerable attention due to their applications in low-cost, large-area and flexible electronics, such as flexible displays and wearable devices.^1–12 Polymer semiconductors are the key component of OTFTs and the recent advancement of CPs promotes the application of these devices.^6–9,12 According to the polarity of the dominant charge carrier in the transistor channel, CPs can be classified as p-type (hole dominant), n-type (electron dominant) and ambipolar (both carriers) semiconductors. In the past several years, a significant amount of high-performance p-type and ambipolar CPs have been reported. In contrast, there has been a paucity of n-type CPs.^13–21 High-performance n-type CPs are crucial to fabricate low-power complementary metal oxide semiconductor (CMOS)-like logic circuits, which is the most important and widely used configuration in integrated circuits.^21–24 Therefore, the scarcity of n-type CPs is one of the most critical issues to be addressed to realize the application of OTFTs in electronic devices.

For n-type CPs, both low-lying highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels are necessary. Empirically, a HOMO energy level of <−5.9 eV and a LUMO energy level of ∼−4.0 eV are required for a stable n-type organic semiconductor.^14,25–30 The low-lying HOMO energy level could build a larger hole injection barrier to block hole injection from the commonly used Au electrodes, realizing unipolar n-type transport. Meanwhile, the low-lying LUMO energy level would facilitate electron injection. Although low work function electrodes, such as Al, Ca and Mg, have been proved to help electron injection, these metals are not environmentally stable.^17,21 More importantly, the low-lying LUMO energy level is beneficial for stabilizing electron transport under air conditions.^18,21,25,26 It has been revealed that the air instability of n-type OTFT devices is not due to the degradation of materials, but arises from the charge carrier trapping by water and oxygen.^17,25 The deep LUMO energy level could prevent the electron from being captured by water or oxygen in air and improve the air stability of the n-type OTFT device. However, the LUMO energy level cannot be far lower than −4.0 eV, because the polymer will be unintentionally “air-doped” if the LUMO energy level is too low (≤−4.5 eV), leading to the loss of its semiconductor properties.²⁰ Therefore, for n-type CPs, a LUMO energy level between −4.0 and −4.4 eV would be desirable for air-stable electron transport. Additionally, HOMO and LUMO distribution can also influence the polarity of charge carriers of CPs. Generally, polymers with delocalized LUMOs and segregated HOMOs tend to display n-type transport characteristics. The main transport pathway of the charge carriers in an organic semiconductor is hopping through the frontier molecular orbitals. Well distributed LUMOs will facilitate LUMO–LUMO interactions, thus electron can easily move from one polymer chain to another, while localized HOMOs may result in less efficient HOMO–HOMO interactions, blocking hole hopping.^31–34

Donor–acceptor (D–A) polymers, in which the D unit and A unit are alternately linked along the polymer backbone, are promising candidates for high mobility polymers due to their strong intermolecular interactions and small π–π stacking distance enabled by the large dipole moment between the D and A units. However, the electron-rich D unit, such as thiophene and bithiophene, would raise the HOMO energy levels of the resultant D–A polymers through orbital hybridization between the D and A units.^35–37 The elevated HOMO energy level would reduce the hole injection barriers. As a result, the majority of D–A polymers tends to exhibit ambipolar or merely p-type charge transport characteristics (depending on the LUMO energy level). Moreover, for D–A polymers, the HOMOs are usually well delocalized along the polymer chains while the LUMOs are mostly localized on the A units,^38–43 which is detrimental to both intrachain and interchain electron transport. Therefore, to achieve high-performance n-type CPs, reasonable design of D–A polymers is required or new polymer backbone design strategies should be developed.

Besides the intrinsic properties of the semiconductor polymers, the OTFT device architecture also exerts an influence on the observed device performance. There are three commonly-used OTFT device architectures, namely, bottom-gate bottom-contact (BGBC), bottom-gate top-contact (BGTC) and top-gate bottom-contact (TGBC). The fabrication of bottom-gate devices is relatively easy since the substrate Si/SiO₂ wafer also serves as a gate electrode and a dielectric layer.⁵⁰ However, TGBC devices frequently exhibit higher electron mobility and better air stability than those with the bottom-gate structure due to the better contact of the semiconductor layer with the organic dielectric and the encapsulation of the dielectric layer, preventing the capture of electron by moisture and oxygen.^21,50

Since the first demonstration of a high-performance n-type OTFT device based on CP in the early 2000s by using a ladder-type polymer, poly[(7-oxo-7,10H-benz[de]imidazo[4′,5′:5,6]benzimidazo[2,1-a]isoquinoline-3,4:10,11-tetrayl)-10-carbonyl] (BBL), as the semiconductor layer,⁴⁴ remarkable progress in the development of n-type CPs has been made via the design and synthesis of novel polymers with appropriate HOMO and LUMO energy levels. Recently, a D–A polymer with unipolar electron mobility (μ_e) of up to 7.16 cm² V⁻¹ s⁻¹ has been reported,⁴⁵ noticeably shortening the performance gap between p- and n-type CPs. In this article, we review the progress of n-type CPs with an emphasis on the polymer design strategies. Four molecular design strategies of n-type CPs, i.e., D–A CPs with strong A units, D–A CPs with weak D units, D–A CPs with dual-A and A–A type CPs, are introduced. The relative merits of these strategies and the future development trend of n-type CPs for OTFTs are also discussed.

Design strategies of n-type CPs

1. D–A CPs with strong A units

The development of n-type CPs is somewhat restrained by the lack of strong A units that can endow D–A CPs with low-lying HOMO and LUMO energy levels. Currently, the useful A units for n-type CPs are mainly limited to naphthalene diimide (NDI), perylene diimide (PDI), bis(oxoindolinylidene)benzodifurandione and quinoidal compounds.

1.1 D–A CPs based on NDI and PDI. NDI exhibits high electron affinity due to the two sets of imide groups. The first solution-processable NDI-based CP was developed by Watson et al. and Facchetti et al. independently in 2008.^46,47 Watson et al. reported a series of NDI-based CPs with different thiophene units.⁴⁶ They found that the LUMO energy levels of these polymers were invariant and dictated by the NDI unit, while the HOMO energy levels are a function of the thiophene D unit. Facchetti et al. firstly reported the OTFT performance of an NDI-based polymer, P(NDI2OD-T) (P1, also known as N2200), which showed n-type performance with a μ_e of 0.06 cm² V⁻¹ s⁻¹ in BGTC OTFT devices.⁴⁷ Due to the low-lying LUMO energy level of P1, OTFT device performance exhibited good air stability. When measured under air conditions, P1-based BGTC OTFT devices operated for 14 weeks with a μ_e of 0.01 cm² V⁻¹ s⁻¹. Later, the same group found that a μ_e of up to 0.85 cm² V⁻¹ s⁻¹ under ambient conditions could be achieved for P1 in TGBC OTFT devices.⁴⁸ By using an off-centre spin-coating technique, the μ_e of P1-based TGBC OTFT devices was further improved to 1.03 cm² V⁻¹ s⁻¹.⁴⁹ Now, P1 is one of the representative n-type CPs and several strategies have been adopted to improve the OTFT device performance of this polymer (Fig. 1 and Table 1).⁵⁰

Fig. 1 Molecular structures of the n-type D–A conjugated polymers based on NDI and PDI.

Table 1 Molecular weights, energy levels and device properties of the n-type D–A conjugated polymers based on NDI and PDI

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d Estimated from EHOMO = ELUMO − E^optg, in which E^optg is the optical bandgap. e Estimated from ELUMO = EHOMO + E^optg. f Measured by photoelectron spectroscopy. g The electrodes were modified with Cs2CO3. h The electrodes were modified with Ba(OH)2.
P1	47.8	5.53	−5.36^d	−3.91	BGTC	0.06	Vacuum	47
	52.5	5.51	n/a	n/a	TGBC	0.85	Air	48
	n/a	n/a	n/a	n/a	TGBC	1.03	Nitrogen	49
P2	32.0	2.02	−5.27^d	−3.83	TGBC	1.04	n/a	51
P3	28.0	2.03	−5.62^d	−4.01	BGTC	6.50	Nitrogen	52
P4	42.4	2.60	−5.95	−3.94	BGTC	0.07	Nitrogen	54
P5	70.0	1.98	−5.42^d	−4.00	TGBC	1.80^g	Nitrogen	57
P6	80.0	2.10	−5.29^d	−3.98	TGBC	0.50^g	Nitrogen	58
P7	117	2.90	−5.55^f	−4.05^e	BGTC	0.07^h	Nitrogen	63
P8	49.0	2.90	n/a	n/a	BGTC	3.7 × 10^−3	Nitrogen	64
P9	62.5	3.30	−5.45^d	−3.88	BGTC	7.0 × 10^−3	Nitrogen	69
P10	21.3	2.90	−5.31^d	−3.91	BGTC	4.8 × 10^−3	Nitrogen	69
P11	92.4	3.60	−5.29^d	−3.92	BGTC	0.012	Nitrogen	69
P12	33.2	7.00	−5.82	−4.25	BGBC	0.38	Nitrogen	70
P13	15.0	1.50	−5.90	−3.90	BGTC	0.013	Nitrogen	71
P14	15.0	1.10	−5.80	−4.00	BGTC	0.06	Air	73
P15	11.0	2.90	−5.61^d	−3.96	BGTC	2.0 × 10^−3	Vacuum	47
P16	9.97	1.53	−5.85	−3.75	BGTC	0.05	Nitrogen	74
P17	11.1	1.90	−5.71^d	−3.85	BGTC	0.042	Nitrogen	75

---

Yang et al. synthesized an NDI-based polymer P2 by incorporating hybrid siloxane chains.⁵¹ The pre-aggregation of P2 in solution was highly sensitive to the solvent; thus the degrees and types of the aggregates in thin-films could be tuned by the solvent, leading to solvent-dependent mobility. With chloroform as the casting solvent, P2-based TGBC OTFT devices showed the best performance with a μ_e of 1.04 cm² V⁻¹ s⁻¹, which was much higher than that of P1 (0.32 cm² V⁻¹ s⁻¹) when the devices were fabricated under the same conditions. The high μ_e of P2 could be attributed to its balanced face-on and edge-on molecular orientations and strong π–π interactions as evidenced by the X-ray diffraction (XRD) analysis. An NDI-based polymer with semifluoroalkyl side chains, P3, was reported by Cho et al.⁵² The strong self-organization of semifluoroalkyl side chains induced the polymer backbones to form superstructures composed of ‘‘backbone crystals’’ and ‘‘side-chain crystals’’. In BGTC OTFT devices, P3 showed a μ_e of 3.93 cm² V⁻¹ s⁻¹ and the device performance was almost unchanged when exposed to air for three months. Device optimization by adding a small fraction of 1-chloronaphthalene to the casting solvent further improved the device performance, and a μ_e as high as 6.50 cm² V⁻¹ s⁻¹ was achieved.

Compared to sulfur, selenium has a larger atomic radius, which could promote orbital overlap and increase mobility.⁵³ Biselenophene was incorporated into an NDI-based polymer to afford P4 by Jenekhe et al.⁵⁴ Relative to the bithiophene analogue, P4 showed a red-shifted absorption spectrum and a ca. 0.1 eV lower LUMO energy level. In BGTC OTFT devices, P4 displayed a μ_e of 0.07 cm² V⁻¹ s⁻¹, which is higher than that of the bithiophene copolymer (0.04 cm² V⁻¹ s⁻¹) under the same device configuration. This can be ascribed to the enhanced crystallinity of P4 induced by the Se–Se interactions.

Thiophene–vinylene–thiophene (TVT) that has a highly coplanar structure can also promote polymer interchain packing and reduce the π–π stacking distance.^55,56 By incorporating a TVT unit, Kim et al. synthesized P5.⁵⁷ It formed highly ordered thin-films that featured strong π–π intermolecular interactions. However, the existence of the vinylene unit slightly raised the HOMO energy level, reducing the hole injection barrier. Consequently, an electron dominant transport characteristic with weak hole transport was observed in P5-based TGBC OTFT devices. When the Au electrodes were modified with Cs₂CO₃, P5 exhibited unipolar n-type transport with the highest μ_e of up to 1.80 cm² V⁻¹ s⁻¹. Selenophene–vinylene–selenophene (SVS) also has been used as the comonomer in an NDI-based polymer (P6) by the same group.⁵⁸ Similarly, P6 showed an electron dominant transport characteristic with a μ_e of up to 2.4 cm² V⁻¹ s⁻¹ in TGBC OTFT devices, which was better than that of P5. However, the Cs₂CO₃ treated unipolar n-type OTFT devices based on P6 exhibited a low μ_e of 0.50 cm² V⁻¹ s⁻¹.

There is a considerable torsion angle (ca. 40°) between NDI and the adjacent thiophene units.^59–62 This twist may hamper frontier molecular orbital delocalization and π–π stacking of the polymer backbones, thereby impairing electron transport. Efforts have been made to promote the backbone planarity. Heeney et al. found that insertion of a vinylene spacer between NDI and thiophene units could significantly reduce the torsion angle from 48° to 18°.⁶³ As a result, the polymer with the vinylene spacer (P7) showed a red-shifted UV-vis absorption spectrum and a denser lamellar structure with a π–π stacking distance of 3.5 Å compared to P1. In BGTC OTFT devices with Ba(OH)₂ modified Au electrodes, P7 exhibited unipolar n-type transport with the highest μ_e of 0.07 cm² V⁻¹ s⁻¹. An NDI-based polymer with an ethynylene spacer, P8, was also synthesized by Marder et al. This polymer showed inferior performance with a μ_e of 3.70 × 10⁻³ cm² V⁻¹ s⁻¹.⁶⁴

A fused π system can endow molecules with greater π-orbital overlap and provide a more efficient pathway for intermolecular charge transfer.^34,65–68 Luscombe et al. reported a series of NDI-based polymers with fused-thiophene units, namely, P9, P10 and P11.⁶⁹ In order to keep good solution processability, linear n-dodecyl chains were installed on quaterthiophene (P11). Surprisingly, these polymers exhibited poor performance compared to the polymer with a nonfused-thiophene unit. Among these polymers, P11 displayed the best performance with the maximum μ_e of 0.⁰¹² cm² V⁻¹ s⁻¹ in tetramethylsiloxane-bis(benzocyclobutene) derivative (BCB) modified BGTC OTFT devices. The inferior performances of these polymers are associated with their disordered thin-films. Incorporation of fused thiophene units would reduce conformational freedom of the polymer backbone, impeding self-organization of polymer backbones into ordered microstructures. A core-extended NDI-based polymer, P12, was developed by Gao et al.⁷⁰ By combining the electron-withdrawing ability of NDI and cyano groups, P12 showed a deep LUMO energy level of −4.25 eV, which is lower than those of the most reported NDI-based polymers. Despite the amorphous thin-film feature, BGBC OTFT devices of P12 still exhibited a high μ_e of 0.38 cm² V⁻¹ s⁻¹.

PDI is another rylene diimide building block with a low-lying LUMO energy level and has gained much attention in designing n-type CPs. The first PDI-based CP, P13, was synthesized by Zhan et al., which displayed unipolar n-type transport behaviour with a μ_e of 0.013 cm² V⁻¹ s⁻¹ in BGTC OTFT devices.⁷¹ The unipolar n-type transport characteristic of P13 is attributed to its low-lying HOMO and LUMO energy levels, which are −5.90 and −3.90 eV, respectively. An improved μ_e of 0.06 cm² V⁻¹ s⁻¹ for P13 was reported by Liu et al. using TGBC OTFT device geometry.⁷² Acetylene linkages were introduced into P14 by Zhan et al. to reduce the steric hindrance between PDI and dithienothiophene units.⁷³ Compared to P13, P14 showed a red-shifted absorption, deeper LUMO energy level and more ordered microstructures in thin-films. With the combination of a low-lying LUMO energy level and a highly ordered thin-film, P14 exhibited air stable electron transport with a high μ_e of 0.06 cm² V⁻¹ s⁻¹ in BGTC OTFT devices. In contrast, P13-based devices did not function with the same device structure in air, which only worked in nitrogen with a low μ_e of 0.017 cm² V⁻¹ s⁻¹. Facchetti et al. reported a PDI–bithiophene copolymer (P15), showing a μ_e of 2.0 × 10⁻³ cm² V⁻¹ s⁻¹ in BGTC OTFT devices.⁴⁷ By using phenothiazine as the comonomer, P16, with a μ_e approaching 0.05 cm² V⁻¹ s⁻¹, has been synthesized by Zhan et al.⁷⁴ Recently, benzothieno[3,2-b]benzothiophene (BTBT) was incorporated into a PDI-based polymer, yielding P17. In BGTC OTFT devices, P17 showed a μ_e of 0.042 cm² V⁻¹ s⁻¹.⁷⁵

1.2 D–A CPs based on bis(oxoindolinylidene)benzodifurandione. (3E,7E)-3,7-Bis(2-oxoindolin-3-ylidene)benzo[1,2-b:4,5-b′]-difuran-2,6(3H,7H)-dione (BDOPV) (also known as IBDF) has been proved to be an excellent A unit to construct n-type CPs, which was developed by Li et al. and Pei et al. independently.^76,77 Li et al. reported a BDOPV and thiophene copolymer, P18, which showed low-lying LUMO (−4.11 eV) and HOMO (−5.79 eV) energy levels.⁷⁶ In poly(methyl methacrylate) (PMMA) encapsulated BGBC devices, P18 exhibited a unipolar n-type transport characteristic with a μ_e of 5.4 × 10⁻³ cm² V⁻¹ s⁻¹. The low μ_e of P18 can be explained by its disordered thin-film morphology. Surprisingly, without PMMA encapsulation, the devices even showed slightly higher μ_e of 8.2 × 10⁻³ cm² V⁻¹ s⁻¹, while distinct hole transport could also be observed. The appearance of hole transport probably relates to the doping of the semiconductor layer by oxygen and/or moisture in ambient, which would alter the energy levels of the polymer.⁷⁸ Soon after, Pei et al. also reported the synthesis of BDOPV.⁷⁷ In order to balance the solubility and interchain interactions of the resultant polymer, long and further branched alkyl chain 4-octadecyldocosyl groups were installed on BDOPV. Through polymerization with (E)-1,2-bis(tributylstannyl)ethene, an electron-deficient poly(p-phenylenevinylene) (PPV) derivative (P19) was obtained. P19 showed very low LUMO and HOMO energy levels, which are −4.10 and −6.12 eV, respectively. Under ambient conditions, unipolar n-type transport performance with the highest μ_e of up to 1.1 cm² V⁻¹ s⁻¹ was achieved in P19-based TGBC devices. Moreover, the devices showed good air stability and the μ_e value was still as high as 0.31 cm² V⁻¹ s⁻¹ after being stored for 30 days. With this polymer, the effect of side chain branching point on the electron mobility was investigated systematically.⁷⁹ Upon moving the branching point away from the polymer backbones, the π–π stacking distance of the polymers decreased. However, the closer π–π stacking distance did not always lead to higher mobility, which is also correlated with the thin-film order and polymer packing conformation. When the branching point was situated at the 4-position (P19), the polymer exhibited the best performance with a μ_e of up to 1.40 cm² V⁻¹ s⁻¹. Other branching positions resulted in inferior performance. Subsequently, P20 with bithiophene as the D unit was also synthesized by Pei et al.⁸⁰ For the TGBC OTFT devices fabricated in nitrogen, unipolar n-type performance with the highest μ_e of 1.74 cm² V⁻¹ s⁻¹ was obtained. While for the devices fabricated under ambient conditions, P20 showed ambipolar charge transport performance because of oxygen doping. Recently, Pei et al. found that the solid-state microstructures of CPs are strongly dependent on the solution-state supramolecular structures. By controlling the self-assembly of P20 in solution, films with high crystallinity and good inter-domain connectivity can be obtained, thus the μ_e of P20 was significantly increased to 3.2 cm² V⁻¹ s^−1 81 (Fig. 2 and Table 2).

Fig. 2 Molecular structures of the n-type D–A conjugated polymers based on BDOPV, BDOPV derivatives and an IDTO unit.

Table 2 Molecular weights, energy levels and device properties of the n-type D–A conjugated polymers based on BDOPV, BDOPV derivatives and an IDTO unit

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d The device was fabricated with the mixture solvent of toluene and o-dichlorobenzene (1/4 vol%). e Estimated from the high VGS region. f Estimated from the low VGS region.
P18	34.2	1.46	−5.79	−4.11	BGBC	5.4 × 10^−3	Air	76
P19	37.6	2.38	−6.12	−4.10	TGBC	1.10	Air	77
	30.2	2.25	−6.16	−4.01	TGBC	1.40	Air	79
P19-1	37.4	2.04	−6.16	−4.01	TGBC	0.90	Air	77
P19-2	37.7	2.26	−6.15	−4.07	TGBC	0.60	Air	77
P19-3	34.9	2.01	−6.15	−3.97	TGBC	0.05	Air	77
P19-4	35.9	2.26	−6.16	−4.02	TGBC	0.76	Air	77
P19-5	35.2	2.44	−6.12	−4.01	TGBC	0.44	Air	77
P20	77.2	3.00	−5.72	−4.15	TGBC	1.74; 3.20^d	Air	80 and 81
P21	26.5	1.65	−5.57	−4.10	BGTC	0.65	Vacuum	82
P22	66.3	1.94	−6.19	−4.26	TGBC	1.70	Air	84
P23	53.8	3.06	−6.22	−4.30	TGBC	0.81	Air	84
P24	38.0	2.74	−5.96	−4.32	TGBC	1.24;^e 14.9^f	Air	85
P25	51.6	2.62	−5.80	−4.37	TGBC	3.22	Air	86
P26	23.5	2.15	−5.92	−3.98	BGBC	0.14	Nitrogen	93
P27	25.9	2.70	−5.78	−4.09	BGBC	0.18	Nitrogen	95

---

Qiu et al. synthesized P21 by copolymerizing BDOPV with thienothiophene. BGTC OTFT devices based on P21 showed unipolar electron transport behaviour with the highest μ_e of 0.65 cm² V⁻¹ s⁻¹ in high vacuum.⁸² However, the same polymer reported by Pei et al. showed ambipolar transport characteristics under air conditions.⁸³ This contradiction may be explained by their different testing environment since moisture and oxygen could induce hole transport as mentioned above. These results indicate that the testing environment also influences the polarity of charge carriers.

To further lower the HOMO and LUMO energy levels, fluorine (F) atoms were introduced on the BDOPV unit. P22 and P23 with F atoms at different positions on BDOPV have been developed by Pei et al.⁸⁴ Due to the strong electron-withdrawing ability of F atoms, the LUMO/HOMO energy levels of P22 and P23 are significantly lower than those of P19, which are −4.26/−6.19 eV for P22 and −4.30/−6.22 eV for P23. Additionally, compared to P19, P22 and P23 showed more ordered thin-films and smaller π–π stacking distances due to the presence of F⋯H weak intramolecular interactions. P22 exhibited a μ_e of 1.70 cm² V⁻¹ s⁻¹ in TGBC OTFT devices under air conditions, whereas P23 showed a decreased μ_e of 0.81 cm² V⁻¹ s⁻¹ with the same OTFT device configuration. This can be related to their different backbone conformations, leading to different interchain organization and thereby charge carrier transport ability. Afterwards, a strong electron-deficient polymer P24 was developed by Pei et al. based on four F atom substituted BDOPV.⁸⁵ P24 showed LUMO/HOMO energy levels of −4.32/−5.96 eV. The TGBC OTFT devices based on P24 showed an obviously kinked |I_SD|^1/2–V_GS plot. In the low V_GS region, a high μ_e of 14.9 cm² V⁻¹ s⁻¹ was achieved, while a much lower mobility of 1.24 cm² V⁻¹ s⁻¹ in the high V_GS region was obtained. The authors found that this non-linear behaviour was highly correlated with the microstructures in thin-films. Thin-films with highly ordered molecular packing tend to exhibit nonideal |I_SD|^1/2–V_GS characteristics, whereas the less ordered thin-films usually result in linear |I_SD|^1/2–V_GS curves in the device.

Pei et al. synthesized P25 by embedding pyridine units into BDOPV.⁸⁶ The replacement of benzene rings with pyridine units not only lowers the LUMO energy level, but also improves the planarity of polymer backbones because an imine-type nitrogen atom (I–N) is less sterically demanding than a CH moiety.⁸⁷ A typical n-type transport characteristic with the highest μ_e of 3.22 cm² V⁻¹ s⁻¹ was achieved in P25-based TGBC OTFT devices. Compared with P20, the enhanced performance of P25 can be ascribed to its lower LUMO energy level and less conformational disorder.

1.3 D–A polymers based on quinoidal compounds. Thienoquinoidal units, in particular, the ones with low-lying LUMO energy levels, are potential components to construct high performance n-type CPs due to their advantages of high planarity and rigidity.⁸⁸ The challenge in incorporating these units into polymer backbones is the introduction of functional groups for polymerization, because the potential polymerization sites in thienoquinoidal compounds are generally occupied by electron-withdrawing groups. One strategy to address this issue is to synthesize quinoidal units that are flanked or terminated with aromatic units, such as benzene or thiophene.^{32,33,89–91} Indophenine, a well-known blue dyestuff consisting of a quinoidal bithiophene, was reported to show promising ambipolar transport performance.⁹² The outer benzene rings on indophenine enable the incorporation of bromine atoms to construct a polymer. To lower the frontier molecular orbital energy levels and realize n-type transport behaviour, Li et al. elaborately oxidized the sulphur atoms in the thiophene units of indophenine to prepare thiophene-S,S-dioxidized indophenine (IDTO).⁹³ Importantly, a single isomer could be obtained owing to the O–O crowding effect. The polymer consisting of IDTO and thienothiophene, P26, showed low-lying LUMO and HOMO energy levels of −3.98 and −5.92 eV, respectively. BGBC OTFT devices with P26 showed unipolar n-type performance with a μ_e of 0.14 cm² V⁻¹ s⁻¹, despite the fact that polymer thin-films were rather disordered. The high mobility of the poorly crystalline P26 may be correlated with its planar and rigid backbone. It is revealed that the polymers with highly coplanar backbones can still achieve high mobility in the lack of long range order. For such polymers, the backbones play a leading role in the formation of the charge transport pathway owing to their efficient intramolecular charge delocalization, and only occasional π–π stacking is required to relay the charge carriers.⁹⁴ It should be noted that the BGBC devices without any encapsulation showed good air stability when they were annealed and/or measured in air. P26 is the first polymer based on a quinoidal unit exhibiting unipolar n-type transport performance. By replacing thienothiophene with bithiophene, P27, was synthesized by Li et al.⁹⁵ Compared with P26, P27 showed more ordered thin-films, thus an enhanced μ_e of 0.18 cm² V⁻¹ s⁻¹ was achieved in P27-based BGBC devices.

2. D–A CPs with weak D units

As we discussed above, D–A CPs tend to exhibit p-type or ambipolar transport characteristics due to their relatively high HOMO energy levels that originate from the electron-rich D unit. In order to address this issue, a strategy of constructing D–A polymers with a weak D unit has been developed, in which the weak D unit could help maintain a low HOMO energy level. Generally, incorporating electron-withdrawing groups, such as F atoms, I–N atoms and cyano groups (CN), into the electron-rich moieties can deliver the weak D units.

2.1 Incorporating F atoms. Introducing F atoms can not only greatly lower the frontier orbital energy levels of polymers, but also induce intramolecular F⋯S and F⋯H noncovalent interactions to enhance the planarity of the polymer backbones.^11,39,96,97 3,4-Difluorothiophene (2FT) was used as a weak D unit to construct D–A CPs with pyridine flanked diketopyrrolopyrrole (PyDPP) by Thelakkat et al., affording P28.⁹⁸ Compared with the thiophene unit, a pyridine unit can significantly lower the HOMO energy level and maintain good planarity of polymer backbones.⁸⁷ Indeed, P28 showed a low HOMO energy level of −6.11 eV. Therefore, a unipolar n-type characteristic with a μ_e of 0.11 cm² V⁻¹ s⁻¹ was achieved in BGBC OTFT devices based on the pristine P28 thin-film. Interestingly, the device performance was almost independent of the thermal annealing temperature. A similar μ_e of 0.13 cm² V⁻¹ s⁻¹ was recorded when the devices were annealed at 250 °C for 30 min. A fused heterocyclic substituent with increased intermolecular interaction was shown to enhance intermolecular charge carrier hopping. By replacing the pyridine unit with thieno[2,3-b]pyridine, a fused-heterocycle flanked DPP monomer was synthesized by McCulloch et al. and incorporated into P29 together with 2FT.⁹⁹ DFT calculations revealed that the backbone of P29 is almost completely planar because: (1) the less steric demanding nitrogen atom on the pyridine unit could suppress the steric hindrance of the DPP core; (2) the five-membered ring of the outer thiophene could minimize the twist between the monomers. P29 exhibited low-lying LUMO/HOMO of −4.10/−5.80 eV and was successfully utilized as an n-type polymer semiconductor in OTFT devices. A promising μ_e of 0.10 cm² V⁻¹ s⁻¹ was achieved in TGBC OTFT devices (Fig. 3 and Table 3).

Fig. 3 Molecular structures of the n-type D–A conjugated polymers based on F atom substituted weak D units.

Table 3 Molecular weights, energy levels and device properties of the n-type D–A conjugated polymers based on F atom substituted weak D units

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d Measured by photoelectron spectroscopy. e Estimated from EHOMO = ELUMO − E^optg, in which E^optg is the optical bandgap. f Estimated from ELUMO = EHOMO + E^optg.
P28	12.0	2.30	−6.11	−3.81	BGBC	0.11	Nitrogen	98
P29	24.4	5.10	−5.80^d	−4.10^f	TGBC	0.10	Nitrogen	99
P30	19.5	1.68	−5.28^e	−3.30	TGBC	0.82	Nitrogen	100
P31	13.8	2.18	−5.27^e	−3.43	TGBC	1.13	Nitrogen	100
P32	14.0	1.20	−5.75^e	−3.71	TGBC	0.91	Nitrogen	101
P33	102.7	1.77	−5.86	−3.66	TGBC	1.19	Air	105
P34	126.2	1.70	−5.89	−3.67	TGBC	1.35	Air	105
P35	52.9	1.54	−5.95	−3.80	TGBC	4.97	Air	26
P36	120.2	2.26	−5.96	−3.69	TGBC	2.45	Air	26
P37	88.0	1.94	−6.09	−3.92	TGBC	1.35	Air	105
P38	48.9	2.03	−5.95	−3.89	TGBC	0.93	Air	29
P39	17.0	1.94	−6.06	−3.99	TGBC	0.71	Air	29
P40	20.3	2.20	−6.04	−4.01	TGBC	1.24	Air	29
P41	75.0	4.48	−5.74	−3.76^f	BGTC	0.40	n/a	106
	37.0	3.24	−5.72	−3.71^f	TGBC	0.34	Air	107
P42	34.2	1.70	−6.04	−4.10	BGTC	0.77	Air	108
P43	7.8	1.67	−5.50^d	−3.85	TGBC	1.30	Nitrogen	109
P43-1	12.0	1.82	−5.46^e	−3.77	TGBC	0.16	Nitrogen	110
P44	6.6	1.45	−5.75^d	−3.75	TGBC	0.40	Nitrogen	109
P45	13.5	3.26	−5.76	−3.60	BGTC	0.42	Nitrogen	111
P46	3.20	1.40	−5.64	−3.87	BGTC	3.70 × 10^−4	Vacuum	112

---

Guo et al. designed and synthesized two bithiophene imide (BTI) derivatives, namely, single bond connected BTI dimers (s-2BTI) and fused BTI dimers (f-2BTI).¹⁰⁰ The outer thiophene rings in the BTI unit could tackle the steric hindrance challenge found in NDI- and PDI-based polymers, leading to planar polymer backbones. With these two building blocks, P30 and P31 have been developed using 2FT as the comonomer. It was found that the LUMO energy level of P31 was 0.13 eV lower than that of P30, suggesting that ring fusion leads to an enhanced electron-accepting ability. Moreover, geometry optimization revealed that P31 had a more linear and planar backbone than P30, thus resulting in higher structural order with stronger interchain interactions. Therefore, relative to P30, which exhibited a μ_e of 0.82 cm² V⁻¹ s⁻¹, P31 showed an improved μ_e of 1.13 cm² V⁻¹ s⁻¹ in TGBC OTFT devices. Recently, a single bond connected thiazole imide dimer (DTzTI) and its copolymer with 2FT (P32) were synthesized by Guo et al.¹⁰¹ P32 showed lower HOMO and LUMO energy levels as compared to P30, due to the electron-deficient nature of the thiazole unit,^102,103 which would help realize unipolar n-type transport. Another benefit of introducing the thiazole unit is that it leads to a more planar polymer backbone, owing to the less steric demanding I–N atom versus CH moiety in thiophene and the S⋯N intramolecular noncovalent interactions promoted by the N atom.⁹⁷ TGBC devices with P32 showed pure n-type transport characteristics with the highest μ_e of 0.91 cm² V⁻¹ s⁻¹.

TVT has been widely used to construct p-type polymers with very high mobility (>8 cm² V⁻¹ s⁻¹).^55,104 By introducing four F atoms on the β-positions of thiophene units in TVT, a weak D unit, (E)-1,2-bis(3,4-difluorothien-2-yl)ethene (4FTVT), was designed and synthesized by Geng et al.⁹⁶ Notably, the introduction of F atoms at the β-positions remarkably enhanced the direct arylation reactivity of α C–H bonds in the thiophene unit and evaded the unwanted possible β C–H activation. Accordingly, 4FTVT exhibited high reactivity in direct arylation polycondensation (DArP). Based on 4FTVT and PyDPP units, P33 with molecular weight >100 kDa has been synthesized via DArP by using Herrmann's catalyst.¹⁰⁵ P33 showed a deep HOMO level of −5.86 eV. DFT calculations revealed that P33 had a planar skeleton despite the existence of connecting pyridine and thiophene rings, which usually leads to a large torsion angle (∼20°) and a twisted polymer backbone. The planar backbone of P33 can be attributed to the presence of F⋯H noncovalent interactions. In TGBC OTFT devices, P33 displayed a unipolar n-type characteristic with a μ_e of 1.19 cm² V⁻¹ s⁻¹. By moving the bifurcation of the side alkyl chains away from the polymer backbone, P34 was synthesized by Geng et al.¹⁰⁵ P34 showed comparable HOMO and LUMO energy levels with P33, indicating that the alkyl chain branching point has a negligible influence on the energy levels. However, compared to P33, P34 showed an enhanced thin-film order and a smaller π–π stacking distance, thus a higher μ_e of 1.35 cm² V⁻¹ s⁻¹ was achieved in P34-based TGBC OTFT devices. With 4FTVT and fluorinated isoindigo (FIID) units, another n-type polymer, P35, was synthesized by Geng et al. via DArP.²⁶ The LUMO and HOMO energy levels for P35 are −3.80 and −5.95 eV, respectively, which are lower than those of P34. This can be explained by the multifluorination effect, that is, P35 was fluorinated on both D and A units while P34 was fluorinated only on the D unit. A unipolar high μ_e of 4.97 cm² V⁻¹ s⁻¹ was achieved in P35-based TGBC OTFT devices. It should be noted that the OTFT devices based on P35 exhibited good air stability. No obvious degradation was observed for TGBC OTFT devices after being stored under air conditions for three months, and the μ_e value was still as high as 0.05 cm² V⁻¹ s⁻¹ after two month storage in air for BGTC OTFT devices without any encapsulation. The good air stability can be attributed to the low-lying LUMO energy level of P35, which can resist the electron capture by water and oxygen during the device operation in air. Additionally, the hydrophobic nature of P35 caused by the multifluorination can also prevent the penetration of oxygen and water into the polymer thin-film.

Although P33, P34 and P35 showed unipolar n-type characteristics, an increase of the current was found in their transfer characteristics when the V_GS was below the turn-on voltage, indicating that weak hole transport still existed in the OTFT devices. It suggests that the HOMO energy levels of P33, P34 and P35 are not low enough to completely block the hole injection. As we discussed above, the existence of a vinylene unit would elevate the HOMO energy level. In order to obtain strictly unipolar n-type characteristics, another weak D unit, 3,3′,4,4′-tetrafluoro-2,2′-bithiophene (4FBT), was synthesized by Geng et al. Like 4FTVT, 4FBT also showed high reactivity in DArP, and high molecular weight CPs without defects can also be synthesized via DArP. With 4FBT as the comonomer, P36 and P37 have been developed via DArP using PyDPP and FIID as the A units, respectively.^26,105 Compared to the 4FTVT-based analogues, 4FBT-based polymers indeed showed lower HOMO energy levels, which are −5.96 and −6.09 eV for P36 and P37, respectively. As expected, P36 and P37 exhibited purely unipolar electron transport behaviour in TGBC OTFT devices, reflected by their typical unipolar n-type output characteristics and clear off-regimes in their transfer characteristics. The μ_e values for P36 and P37 were 2.45 and 1.35 cm² V⁻¹ s⁻¹, respectively. Recently, a series of n-type CPs P38, P39 and P40 were developed by Geng et al. based on 4FBT and bisisoindigo (bis-IID) derivatives containing different numbers of F or I–N atoms.²⁹ All these polymers showed low-lying LUMO and HOMO energy levels, and their energy levels can be feasibly tuned by the numbers of F and I–N atoms. TGBC OTFT devices based on these polymers all displayed unipolar n-type transport behaviours. Among these polymers, P40 exhibited the highest μ_e of 1.24 cm² V⁻¹ s⁻¹.

3,3′-Difluoro-2,2′-bithiophene (2FBT) and 1,2,4,5-tetrafluorobenzene (TFB) have also been used as the weak D units to construct n-type polymer. P41 comprising 2FBT and dithienylbenzodiimide (TBDI) was developed by Guo et al.¹⁰⁶ Unipolar n-type charge transport properties with the highest μ_e of 0.40 cm² V⁻¹ s⁻¹ were achieved in P41-based BGTC OTFT devices. Soon after, Liao et al. also reported the synthesis of P41, and they found that P41 exhibited a similar μ_e of 0.34 cm² V⁻¹ s⁻¹ in TGBC OTFT devices.¹⁰⁷ Another example of an n-type polymer with 2FBT as the weak D unit is P42, which was developed by Zhang et al.¹⁰⁸ The BGTC OTFT devices with P42 showed unipolar electron transport characteristics with a high μ_e of 0.77 cm² V⁻¹ s⁻¹. However, obviously an undesirable kink was observed in the |I_SD|^1/2–V_GS curves, which led to the overestimation of mobility. Therefore, the effective mobility was calculated using the reliability factor (r ≈ 30%), which was 0.23 cm² V⁻¹ s⁻¹. Notably, P42 exhibited excellent air stability of electron transport. For BGTC OTFT devices of P42 without any encapsulation, the unipolar electron mobility was still as high as 0.10 cm² V⁻¹ s⁻¹ with on/off ratios of >10⁶ after 60 days of air storage.

Like 4FTVT and 4FBT, TFB also showed high reactivity in DArP. With TFB and NDI units, P43 and P44 were synthesized by Sommer et al. via DArP.¹⁰⁹ The linkage unit, thiophene in P43 and furan in P44, significantly reduced the steric hindrance between NDI and TFB units due to the weak interactions of F⋯S/F⋯O and the avoidance of directly connecting benzene rings. Due to the presence of a debromination side reaction, M_ns of the polymers are relatively low, which are 7.8 kDa for P43 and 6.6 kDa for P44. Compared to P44, which exhibited a μ_e of 0.40 cm² V⁻¹ s⁻¹, P43 showed a more efficient electron transport with the highest μ_e of up to 1.30 cm² V⁻¹ s⁻¹. The enhanced OTFT performance of P43 can be explained by the reduced orientational disorder and the formation of a terrace-like thin-film morphology. Recently, P43-1 with a M_n of 12 kDa was developed by Jin et al. via Stille coupling polycondensation.¹¹⁰ In TGBC OTFT devices, P43-1 displayed a unipolar μ_e of 0.16 cm² V⁻¹ s⁻¹, which is almost one order of magnitude lower than that of P43. The dramatically decreased electron transport performance of P43-1 can be attributed to its disordered thin-film. As mirrored by the XRD analysis, up to 5 orders of (h00) diffraction peaks were observed in the P43 thin-film, while P43-1 showed an almost amorphous feature. This can be ascribed to the poor solubility of P43-1 due to the small size of side alkyl chains, leading to the difficulty in forming smooth thin-films. Other than NDI, rylene diimide with an azulene unit was recently reported to be another promising candidate for constructing n-type CPs. Gao et al. synthesized P45 by copolymerizing TFB with 2,2′-biazulene-1,1′,3,3′-tetracarboxylic diimides (TBAzDI).¹¹¹ In BGTC OTFT devices, P45 showed a μ_e of 0.42 cm² V⁻¹ s⁻¹. Very recently, Wang et al. reported a CP comprising TFB and benzobisthiadiazole, namely, P46.¹¹² Relative to its non-fluorinated counterpart that showed a p-type charge transport characteristic in BGTC OTFT devices,¹¹² P46 exhibited unipolar n-type charge transport behaviour due to its deep HOMO energy level, which could build large hole injection barriers. This result again highlights the advantage of weak D units for constructing n-type CPs.

2.2 Incorporating CN groups. Other than F, CN is also a strong electron-withdrawing group. However, weak D units with CN groups have been somewhat overlooked compared to the extensively studied F-substituted ones. Due to their relatively larger size, introducing CN groups into polymer backbones may cause steric hindrance between the comonomers and distort the polymer backbones. Thus, the substituted position of CN groups should be carefully considered to keep the planarity of the polymer backbone. Hwang et al. reported P47 by copolymerizing (E)-1,2-di-(3-cyanothiophen-2-yl)ethene (TVT2CN), which comprises CN groups at the 3 and 3′ positions of the TVT unit, with a thiophene flanked diketopyrrolopyrrole (TDPP) unit.¹¹³ DFT calculations revealed that P47 adopted a planar molecular structure with very small dihedral angles between neighbouring aromatic units. The LUMO orbital of P47 was well spread over the entire conjugated backbone, while the HOMO orbital was slightly localized on the DPP unit. These features are beneficial for efficient electron transport. Compared to its non-CN-substituted counterpart, P47 has significantly deeper HOMO and LUMO energy levels, which are −5.40 and −4.10 eV, respectively. Although the HOMO energy level of −5.40 eV is not deep enough to suppress the hole injection, electron dominant transport properties were observed in P47-based TGBC devices. This phenomenon can be related to the relatively localized HOMO orbital of P47, resulting in lower HOMO–HOMO interactions and blocked intermolecular hole hopping. The device performance of P47 was strongly dependent on the annealing temperature. The 100 °C annealed thin-film only showed a μ_e of 0.10 cm² V⁻¹ s⁻¹. When the annealing temperature was increased to 250 °C, the highest μ_e of 1.20 cm² V⁻¹ s⁻¹ was achieved. Very recently, Geng et al. developed a series of n-type polymers P48, P49 and P50 based on 3,3′-dicyano-2,2′-bithiophene (BT2CN).²⁷ All the polymers showed highly ordered molecular packing structures with short π–π stacking distances (3.54 Å to 3.60 Å), as revealed by two-dimensional grazing incidence wide angle X-ray scattering (2D-GIWAXS) studies, respectively. Those results suggest that the CN group has a negligible effect on the molecular packing if it locates at an appropriate position. Similar to P47, P48 also displayed electron dominant transport behaviour in TGBC OTFT devices, while P49 and P50 showed strictly unipolar electron transport due to their low-lying HOMO energy levels. The μ_e values of P48, P49 and P50 were 0.35, 0.30 and 0.27 cm² V⁻¹ s⁻¹, respectively, which were comparable to that of the benchmark n-type CP N2200 when the devices were fabricated under the same conditions. Interestingly, the electron mobility of those polymers was weakly dependent on the annealing temperature, and the values of μ_e were all above 0.2 cm² V⁻¹ s⁻¹ with the annealing temperatures in the range from 100 to 250 °C.

2.3 Incorporating I–N. Besides F atoms and CN groups, incorporating I–N can also lead to weak D units. Incorporating I–N not only deepens the energy levels but also promotes the planarity of polymer backbones because I–N is less steric demanding versus CH moiety.¹⁰¹ Thiazole is an attractive weak D unit to construct electron transport CPs because the large dipole of thiazole could induce strong dipole–dipole interactions.¹¹⁴ Holding these features, diverse n-type polymers with thiazole units have been explored. Reichmanis et al. reported a one-step synthetic strategy for preparing 2,2′-bithiazole and copolymerized it with a TDPP unit via Stille cross-coupling polycondensation, affording P51.¹¹⁵ Afterward, a similar polymer P51-1 was synthesized by Li et al. via DArP.¹¹⁶ Both polymers showed electron dominant properties in OTFT devices when Au was used as the source and drain electrodes. This contrasts with their isoelectronic conjugated polymer comprising electron-rich bithiophene and TDPP, which displayed merely hole transport characteristics.¹¹⁷ The conversion of charge carrier polarity confirms the potential of the thiazole unit in the development of electron transport polymers. BGTC OTFT devices based on P51 with low work function calcium and ethoxylated polyethylenimine (PEIE) treated silver as the source and drain electrodes were fabricated, which showed unipolar electron transport behaviour with the highest μ_e of 0.31 cm² V⁻¹ s⁻¹. It should be mentioned that the devices prepared from nonchlorinated solvents, such as xylene and tetrahydronaphthalene, exhibited similar performance to those produced from o-dichlorobenzene. By replacing the flanked thiophene units with thiazoles, P52 was developed by Reichmanis et al.³⁷ In comparison with P51, P52 showed a ∼0.2 eV deeper HOMO energy level under identical measurement conditions, indicating that higher hole injection barriers would be established. Therefore, BGBC OTFT devices based on P52 exhibited purely unipolar electron transport with the highest μ_e of 0.067 cm² V⁻¹ s⁻¹. No appreciable decrease was observed in μ_e after storing the device in air over 4 months, indicative of the good air stability of the devices. A similar polymer P52-1 was also reported by Choi et al.,¹¹⁸ which showed a μ_e of 0.041 cm² V⁻¹ s⁻¹ in TGBC OTFT devices under air conditions (Fig. 4 and Table 4).

Fig. 4 Molecular structures of the n-type D–A conjugated polymers based on CN groups and I–N substituted weak donor units.

Table 4 Molecular weights, energy levels and device properties of the n-type D–A conjugated polymers based on CN groups and I–N substituted weak donor units

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d Measured by photoelectron spectroscopy. e Estimated from ELUMO = EHOMO + E^optg. f Measured by photoemission yield spectroscopy. g Measured by low-energy inverse photoemission spectroscopy. h With calcium and ethoxylated polyethylenimine (PEIE) modified silver as source and drain electrodes. i The device was fabricated by blade-coating. j The device was fabricated by off-centre spin-coating.
P47	53.0	1.30	−5.40	−4.10^e	TGBC	1.20	Nitrogen	113
P48	155.0	1.34	−5.41	−3.67	TGBC	0.35	Air	27
P49	275.0	1.30	−5.83	−3.75	TGBC	0.30	Air	27
P50	111.0	1.54	−6.15	−3.92	TGBC	0.25	Air	27
P51	64.0	3.60	−5.54^d	−3.75	BGTC	0.31^h	Nitrogen	115
P51-1	18.0	3.80	−5.56	−3.63	TGBC	0.53	Air	116
P52	26.0	1.35	−5.71^d	−4.10^e	BGBC	0.067	Nitrogen	37
P52-1	44.0	1.74	−5.92	−4.32^e	BGTC	0.041	Air	118
P53	31.9	1.70	−5.87	−4.13^e	TGBC	0.83	Air	119
P54	40.4	2.58	−6.06	−4.28	BGTC	0.055	Vacuum	120
P55	14.3	2.50	−5.99	−4.18	BGBC	0.017	Nitrogen	95
P56	45.0	3.50	−5.78^d	−4.40	BGBC	0.05^i	Nitrogen	121
P56-1	89.0	4.98	n/a	−3.83	TGBC	0.085^j	Nitrogen	122
P57	79.4	1.80	−6.24	−3.89	TGBC	0.57	Air	28
P58	115.6	1.80	−6.01	−3.85	TGBC	0.37	Air	28
P59	118.3	1.50	−5.91	−3.83	TGBC	0.39	Air	28
P60	56.1	2.03	−6.12^f	−3.56^g	BGTC	0.05	Air	123

---

With bithiazole and thiophene flanked difluorobenzoxadiazole, P53 was developed by Guo et al.¹¹⁹ Notably, although electron-rich thiophene units were flanked on the difluorobenzoxadiazole unit, P53 had a low-lying HOMO energy level of −5.87 eV, which could block hole injection. P53-based TGBC OTFT devices with CYTOP as the dielectric showed electron transport behaviour with a μ_e of 0.71 cm² V⁻¹ s⁻¹. During device optimization, they found that the devices with PMMA dielectric displayed an improved performance with a higher μ_e of up to 0.83 cm² V⁻¹ s⁻¹. Subsequently, Cho et al. reported P54 based on bithiazole and a BDOPV derivative.¹²⁰ P54 showed deep LUMO/HOMO energy levels of −4.28/−6.06 eV, which were low enough to stabilize electron transport in air and to block hole injection. Consequently, P54 BGTC OTFT devices without any encapsulation exhibited unipolar electron transport with a μ_e of 0.055 cm² V⁻¹ s⁻¹ under air conditions. Importantly, the BGTC OTFT devices displayed excellent air stability and maintained unipolar electron transport with a μ_e of up to 0.01 cm² V⁻¹ s⁻¹ after one year storage in air. The copolymer (P55) of quinoidal acceptor IDTO and a bithiazole unit was synthesized by Li et al.⁹⁵ In BGBC OTFT devices, P55 exhibited inferior performance with a μ_e of 0.017 cm² V⁻¹ s⁻¹ due to its low molecular weight and the amorphous thin-film.

As aforementioned, the NDI unit has been widely used to construct n-type CPs. However, weak hole injection can be observed for the most of NDI-based CPs due to their rather high HOMO energy levels.²⁸ To address this issue, Reichmanis et al. synthesized a n-type polymer P56 containing bithiazole and NDI units.¹²¹ As expected, P56 showed ∼0.4 eV deeper HOMO energy level than the NDI–bithiophene copolymer (N2200). Therefore, P56 exhibited a strictly unipolar n-type transport characteristic. The BGBC OTFT devices of P56 fabricated by blade-coating displayed a μ_e of 0.05 cm² V⁻¹ s⁻¹, which is lower than that of N2200. The low μ_e of P56 may be correlated with its twisted conjugated backbones derived from the steric hindrance between NDI and thiazole units, hindering polymer backbone organization into highly ordered microstructures that would be expected to result in efficient intermolecular charge transport. A similar polymer P56-1 was synthesized by Sommer et al. via DArP with Pd₂dba₃ and tris(o-anisyl)phosphine as the catalyst and ligand, respectively.¹²² It should be noted that homo-coupling free P56-1 with a high molecular weight of up to 89.0 kDa can be synthesized by DArP in less than 1 hour, suggesting that DArP is an efficient protocol to synthesize CPs. P56-1 showed unipolar electron transport with the highest μ_e of 0.085 cm² V⁻¹ s⁻¹ in TGBC OTFT devices fabricated by the off-centre spin-coating method. In order to simultaneously reduce the HOMO energy level and improve the planarity of polymer skeletons, Cao et al. designed and synthesized a thiazole flanked NDI unit (TzNDI) with the thiazole N atom pointing toward the NDI unit.²⁸ With this linking method, the twist angle between NDI and thiazole could be significantly reduced due to the less steric demanding I–N. Based on TzNDI, a series of n-type polymers have been developed, including P57, P58 and P59. Indeed, the TzNDI unit endows the resultant polymers with good backbone planarity and low-lying HOMO energy levels. In TGBC OTFT devices, all these polymers exhibited strictly unipolar electron transport performance with the highest μ_e of up to 0.57 cm² V⁻¹ s⁻¹, outperforming N2200 (0.41 cm² V⁻¹ s⁻¹) under the same device fabrication conditions.

f-2BTI with the imide groups located at the centre of bithiophene should minimize the steric hindrance between the carbonyl bridge and the adjacent atoms to achieve a planar polymer backbone. By copolymerizing f-2BTI and a bithiazole unit, Takimiya et al. synthesized P60.¹²³ The polymer showed a deep-lying HOMO energy level of −6.12 eV as determined by photoemission yield spectroscopy (PYS). In OTFT devices with the BGTC configuration, P60 exhibited a unipolar μ_e of 0.05 cm² V⁻¹ s⁻¹.

3. D–A CPs with dual-A

Introducing two electron acceptor building blocks into a polymer backbone to construct dual-A architecture (A₁–D–A₂–D) was an effective approach to tune the energy levels of D–A polymers.^124–126 This strategy could increase the loading of A unit in polymer chains, enabling the ease of realizing unipolar n-type transport via further energy level modulation (Fig. 5 and Table 5).

Fig. 5 Molecular structures of the n-type D–A conjugated polymers based on dual-A.

Table 5 Molecular weights, energy levels and device properties of the n-type D–A conjugated polymers based on dual-A

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d Measured by photoelectron spectroscopy. e Estimated from ELUMO = EHOMO + E^optg. f The Au electrodes were modified with PFBT. g The device was modified with NTMS.
P61	54.0	1.25	−5.84^d	−3.99^e	TGBC	0.01	n/a	127
P62	147.0	2.65	−5.98^d	−4.22^e	BGTC	1.3 × 10^−3	n/a	128
P63	48.5	4.56	−6.24	−3.85	TGBC	2.2^f	Air	129
P64	56.1	4.24	−6.20	−3.88	TGBC	3.5^f	Air	129
P65	11.1	1.84	−5.35	−3.72	BGTC	1.1 × 10^−3	Vacuum	133
P66	10.2	1.53	−5.20	−3.85	BGTC	4.1 × 10^−3	Vacuum	133
P67	11.9	1.54	−5.68	−3.93	BGTC	0.039	Vacuum	133
P68	10.3	1.61	−5.34	−4.03	BGTC	2.7 × 10^−3	Vacuum	133
P69	313.6	3.40	−5.77	−3.78	BGTC	0.92	Vacuum	134
P70	122.5	2.70	−6.01	−3.86	BGTC	0.46	Vacuum	134
P71	141.0	2.00	−5.87	−3.81	BGTC	2.11	Vacuum	134
P72	133.1	2.50	−5.45	−3.88	BGTC	5.35^g	Vacuum	134
P73	31.6	2.50	−5.71	−3.80	BGTC	3.87^g	Vacuum	45
P74	54.9	1.80	−5.40	−3.87	BGTC	7.16^g	Vacuum	45
P75	22.1	3.68	−5.82	−3.76	BGTC	0.42	Nitrogen	111

---

Benzodipyrrolidone (BDP), which is similar in structure to the DPP unit, is a promising building block for CPs with high mobility. McCulloch et al. reported an A₁–D–A₂–D polymer P61 by the combination of BDP with a benzothiadiazole (BT) unit.¹²⁷ Compared with the typical D–A polymers with BDP as the A unit,¹²⁷ P61 indeed showed a much lower HOMO energy level of −5.84 eV, being suitably deep to block hole injection from the commonly used Au electrodes. In TGBC OTFT devices, P61 displayed electron transport with a μ_e of 0.01 cm² V⁻¹ s⁻¹. Afterward, another A₁–D–A₂–D polymer P62 based on BDP and thienopyrroledione (TPD) was synthesized by Kanbara et al. via DArP with Pd(AcO)₂ and PCy₃·HBF₄ as the catalyst and ligand, respectively.¹²⁸ However, P62 exhibited inferior electron transport performance with a μ_e of 0.0013 cm² V⁻¹ s⁻¹, due to its amorphous thin-film likely originating from the steric hindrance of the bulky alkyl chains.

NDI was often selected as one of the A moieties in dual-A polymers due to its strong electron-withdrawing nature. Based on NDI and fluorinated benzothiadiazole (FBTz), P63 with a A₁–D–A₂–D structure was synthesized by Liu et al.¹²⁹ Using thiophene as the linkage could avoid the steric hindrance between NDI and FBTz, and thus maintain good planarity of the polymer backbone. Cyclic voltammetry experiments revealed that P63 exhibited a strong reduction process and a weak oxidation process with the LUMO/HOMO energy levels of −3.85 eV/−6.24 eV. P63 showed unipolar electron transport with the highest μ_e of 2.20 cm² V⁻¹ s⁻¹ in TGBC OTFT devices, where pentafluorobenzenethiol (PFBT) modified Au was used as the source and drain electrodes to facilitate the electron injection. It has been demonstrated that the work function of Au electrodes was tuned from 5.13 eV (bare Au) to 4.77 eV (PFBT modified Au),¹³⁰ and therefore the energetic mismatch between Au electrodes and the LUMO energy level of the polymer was reduced. Compared with sulfur, selenium has a larger atomic radius, which could facilitate orbital overlap and charge carrier transport. Besides, the empty orbital in selenium could improve the electron-accepting ability of selenophene, and in turn decrease the LUMO energy level of selenophene-containing polymers.^131,132 With these advantages, P64 was developed by Liu et al., in which selenophene was used as the linking unit.¹²⁹ It was found that P64 possessed a narrower bandgap and deeper LUMO energy level than P63. Thanks to the smooth surface morphology and short π–π stacking distance, P64 exhibited a high μ_e of 3.50 cm² V⁻¹ s⁻¹ in TGBC OTFT devices.

Soon after, Michinobu et al. reported a series of A₁–D–A₂–D polymers, namely, P65, P66, P67 and P68, composed of NDI and benzobisthiadiazole (BBT) or its heteroatom-substituted derivatives (SN).¹³³ Similar to Liu's results, replacing the sulfur atom in the thiadiazole heterocycles with a selenium atom led to deeper LUMO energy levels and stronger π–π interactions of the polymers. Consequently, P66 (0.0041 cm² V⁻¹ s⁻¹) exhibited three times higher μ_e than P65 (0.0011 cm² V⁻¹ s⁻¹) at the optimized annealing temperature of 250 °C in BGTC OTFT devices. However, the other selenium containing polymer, P68 (0.0027 cm² V⁻¹ s⁻¹), showed much lower μ_e than its thiophene containing analogue P67 (0.039 cm² V⁻¹ s⁻¹) due to its low thermal stability and decomposition problem during the thermal annealing process. Subsequently, Michinobu et al. developed another four A₁–D–A₂–D polymers, P69, P70, P71 and P72, based on NDI and BT units.¹³⁴ With these polymers, the effect of different N-substituted positions on the OTFT device performance was studied in detail. It should be noted that an optimized Stille polymerization protocol with chlorobenzene as the solvent and Pd(0)/Cu(I) as the cocatalyst was developed to synthesize high molecular weight polymers. Chlorobenzene provides good solubility for these polymers, while Cu(I) can react with organostannanes to produce transient organocopper intermediates and thereby accelerate the transmetallation reaction in the catalytic circle,¹³⁵ thus giving rise to high molecular weight polymers. BGTC OTFT devices with P69 showed electron dominant transport properties with a μ_e of 0.92 cm² V⁻¹ s⁻¹. Replacing the thiophene unit in P69 with thiazole produced P70, which showed a unipolar electron transport characteristic with a μ_e of 0.46 cm² V⁻¹ s⁻¹. P71, which contained I–N on the electron-withdrawing BT unit, exhibited an even higher μ_e of 2.11 cm² V⁻¹ s⁻¹. P72 with two I–N atoms embedding into the triple fused ring structure of the BT unit displayed the highest μ_e of 4.87 cm² V⁻¹ s⁻¹. However, due to the high HOMO energy level of P72, weak hole transport was observed in the OTFT devices. In order to realize unipolar electron transport properties, the BGTC OTFT devices based on P72 was modified with [3-(N,N-dimethylamino)-propyl]trimethoxysilane (NTMS), in which the amine groups could trap the hole and thus suppress the hole transport. In such a case, a unipolar μ_e of up to 5.35 cm² V⁻¹ s⁻¹ was obtained. P65 and P72 possessed similar polymer backbones and only differed in the alkyl chains. P65 contained small alkyl chains on both D and A units while P72 carried long alkyl chains on the A unit. However, the μ_e of P72 is three orders of magnitude higher than that of P65. This result suggests that the position and selection of alkyl chains are as important as optimizing the polymer backbone, since the alkyl chains not only affect the solubility of the polymer but also influence the packing behaviour of the polymer and therefore the device performance.^136,137

As we discussed above, the connection of NDI and a thiophene unit (NDT–T) would create large torsion and lead to a twisted polymer backbone, which impairs charge transport. Inserting a vinylene group between NDI and the thiophene unit (NDI–V–T) provides an effective way to solve this issue, due to the formation of H⋯O intramolecular hydrogen bonds between vinylene and NDI units, ensuring planar backbones of the polymer.⁹⁷ With this design strategy, P73 and P74 were recently developed by Michinobu et al.⁴⁵ In P73, the F atom on the BT unit can form additional intramolecular hydrogen bonds via H⋯F interactions, further locking the conformation of polymer backbones. As demonstrated by the DFT calculations, the torsion angles were significantly decreased after incorporating vinylene groups and F atoms, thus planar backbones were obtained in P73 and P74, which in turn led to high order polymer thin-films. In BGTC OTFT devices modified with NTMS, P73 displayed unipolar electron transport with a high μ_e of 3.87 cm² V⁻¹ s⁻¹. P74 exhibited a much higher μ_e of 7.16 cm² V⁻¹ s⁻¹, which is the highest value for unipolar n-type CPs. 2D-GIWAXS revealed that P73 adopted an edge-on orientation molecular packing with a π–π stacking distance of 3.49 Å, while P74 showed a bimodal packing behaviour with a mixture of face-on and edge-on backbone orientations and an extremely short π–π stacking distance of 3.40 Å. It is believed that the bimodal packing is promising for achieving high mobility because the face-on orientation packing molecules can provide a chance for charge carriers to move around the boundaries between crystalline and amorphous domains.^138–140 Moreover, compared with edge-on orientation, face-on molecular packing is beneficial for the injection of charge carriers from electrodes to the semiconductor layer.¹⁴¹ These events were facilitated by the extremely short π–π stacking distance. Therefore, better electron transport was realized for P74 in comparison with P73.

TBAzDI, an imide-functionalized azulene acceptor unit, was considered as another promising building block for n-type CPs. Recently, an A₁–D–A₂–D polymer P75 was developed by Gao et al.¹¹¹ DFT calculations revealed that the LUMO orbital of P75 was well delocalized over the entire polymer backbone, which was greatly different from those of NDI- and PDI-based n-type D–A polymers, where the LUMO orbitals were usually localized on the A units.^73,142 The delocalized LUMO orbital could encourage LUMO–LUMO interactions, improving electron transfer between polymer chains. The LUMO and HOMO energy levels of P75 are −3.76 and −5.82 eV, respectively, lower than those of TBAzDI-based polymer with a weak D unit of TFB (P45). The above results indicate that the dual-A design strategy not only significantly brings down the energy levels of polymers but also promotes the delocalization of LUMO orbitals, which are of critical significance for n-type polymers. In BGTC OTFT devices, P75 exhibited a high μ_e of 0.42 cm² V⁻¹ s⁻¹, making P75 among the best n-type CPs for BGTC OTFT devices.

4. Acceptor–acceptor (A–A) type CPs

Recently, a new design concept, called “A–A” type polymer, has been proposed to create n-type CPs. This strategy can eliminate the adverse effect from the electron-rich D unit completely, yielding unipolar n-type CPs feasibly. However, the synthesis of A–A type CPs with high molecular weights, which dramatically affects the charge transport performance of CPs,^143,144 remains a great challenge due to: (1) the difficulty in synthesizing high purity organometal functionalized A monomers because of their reduced chemical reactivity; (2) the instability nature of the organometal functionalized electron-accepting heteroarenes.¹⁴⁵ Until now, only a few A–A type CPs have been reported, and they are commonly synthesized via Yamamoto^146–148 or Stille polycondensation.^{101,149–151} Certainly, some A–A type CPs, such as the polymers comprising a IID or BT unit, also can be synthesized by Suzuki polycondensation owing to the accessibility and stability of their organoboron derivatives.^152–155 Notably, the recent advances in DArP have significantly promoted the development of this class of polymers.^156,157 From the perspective of material design, A–A type CPs are prone to suffer from the problem of twisted polymer backbones because the electron-withdrawing substituents/groups, such as imide and cyano groups that can endow the A units with low-lying frontier molecular orbital energy levels, are usually bulky and induce significant steric crowding between the adjacent A units.^45,149 The twisted polymer chains would cause poor molecular packing and less ordered thin-films, impairing charge transport. Therefore, in order to reduce the backbone torsion of A–A type CPs, elaborate molecular design is required (Fig. 6 and Table 6).

Fig. 6 Molecular structures of the n-type A–A conjugated polymers.

Table 6 Molecular weights, energy levels and device properties of the n-type A–A conjugated polymers

Polymers	M n (kDa)	Đ	E HOMO (eV)	E LUMO (eV)	OTFT structure	μ e (cm^2 V^−1 s^−1)	Measured environment	Ref.
a Polydispersity. b HOMO energy level measured by cyclic voltammetry, unless explicitly stated. c LUMO energy level measured by cyclic voltammetry, unless explicitly stated. d Measured by photoelectron spectroscopy. e Estimated from EHOMO = ELUMO − E^optg, in which E^optg is the optical bandgap. f Estimated from ELUMO = EHOMO + E^optg. g Estimated from TGBC OTFT devices with polyolefin–polyacrylate dielectric. h The device was modified with CsF. i The device was fabricated by off-centre spin coating.
P76	71.1	2.90	−6.50^e	−3.76	BGTC	6.0 × 10^−4	Nitrogen	146
P77	5.2	1.74	−5.14	−3.03	BGTC	0.20	n/a	152
P78	15.7	1.73	−6.20^d	−4.20	BGTC	0.21	Air	159
P79	20.5	2.53	−5.50	−4.40	BGTC	0.31	Air	160
P80	30.6	1.16	−5.49	−3.08	BGTC	0.032	n/a	152
P81	8.2	1.38	−5.94	−4.00	BGTC	0.01	Air	161
P82-L	3.6	2.20	−6.28	−3.47	BGTC	0.011	Vacuum	147
P82-M	7.2	1.98	−6.28	−3.47	BGTC	0.038, 0.14^g	Air, vacuum	148
P82-M*	5.4	1.43	n/a	n/a	n/a	0.018^h	Nitrogen	149
P82-H	12.7	2.12	−5.46^e	−3.48	TGBC	1.53,^h 3.71^h^,^i	Nitrogen	149
P83	13.3	2.20	−5.39^e	−3.53	TGBC	1.34^h	Nitrogen	149
P84	11.7	1.90	−5.41^e	−3.59	TGBC	0.53^h	Nitrogen	149
P85	9.4	1.60	−5.47^e	−3.67	TGBC	0.21^h	Nitrogen	149
P86	9.2	1.10	−5.49^e	−3.72	TGBC	0.014^h	Nitrogen	149
P87	5.8	2.60	−6.17^e	−3.94	TGBC	0.015^i	Nitrogen	101
P88	7.0	1.10	−5.78^e	−3.77	TGBC	1.61^i	Nitrogen	163
P89	20	2.20	−6.10	−4.20	BGBC	3.5 × 10^−3	Nitrogen	156
P90	24	2.20	−6.00	−4.20	BGBC	3.0 × 10^−4	Nitrogen	156
P90-1	57.1	3.30	−5.70	−3.66	BGTC	0.01	Nitrogen	153
P91	25.0	1.32	−5.68	−3.54	BGTC	0.22	Nitrogen	153
P92	14.0	1.60	−5.80^d	−4.10^f	TGBC	1.0	n/a	154
P93	120	1.50	−5.40^d	−3.40^f	TGBC	0.020	Nitrogen	145
P94	n/a	n/a	n/a	−4.12	TGBC	2.2 × 10^−6	Air-free	151

---

The promising electron transport properties of NDI-based polymers inspire the exploration of A–A type polymers with an NDI unit. An NDI-based homopolymer P76 with branched 2-octyldodecyl substituents was first synthesized by Luscombe et al. via Yamamoto coupling polycondensation.¹⁴⁶ Compared with the NDI–thiophene D–A polymers, P76 showed much lower HOMO energy level, which is −6.50 eV, suggesting that A–A type CPs are highly desirable as unipolar electron transport semiconductors. However, P76 exhibited a limited μ_e of 6.0 × 10⁻⁴ cm² V⁻¹ s⁻¹ due to its amorphous thin-film, which was probably derived from the steric repulsion of the NDI unit and the absence of space for alkyl chain interdigitation. Later, an NDI- and BT-based A–A type polymer P77 was developed by Iyer et al. via Suzuki polycondensation.¹⁵² As a spacer, the BT unit could avoid the steric hindrance of the NDI unit and encourage alkyl chain interdigitation, thus promoting polymer chain packing and the formation of lamellar nanostructures with compact alkyl chain interdigitation. As a result, P77 displayed highly ordered thin-films as revealed by the obvious diffraction peaks in the out-of-plane XRD. The well-organized microstructures of P77 enabled more efficient electron transport than P76. In BGTC OTFT devices, P77 showed the highest μ_e of 0.20 cm² V⁻¹ s⁻¹.

A thiophene fused NDI derivative, 4,5,9,10-naphtho[2,3-b:6,7-b′]dithiophenediimide (NDTI), was reported by Takimiya et al.¹⁵⁸ It is an attractive archetypical building block for CPs for the following reasons: (1) the planar and rigid structure is beneficial to charge transport; (2) the outer thiophene rings help create polymers with good planarity. A–A type polymers P78 and P79 were synthesized via CuI assisted Stille coupling polycondensation by copolymerizing NDTI with naphthobisthiadiazole (NTz) and benzobisthiadiazole (BBT), respectively.^159,160 Both polymers showed near-infrared absorption spectra and remarkable low-lying LUMO energy levels of ∼−4.40 eV. As expected, both polymers afforded air stable unipolar electron transport with μ_e of 0.21 and 0.31 cm² V⁻¹ s⁻¹ for P78 and P79, respectively, in BGTC OTFT devices.

A PDI unit has also been used as an A unit to develop A–A type polymers with unipolar electron transport characteristics. P80¹⁵² and P81¹⁶¹ were synthesized by Iyer et al. and Zhan et al., respectively. The μ_e for P80 and P81 were 0.032 and 0.01 cm² V⁻¹ s⁻¹, respectively.

A BTI unit was designed and synthesized by Marks et al. in 2008.¹⁴⁷ Moving the imide group to the centre of bithiophene could greatly attenuate the steric hindrance, therefore improving the backbone planarity of the resultant polymers. A BTI homopolymer P82 was synthesized using Yamamoto coupling polycondensation. When the reaction was conducted at 60 °C in DMF, a polymer with a low number-average molecular weight (M_n) of 3.6 kDa was obtained (P82-L), after being purified by multiple precipitations.¹⁴⁷ By increasing the polymerization temperature from 60 °C to 80 °C, P82 with an increased M_n of 7.2 kDa (P82-M) was yielded.¹⁴⁸ Polymerization at 80 °C using DMF and toluene as the co-solvent led to insoluble P82.¹⁴⁸ In BGTC OTFT devices, P82-M showed a μ_e of 0.038 cm² V⁻¹ s⁻¹, two times higher than that of P82-L (0.011 cm² V⁻¹ s⁻¹). This trend is common for CPs and in good agreement with the XRD and atomic force microscopy (AFM) results, where more ordered thin-film and better-defined grains were observed for P82-M. TGBC OTFT devices with polyolefin–polyacrylate dielectric afforded an improved μ_e of 0.14 cm² V⁻¹ s⁻¹ using P82-M as the channel material.¹⁴⁸

Stille coupling polycondensation using hexamethylditin is another protocol to synthesize A–A type polymers. In order to investigate the effect of the polymerization method on the molecular weight, polymer structure and device performance, P82 was prepared by Stille and Yamamoto coupling polycondensation.¹⁴⁹ Compared with Yamamoto polycondensation, which yielded P82 with a M_n of 5.4 kDa (P82-M*), Stille polycondensation furnished P82 with a high M_n of 12.7 kDa (P82-H), indicating that Stille polycondensation is a more effective method to prepare high M_n A–A type polymers. P82-H showed much better performance with the highest μ_e of up to 1.53 cm² V⁻¹ s⁻¹ in TGBC OTFT devices modified with CsF for reducing the electron injection barriers and suppressing the weak hole injection. In contrast, P82-M* displayed a low μ_e of 0.018 cm² V⁻¹ s⁻¹ under the same device configuration. Apart from M_n, the structural defects and Br end groups that act as charge carrier traps may also contribute to the poor performance of P82-M*. This result clearly indicates that the polymerization method and hence polymer quality are critical to the device performance. By using the off-centre spin-coating technique, the μ_e of P82-H can be further improved to 3.71 cm² V⁻¹ s⁻¹, which is the highest reported value for A–A type n-type polymers.

With a combination of Yamamoto coupling, ester hydrolysis, carboxylic acid dehydration, and immidization reactions, a series of highly electron-deficient (semi)ladder BTI derivatives (BTIn) with up to 5 imide groups and 15 rings in a row have been synthesized by Guo et al.¹⁶² Based on these units, homopolymers PBTIn, namely, P83, P84, P85 and P86, have been developed using Stille polycondensation.¹⁴⁹ Compared with P82, PBTIn showed red-shifted absorption and gradually down-shifted HOMO and LUMO energy levels. However, the electron transport performance of PBTIn was lower than that of P82, and a monotonic decrease of μ_e from 1.34 cm² V⁻¹ s⁻¹ for P83 to 0.014 cm² V⁻¹ s⁻¹ for P86 was observed with an extension of the conjugation length of the monomer. It is likely because that the larger π-conjugation monomer causes a reduced freedom of molecular motion in the solid state, impacting self-organization of polymer chains into ordered microstructures.

Although BTI-based A–A type polymers delivered good electron transport properties, weak hole injection could be observed in the OTFT devices without CsF modification, as indicated by the lack of clear off-regime in their transfer curves. This phenomenon could be attributed to the electron-rich thiophene unit in the BTI core, resulting in relatively high-lying HOMO energy levels of the resultant polymers. In order to bring down the HOMO energy level of the polymer and hence block hole injection, bithiazole imide (BTzI) was designed and synthesized by Guo et al. and its homopolymer P87 was developed by Stille polycondensation.¹⁰¹ It should be noted that the polymerization of the BTzI monomer appeared to be problematic under typical Pd-mediated Stille coupling conditions, possibly due to the chelation of the Pd catalyst with the N atoms on thiazole moieties. After adding CuI as the co-catalyst, P87 with a M_n of 5.8 kDa was obtained. As expected, P87 showed a deep HOMO energy level of −6.17 eV, which was ∼0.7 eV lower than that of P82-H when measured by the same method. However, it was found that the thin-film order of the polymer was significantly reduced after replacing the two thiophene units in the BTI core with thiazole moieties. Thus P87 showed inferior electron mobility. The highest unipolar μ_e of P87 is 0.015 cm² V⁻¹ s⁻¹ in TGBC OTFT devices without CsF modification. Subsequently, the same group developed an asymmetry BTI derivative, BTzTI, by replacing one thiophene unit in the BTI core with a thiazole moiety to address the relatively high-lying HOMO energy level of BTI-based A–A type polymers and tackle the problem of low ordered thin-films for P87.¹⁶³ In addition, to ensure the regioregularity of the polymer backbone and overcome the low reactivity of the C–Br bond on the thiazole unit, the BTzTI unit was first dimerized and the resulting dimer (DTzTI) was polymerized by using Stille polycondensation. As a result, the A–A type polymer P88 with a decent M_n of 7.0 kDa was obtained.¹⁶³ The polymer showed a low-lying HOMO energy level of −5.78 eV, and highly ordered thin-films as revealed by the presence of diffraction peaks up to the fifth order in film XRD patterns. Consequently, P88 exhibited a remarkably high μ_e of 1.61 cm² V⁻¹ s⁻¹ when TGBC device configuration was employed.

Based on the amide containing isoindigo (IID) and thienopyrroledione (TPD), P89 and P90 were synthesized via DArP by Leclerc et al.¹⁵⁶ This was the first report on A–A type polymers made by DArP. After detailed optimization of the reaction conditions including the ligand and base, P89 and P90 with rather high M_ns of 20 and 24 kDa, respectively, were obtained. However, using standard Suzuki polycondensation produced no polymeric product for P89 and low M_n product for P90 (13 kDa), implying the advantages of DArP for the synthesis of A–A type polymers. BGBC OTFT devices with P89 and P90 as semiconductor materials showed unipolar electron transport behaviour, and the maximum μ_e for P89 and P90 were 3.5 × 10⁻³ and 3.0 × 10⁻⁴ cm² V⁻¹ s⁻¹, respectively. Subsequently, A–A type polymers P90-1 and P91 comprising IID-TPD and IID-BT, respectively, were synthesized by Yang et al., using Suzuki polycondensation.¹⁵³ Both polymers showed similar HOMO energy levels of ∼−5.70 eV, while the LUMO energy level of P90-1 (−3.66 eV) was obviously lower than that of P91 (−3.54 eV). This indicates that the TPD unit has a stronger electron-withdrawing ability compared to the BT unit. After annealing at 150 °C, P91 exhibited a μ_e of 0.22 cm² V⁻¹ s⁻¹. On the other hand, about one order of magnitude lower μ_e was obtained for the annealed P90-1 film (∼0.01 cm² V⁻¹ s⁻¹). The poorer device performance of P90-1 is presumably attributed to the smaller grains and lower degree of molecular packing in thin-films. Embedding I–N in the IID unit could significantly improve the planarity of the polymer backbone and hence enhance the molecular packing order in thin-films. Thus, McCulloch et al. synthesized P92 via coupling diazaisoindigo (AIID) with a BT unit.¹⁵⁴ Thanks to the more ordered thin-film and lower energetic disorder enabled by the highly planar backbones, high μ_e of up to 1.0 cm² V⁻¹ s⁻¹ was achieved for the devices based on P92.

Ober et al. reported P93 comprising benzotriazole and BT units,¹⁴⁵ which was synthesized by Suzuki polycondensation between dipotassium bis(trifluoroborate) of 2-alkylbenzotriazole and dibromobenzothiadiazole. By using LiOH as base, which can effectively hydrolyze trifluoroborate to the boronic acid, P93 with a high M_n of up to 120 kDa was obtained. P93 showed a unipolar electron transport characteristic with a μ_e of 0.020 cm² V⁻¹ s⁻¹.

Swager et al. synthesized an A–A type polymer P94 carrying perfluorinated side-chains via Stille polycondensation.¹⁵¹ P94 was only soluble in perfluorinated solvents and insoluble in common organic solvents. This feature would allow for orthogonal processing of devices with multi-layered configurations. P94-based TGBC OTFT devices showed a μ_e of 2.2 × 10⁻⁶ cm² V⁻¹ s⁻¹.

Conclusion and outlook

In this review, we have summarized the recent progress of n-type CPs from the aspect of molecule design. It is clear that controlling the HOMO/LUMO energy levels and orbital distributions of CPs are critical to achieve unipolar n-type transport behaviour. In view of these key points, four molecule design strategies were outlined and discussed in detail, i.e., D–A CPs with strong A units, D–A CPs with weak D units, D–A CPs with dual-A and A–A type CPs. On the basis of energetic consideration, the strategies of “D–A CPs with weak D units” and “A–A type CPs” are highly desired for constructing n-type CPs because they can eliminate the adverse effect of an electron-rich D unit, which tends to elevate the HOMO energy levels of the polymers and thus reduce the hole injection barriers. Additionally, although we classified some polymers into one specific strategy, several polymers were designed by multiple strategies. For instance, P40 is synthesized by the combination of “dual-A” and “weak D” design strategies. In fact, multiple strategies may be more effective than a single strategy in terms of energy level modulation. We hope that the findings summarized in this review will provide some guidelines for further development of high-performance n-type CPs and accelerate their applications in organic electronics.

Although some of the n-type CPs have exhibited comparable performance with their p-type counterparts, the overall development of n-type CPs still lags far behind p-type CPs. Improving the electron mobility is a common theme in the development of n-type CPs because the range of possible applications for OTFT devices increases with the enhancement of mobility. Besides, following challenges also deserve further attention: (1) there are still very few purely unipolar n-type CPs. Many reported n-type CPs also exhibited weak hole transport characteristics as evidenced by the lack of clear off-regime in their transfer curves and the observation of linear increased current in their output curves at the low gate voltage and high source–drain voltage. This undesired hole transport will result in difficult switch off and high power consumption of the devices.^21,105 (2) Most of the high-performance n-type CPs reported so far are only dissolved in chlorinated solvents, such as chloroform and dichlorobenzene, due to their strong intermolecular interactions. Processing with “greener” non-chlorinated solvents is necessary for the practical use of polymeric semiconductors in printing electronics. Therefore, the development of polymeric semiconductors that can be processed with non-chlorinated solvents is still a critical issue. (3) Uniaxial alignment of conjugated polymer backbones in thin-films can significantly enhance the charge carrier mobility of OTFT devices. Several techniques, such as dip coating, off-centre spin coating, shear coating and wire-bar coating, have been reported for the preparation of uniaxially aligned films of CPs. Among them, shear coating and wire-bar coating are highly desirable because they can be adopted in large-scale manufacturing processes. It was found that the alignment of polymer backbones produced by these deposition techniques was closely related to the aggregation of polymer chains in solution.¹⁶⁴ Therefore, rational design of CPs to control the aggregation structures in solution is necessary. In addition, new professional techniques to characterize solution-state supramolecular structures should be developed. In this regard, a recent work reported by Pei et al. provided a guideline.⁸¹ (4) Mass production of high-performance CPs with small batch-to-batch variation is still a great challenge. Most of the reported high mobility CPs are synthesized via traditional polycondensation, such as Stille polycondensation, which requires the preparation of an organotin monomer by using a flammable organolithium reagent and produces a stoichiometric amount of toxic by-products. Furthermore, as we discussed above, polymers with an A–A backbone are highly desirable for the development of n-type CPs. However, A–A type polymers with high M_n, which frequently provide more efficient electron transport, usually cannot be obtained by the traditional polycondensation methods. Therefore, an efficient and environmentally benign synthetic strategy must be developed. DArP is emerging as an alternative protocol to traditional polycondensation methods for the synthesis of CPs.^{157,165–167} However, DArP is greatly restricted by the limited substrate scope and the narrow applicability of reaction conditions.^{157,165–167} It is obvious that there are significant opportunities to develop high-performance n-type CPs by combining synthetic chemistry with molecule design.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21774093) and the National Key R&D Program of “Strategic Advanced Electronic Materials” (No. 2016YFB0401100) of the Chinese Ministry of Science and Technology.

Notes and references

1. A. C. Arias, J. D. MacKenzie, I. McCulloch, J. Rivnay and A. Salleo, Chem. Rev., 2010, 110, 3–24.
2. X. Guo, M. Baumgarten and K. Müllen, Prog. Polym. Sci., 2013, 38, 1832–1908.
3. H. Sirringhaus, Adv. Mater., 2014, 26, 1319–1335.
4. K. Fukuda, Y. Takeda, Y. Yoshimura, R. Shiwaku, L. T. Tran, T. Sekine, M. Mizukami, D. Kumaki and S. Tokito, Nat. Commun., 2014, 5, 4147.
5. X. Gao and Z. Zhao, Sci. China: Chem., 2015, 58, 947–968.
6. L. Shi, Y. Guo, W. Hu and Y. Liu, Mater. Chem. Front., 2017, 1, 2423–2456.
7. I. Osaka and K. Takimiya, Adv. Mater., 2017, 29, 1605218.
8. Z. Liu, G. Zhang and D. Zhang, Acc. Chem. Res., 2018, 51, 1422–1432.
9. J. Yang, Z. Zhao, S. Wang, Y. Guo and Y. Liu, Chem, 2018, 4, 2748–2785.
10. M. Muccini, Nat. Mater., 2006, 5, 605.
11. Y. Wang and T. Michinobu, J. Mater. Chem. C, 2018, 6, 10390–10410.
12. J. Xu, S. Wang, G.-J. N. Wang, C. Zhu, S. Luo, L. Jin, X. Gu, S. Chen, V. R. Feig, J. W. To, S. Rondeau-Gagné, J. Park, B. C. Schroeder, C. Lu, J. Y. Oh, Y. Wang, Y.-H. Kim, H. Yan, R. Sinclair, D. Zhou, G. Xue, B. Murmann, C. Linder, W. Cai, J. B.-H. Tok, J. W. Chung and Z. Bao, Science, 2017, 355, 59–64.
13. C. R. Newman, C. D. Frisbie, D. A. da Silva Filho, J.-L. Brédas, P. C. Ewbank and K. R. Mann, Chem. Mater., 2004, 16, 4436–4451.
14. J. Zaumseil and H. Sirringhaus, Chem. Rev., 2007, 107, 1296–1323.
15. X. Zhao and X. Zhan, Chem. Soc. Rev., 2011, 40, 3728–3743.
16. J. E. Anthony, A. Facchetti, M. Heeney, S. R. Marder and X. Zhan, Adv. Mater., 2010, 22, 3876–3892.
17. Y. Zhao, Y. Guo and Y. Liu, Adv. Mater., 2013, 25, 5372–5391.
18. X. Gao and Y. Hu, J. Mater. Chem. C, 2014, 2, 3099–3117.
19. H. Usta, A. Facchetti and T. J. Marks, Acc. Chem. Res., 2011, 44, 501–510.
20. J. Choi, H. Song, N. Kim and F. S. Kim, Semicond. Sci. Technol., 2015, 30, 064002.
21. J. T. Quinn, J. Zhu, X. Li, J. Wang and Y. Li, J. Mater. Chem. C, 2017, 5, 8654–8681.
22. R. J. Baker, CMOS Circuit Design, Layout, and Simulation, John Wiley & Sons, 2010, pp. 6–7.
23. K. J. Baeg, M. Caironi and Y. Y. Noh, Adv. Mater., 2013, 25, 4210–4244.
24. B. Sun, W. Hong, C. Guo, S. Sutty, H. Aziz and Y. Li, Org. Electron., 2016, 37, 190–196.
25. D. De Leeuw, M. Simenon, A. Brown and R. Einerhand, Synth. Met., 1997, 87, 53–59.
26. Y. Gao, Y. Deng, H. Tian, J. Zhang, D. Yan, Y. Geng and F. Wang, Adv. Mater., 2017, 29, 1606217.
27. Y. Sui, Y. Deng, Y. Han, J. Zhang, W. Hu and Y. Geng, J. Mater. Chem. C, 2018, 6, 12896–12903.
28. L. Zhang, Z. Wang, C. Duan, Z. Wang, Y. Deng, J. Xu, F. Huang and Y. Cao, Chem. Mater., 2018, 30, 8343–8351.
29. F. Chen, Y. Jiang, Y. Sui, J. Zhang, H. Tian, Y. Han, Y. Deng, W. Hu and Y. Geng, Macromolecules, 2018, 51, 8652–8661.
30. H. T. Nicolai, M. Kuik, G. Wetzelaer, B. De Boer, C. Campbell, C. Risko, J. Brédas and P. Blom, Nat. Mater., 2012, 11, 882.
31. G. W. van Pruissen, E. A. Pidko, M. M. Wienk and R. A. Janssen, J. Mater. Chem. C, 2014, 2, 731–735.
32. K. Kawabata, I. Osaka, M. Nakano, N. Takemura, T. Koganezawa and K. Takimiya, Adv. Electron. Mater., 2015, 1, 1500039.
33. K. Kawabata, M. Saito, I. Osaka and K. Takimiya, J. Am. Chem. Soc., 2016, 138, 7725–7732.
34. H. Song, Y. Deng, Y. Gao, Y. Jiang, H. Tian, D. Yan, Y. Geng and F. Wang, Macromolecules, 2017, 50, 2344–2353.
35. X. Lv, W. Li, M. Ouyang, Y. Zhang, D. S. Wright and C. Zhang, J. Mater. Chem. C, 2017, 5, 12–28.
36. Y.-W. Su, Y.-C. Lin and K.-H. Wei, J. Mater. Chem. A, 2017, 5, 24051–24075.
37. Z. Yuan, B. Fu, S. Thomas, S. Zhang, G. DeLuca, R. Chang, L. Lopez, C. Fares, G. Zhang, J.-L. Bredas and E. Reichmanis, Chem. Mater., 2016, 28, 6045–6049.
38. T. Lei, Y. Cao, X. Zhou, Y. Peng, J. Bian and J. Pei, Chem. Mater., 2012, 24, 1762–1770.
39. T. Lei, J.-H. Dou, Z.-J. Ma, C.-H. Yao, C.-J. Liu, J.-Y. Wang and J. Pei, J. Am. Chem. Soc., 2012, 134, 20025–20028.
40. H. Zhou, L. Yang and W. You, Macromolecules, 2012, 45, 607–632.
41. R. S. Ashraf, A. J. Kronemeijer, D. I. James, H. Sirringhaus and I. McCulloch, Chem. Commun., 2012, 48, 3939–3941.
42. G. K. Dutta, A. R. Han, J. Lee, Y. Kim, J. H. Oh and C. Yang, Adv. Funct. Mater., 2013, 23, 5317–5325.
43. G. Kim, A.-R. Han, H. R. Lee, J. Lee, J. H. Oh and C. Yang, Chem. Commun., 2014, 50, 2180–2183.
44. A. Babel and S. A. Jenekhe, J. Am. Chem. Soc., 2003, 125, 13656–13657.
45. Y. Wang, T. Hasegawa, H. Matsumoto and T. Michinobu, J. Am. Chem. Soc., 2019, 141, 3566–3575.
46. X. Guo and M. D. Watson, Org. Lett., 2008, 10, 5333–5336.
47. Z. Chen, Y. Zheng, H. Yan and A. Facchetti, J. Am. Chem. Soc., 2009, 131, 8–9.
48. H. Yan, Z. Chen, Y. Zheng, C. Newman, J. R. Quinn, F. Dötz, M. Kastler and A. Facchetti, Nature, 2009, 457, 679–687.
49. H. Opoku, B. Nketia-Yawson, E. S. Shin and Y.-Y. Noh, Org. Electron., 2017, 41, 198–204.
50. X. Guo, A. Facchetti and T. J. Marks, Chem. Rev., 2014, 114, 8943–9021.
51. Y. Kim, D. X. Long, J. Lee, G. Kim, T. J. Shin, K.-W. Nam, Y.-Y. Noh and C. Yang, Macromolecules, 2015, 48, 5179–5187.
52. B. Kang, R. Kim, S. B. Lee, S.-K. Kwon, Y.-H. Kim and K. Cho, J. Am. Chem. Soc., 2016, 138, 3679–3686.
53. F. So, World Scientific Handbook of Organic Optoelectronic Devices (Volumes 1 & 2): Volume 1: Perovskite Electronics Volume 2: Organic Photovoltaics (OPVs), World Scientific, 2018.
54. Y.-J. Hwang, G. Ren, N. M. Murari and S. A. Jenekhe, Macromolecules, 2012, 45, 9056–9062.
55. H. Chen, Y. Guo, G. Yu, Y. Zhao, J. Zhang, D. Gao, H. Liu and Y. Liu, Adv. Mater., 2012, 24, 4618–4622.
56. I. Kang, T. K. An, J. A. Hong, H. J. Yun, R. Kim, D. S. Chung, C. E. Park, Y. H. Kim and S. K. Kwon, Adv. Mater., 2013, 25, 524–528.
57. R. Kim, P. S. K. Amegadze, I. Kang, H. J. Yun, Y. Y. Noh, S. K. Kwon and Y. H. Kim, Adv. Funct. Mater., 2013, 23, 5719–5727.
58. M. J. Sung, A. Luzio, W. T. Park, R. Kim, E. Gann, F. Maddalena, G. Pace, Y. Xu, D. Natali, C. de Falco, L. Dang, C. R. McNeill, M. Caironi, Y. Y. Noh and Y. H. Kim, Adv. Funct. Mater., 2016, 26, 4984–4997.
59. H. Chen, Y. Guo, Z. Mao, G. Yu, J. Huang, Y. Zhao and Y. Liu, Chem. Mater., 2013, 25, 3589–3596.
60. D. Fazzi, M. Caironi and C. Castiglioni, J. Am. Chem. Soc., 2011, 133, 19056–19059.
61. T. Schuettfort, S. Huettner, S. Lilliu, J. E. Macdonald, L. Thomsen and C. R. McNeill, Macromolecules, 2011, 44, 1530–1539.
62. 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, 18303–18317.
63. Z. Fei, Y. Han, J. Martin, F. H. Scholes, M. Al-Hashimi, S. Y. AlQaradawi, N. Stingelin, T. D. Anthopoulos and M. Heeney, Macromolecules, 2016, 49, 6384–6393.
64. T. Sajoto, S. P. Tiwari, H. Li, C. Risko, S. Barlow, Q. Zhang, J.-Y. Cho, J.-L. Brédas, B. Kippelen and S. R. Marder, Polymer, 2012, 53, 1072–1078.
65. M. He, W. Li, Y. Gao, H. Tian, J. Zhang, H. Tong, D. Yan, Y. Geng and F. Wang, Macromolecules, 2016, 49, 825–832.
66. Y. Deng, Y. Chen, X. Zhang, H. Tian, C. Bao, D. Yan, Y. Geng and F. Wang, Macromolecules, 2012, 45, 8621–8627.
67. C.-E. Tsai, R.-H. Yu, F.-J. Lin, Y.-Y. Lai, J.-Y. Hsu, S.-W. Cheng, C.-S. Hsu and Y.-J. Cheng, Chem. Mater., 2016, 28, 5121–5130.
68. P. Prins, F. Grozema, J. Schins, S. Patil, U. Scherf and L. Siebbeles, Phys. Rev. Lett., 2006, 96, 146601.
69. M. Yuan, M. M. Durban, P. D. Kazarinoff, D. F. Zeigler, A. H. Rice, Y. Segawa and C. K. Luscombe, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 4061–4069.
70. Z. Zhao, F. Zhang, Y. Hu, Z. Wang, B. Leng, X. Gao, C.-A. Di and D. Zhu, ACS Macro Lett., 2014, 3, 1174–1177.
71. X. Zhan, Z. a. Tan, B. Domercq, Z. An, X. Zhang, S. Barlow, Y. Li, D. Zhu, B. Kippelen and S. R. Marder, J. Am. Chem. Soc., 2007, 129, 7246–7247.
72. L. Zhang, C. a. Di, Y. Zhao, Y. Guo, X. Sun, Y. Wen, W. Zhou, X. Zhan, G. Yu and Y. Liu, Adv. Mater., 2010, 22, 3537–3541.
73. X. Zhao, L. Ma, L. Zhang, Y. Wen, J. Chen, Z. Shuai, Y. Liu and X. Zhan, Macromolecules, 2013, 46, 2152–2158.
74. W. Zhou, Y. Wen, L. Ma, Y. Liu and X. Zhan, Macromolecules, 2012, 45, 4115–4121.
75. S. K. Samanta, I. Song, J. H. Yoo and J. H. Oh, ACS Appl. Mater. Interfaces, 2018, 10, 32444–32453.
76. Z. Yan, B. Sun and Y. Li, Chem. Commun., 2013, 49, 3790–3792.
77. T. Lei, J.-H. Dou, X.-Y. Cao, J.-Y. Wang and J. Pei, J. Am. Chem. Soc., 2013, 135, 12168–12171.
78. Y. Wang, S. D. Motta, F. Negri and R. Friedlein, J. Am. Chem. Soc., 2011, 133, 10054–10057.
79. J. H. Dou, Y. Q. Zheng, T. Lei, S. D. Zhang, Z. Wang, W. B. Zhang, J. Y. Wang and J. Pei, Adv. Funct. Mater., 2014, 24, 6270–6278.
80. T. Lei, J. H. Dou, X. Y. Cao, J. Y. Wang and J. Pei, Adv. Mater., 2013, 25, 6589–6593.
81. Y. Q. Zheng, Z. F. Yao, T. Lei, J. H. Dou, C. Y. Yang, L. Zou, X. Meng, W. Ma, J. Y. Wang and J. Pei, Adv. Mater., 2017, 29, 1701072.
82. G. Zhang, J. Guo, M. Zhu, P. Li, H. Lu, K. Cho and L. Qiu, Polym. Chem., 2015, 6, 2531–2540.
83. X. Zhou, N. Ai, Z.-H. Guo, F.-D. Zhuang, Y.-S. Jiang, J.-Y. Wang and J. Pei, Chem. Mater., 2015, 27, 1815–1820.
84. T. Lei, X. Xia, J.-Y. Wang, C.-J. Liu and J. Pei, J. Am. Chem. Soc., 2014, 136, 2135–2141.
85. Y. Q. Zheng, T. Lei, J. H. Dou, X. Xia, J. Y. Wang, C. J. Liu and J. Pei, Adv. Mater., 2016, 28, 7213–7219.
86. Y.-Z. Dai, N. Ai, Y. Lu, Y.-Q. Zheng, J.-H. Dou, K. Shi, T. Lei, J.-Y. Wang and J. Pei, Chem. Sci., 2016, 7, 5753–5757.
87. B. Sun, W. Hong, Z. Yan, H. Aziz and Y. Li, Adv. Mater., 2014, 26, 2636–2642.
88. Y. Kim, H. Hwang, N. K. Kim, K. Hwang, J. J. Park, G. I. Shin and D. Y. Kim, Adv. Mater., 2018, 30, 1706557.
89. W. Hong, C. Guo, Y. Li, Y. Zheng, C. Huang, S. Lu and A. Facchetti, J. Mater. Chem., 2012, 22, 22282–22289.
90. J. W. Rumer, M. Levick, S.-Y. Dai, S. Rossbauer, Z. Huang, L. Biniek, T. D. Anthopoulos, J. R. Durrant, D. J. Procter and I. McCulloch, Chem. Commun., 2013, 49, 4465–4467.
91. W. Cui and F. Wudl, Macromolecules, 2013, 46, 7232–7238.
92. H. Hwang, D. Khim, J. M. Yun, E. Jung, S. Y. Jang, Y. H. Jang, Y. Y. Noh and D. Y. Kim, Adv. Funct. Mater., 2015, 25, 1146–1156.
93. Y. Deng, B. Sun, Y. He, J. Quinn, C. Guo and Y. Li, Angew. Chem., Int. Ed., 2016, 55, 3459–3462.
94. X. Zhang, H. Bronstein, A. J. Kronemeijer, J. Smith, Y. Kim, R. J. Kline, L. J. Richter, T. D. Anthopoulos, H. Sirringhaus, K. Song, M. Heeney, W. Zhang, I. McCulloch and D. M. DeLongchamp, Nat. Commun., 2013, 4, 2238.
95. Y. Deng, J. Quinn, B. Sun, Y. He, J. Ellard and Y. Li, RSC Adv., 2016, 6, 34849–34854.
96. Y. Gao, X. Zhang, H. Tian, J. Zhang, D. Yan, Y. Geng and F. Wang, Adv. Mater., 2015, 27, 6753–6759.
97. H. Huang, L. Yang, A. Facchetti and T. J. Marks, Chem. Rev., 2017, 117, 10291–10318.
98. C. J. Mueller, C. R. Singh, M. Fried, S. Huettner and M. Thelakkat, Adv. Funct. Mater., 2015, 25, 2725–2736.
99. H.-Y. Chen, M. Nikolka, A. Wadsworth, W. Yue, A. Onwubiko, M. Xiao, A. J. P. White, D. Baran, H. Sirringhaus and I. McCulloch, Macromolecules, 2017, 51, 71–79.
100. Y. Wang, Z. Yan, H. Guo, M. A. Uddin, S. Ling, X. Zhou, H. Su, J. Dai, H. Y. Woo and X. Guo, Angew. Chem., Int. Ed., 2017, 56, 15304–15308.
101. Y. Shi, H. Guo, M. Qin, Y. Wang, J. Zhao, H. Sun, H. Wang, Y. Wang, X. Zhou, A. Facchetti, X. Lu, M. Zhou and X. Guo, Chem. Mater., 2018, 30, 7988–8001.
102. Y. Ie, M. Nitani, M. Karakawa, H. Tada and Y. Aso, Adv. Funct. Mater., 2010, 20, 907–913.
103. I. Osaka, R. Zhang, J. Liu, D.-M. Smilgies, T. Kowalewski and R. D. McCullough, Chem. Mater., 2010, 22, 4191–4196.
104. I. Kang, H.-J. Yun, D. S. Chung, S.-K. Kwon and Y.-H. Kim, J. Am. Chem. Soc., 2013, 135, 14896–14899.
105. K. Guo, J. Bai, Y. Jiang, Z. Wang, Y. Sui, Y. Deng, Y. Han, H. Tian and Y. Geng, Adv. Funct. Mater., 2018, 28, 1801097.
106. J. Chen, X. Zhang, G. Wang, M. A. Uddin, Y. Tang, Y. Wang, Q. Liao, A. Facchetti, T. J. Marks and X. Guo, J. Mater. Chem. C, 2017, 5, 9559–9569.
107. F. P. Wu, H. I. Un, Y. Li, H. Hu, Y. Yuan, B. Yang, K. Xiao, W. Chen, J. Y. Wang, Z. Q. Jiang, J. Pei and L. S. Liao, Chem. – Eur. J., 2017, 23, 14723–14727.
108. F. Wang, Y. Dai, W. Wang, H. Lu, L. Qiu, Y. Ding and G. Zhang, Chem. Mater., 2018, 30, 5451–5459.
109. A. Luzio, D. Fazzi, F. Nübling, R. Matsidik, A. Straub, H. Komber, E. Giussani, S. E. Watkins, M. Barbatti, W. Thiel, E. Gann, L. Thomsen, C. R. McNeill, M. Caironi and M. Sommer, Chem. Mater., 2014, 26, 6233–6240.
110. K. Kranthiraja, D. X. Long, V. G. Sree, W. Cho, Y.-R. Cho, A. Zaheer, J.-C. Lee, Y.-Y. Noh and S.-H. Jin, Macromolecules, 2018, 51, 5530–5536.
111. H. Xin, C. Ge, X. Jiao, X. Yang, K. Rundel, C. R. McNeill and X. Gao, Angew. Chem., Int. Ed., 2018, 57, 1322–1326.
112. Y. Wang, A. T.-R. Tan, T. Mori and T. Michinobu, J. Mater. Chem. C, 2018, 6, 3593–3603.
113. H. S. Kim, G. Huseynova, Y.-Y. Noh and D.-H. Hwang, Macromolecules, 2017, 50, 7550–7558.
114. H. Usta, W. C. Sheets, M. Denti, G. Generali, R. Capelli, S. Lu, X. Yu, M. Muccini and A. Facchetti, Chem. Mater., 2014, 26, 6542–6556.
115. B. Fu, C.-Y. Wang, B. D. Rose, Y. Jiang, M. Chang, P.-H. Chu, Z. Yuan, C. Fuentes-Hernandez, B. Kippelen, J.-L. Bredas, D. M. Collard and E. Reichmanis, Chem. Mater., 2015, 27, 2928–2937.
116. C. Guo, J. Quinn, B. Sun and Y. Li, Polym. Chem., 2016, 7, 4515–4524.
117. Y. Li, P. Sonar, S. P. Singh, M. S. Soh, M. van Meurs and J. Tan, J. Am. Chem. Soc., 2011, 133, 2198–2204.
118. H. N. Hong, H. J. Kim, A. Kim, S. Choi, Y. U. Kim, M. J. Cho and D. H. Choi, Synth. Met., 2018, 236, 1–7.
119. S. Shi, Y. Wang, M. A. Uddin, X. Zhou, H. Guo, Q. Liao, X. Zhu, X. Cheng, H. Y. Woo and X. Guo, Adv. Electron. Mater., 2017, 3, 1700100.
120. S. Ma, G. Zhang, F. Wang, Y. Dai, H. Lu, L. Qiu, Y. Ding and K. Cho, Macromolecules, 2018, 51, 5704–5712.
121. Z. Yuan, C. Buckley, S. Thomas, G. Zhang, I. Bargigia, G. Wang, B. Fu, C. Silva, J.-L. Brédas and E. Reichmanis, Macromolecules, 2018, 51, 7320–7328.
122. R. Matsidik, M. Giorgio, A. Luzio, M. Caironi, H. Komber and M. Sommer, Eur. J. Org. Chem., 2018, 6121–6126.
123. M. Saito, I. Osaka, Y. Suda, H. Yoshida and K. Takimiya, Adv. Mater., 2016, 28, 6921–6925.
124. C. Zhu, Z. Zhao, H. Chen, L. Zheng, X. Li, J. Chen, Y. Sun, F. Liu, Y. Guo and Y. Liu, J. Am. Chem. Soc., 2017, 139, 17735–17738.
125. J. Yang, H. Wang, J. Chen, J. Huang, Y. Jiang, J. Zhang, L. Shi, Y. Sun, Z. Wei, G. Yu, Y. Guo, S. Wang and Y. Liu, Adv. Mater., 2017, 29, 1606162.
126. L. Xu, Z. Zhao, M. Xiao, J. Yang, J. Xiao, Z. Yi, S. Wang and Y. Liu, ACS Appl. Mater. Interfaces, 2017, 9, 40549–40555.
127. J. Rumer, S. Rossbauer, M. Planells, S. Watkins, T. Anthopoulos and I. McCulloch, J. Mater. Chem. C, 2014, 2, 8822–8828.
128. Z. A. Wang, J. Kuwabara, A. Ichige, T. Yasuda and T. Kanbara, Synth. Met., 2016, 222, 383–387.
129. Z. Zhao, Z. Yin, H. Chen, L. Zheng, C. Zhu, L. Zhang, S. Tan, H. Wang, Y. Guo, Q. Tang and Y. Liu, Adv. Mater., 2017, 29, 1602410.
130. B. R. Conrad, C. K. Chan, M. A. Loth, S. R. Parkin, X. Zhang, D. M. Delongchamp, J. E. Anthony and D. J. Gundlach, Appl. Phys. Lett., 2010, 97, 133306.
131. K. Kawashima, I. Osaka and K. Takimiya, Chem. Mater., 2015, 27, 6558–6570.
132. 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, 1314–1321.
133. Y. Wang, T. Hasegawa, H. Matsumoto, T. Mori and T. Michinobu, Adv. Funct. Mater., 2017, 27, 1701486.
134. Y. Wang, T. Hasegawa, H. Matsumoto, T. Mori and T. Michinobu, Adv. Mater., 2018, 30, 1707164.
135. D. M. Hodgson, J. Witherington, B. A. Moloney, I. C. Richards and J.-L. Brayer, Synlett, 1995, 32–34.
136. T. Lei, J. Y. Wang and J. Pei, Chem. Mater., 2013, 26, 594–603.
137. J. Mei and Z. Bao, Chem. Mater., 2014, 26, 604–615.
138. S. Park, M. H. Lee, K. S. Ahn, H. H. Choi, J. Shin, J. Xu, J. Mei, K. Cho, Z. Bao, D. R. Lee, M. S. Kang and D. H. Kim, Adv. Funct. Mater., 2016, 26, 4627–4634.
139. G. Xue, X. Zhao, G. Qu, T. Xu, A. Gumyusenge, Z. Zhang, Y. Zhao, Y. Diao, H. Li and J. Mei, ACS Appl. Mater. Interfaces, 2017, 9, 25426–25433.
140. J. Lee, A. R. Han, H. Yu, T. J. Shin, C. Yang and J. H. Oh, J. Am. Chem. Soc., 2013, 135, 9540–9547.
141. S. Fabiano, H. Yoshida, Z. Chen, A. Facchetti and M. A. Loi, ACS Appl. Mater. Interfaces, 2013, 5, 4417–4422.
142. 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, 18303–18317.
143. H. N. Tsao, D. M. Cho, I. Park, M. R. Hansen, A. Mavrinskiy, D. Y. Yoon, R. Graf, W. Pisula, H. W. Spiess and K. Müllen, J. Am. Chem. Soc., 2011, 133, 2605–2612.
144. K. Guo, Y. Jiang, Y. Sui, Y. Deng and Y. Geng, Chin. J. Polym. Sci., 2019 DOI:10.1007/s10118-019-2277-1.
145. J.-K. Lee, M. C. Gwinner, R. Berger, C. Newby, R. Zentel, R. H. Friend, H. Sirringhaus and C. K. Ober, J. Am. Chem. Soc., 2011, 133, 9949–9951.
146. M. M. Durban, P. D. Kazarinoff and C. K. Luscombe, Macromolecules, 2010, 43, 6348–6352.
147. J. A. Letizia, M. R. Salata, C. M. Tribout, A. Facchetti, M. A. Ratner and T. J. Marks, J. Am. Chem. Soc., 2008, 130, 9679–9694.
148. X. Guo, R. P. Ortiz, Y. Zheng, Y. Hu, Y. Y. Noh, K. J. Baeg, A. Facchetti and T. J. Marks, J. Am. Chem. Soc., 2011, 133, 1405–1418.
149. Y. Wang, H. Guo, A. Harbuzaru, M. A. Uddin, I. Arrechea-Marcos, S. Ling, J. Yu, Y. Tang, H. Sun, J. T. López Navarrete, R. P. Ortiz, H. Y. Woo and X. Guo, J. Am. Chem. Soc., 2018, 140, 6095–6108.
150. Y. Shi, H. Guo, M. Qin, J. Zhao, Y. Wang, H. Wang, Y. Wang, A. Facchetti, X. Lu and X. Guo, Adv. Mater., 2018, 30, 1705745.
151. Y. Takeda, T. L. Andrew, J. M. Lobez, A. J. Mork and T. M. Swager, Angew. Chem., Int. Ed., 2012, 51, 9042–9046.
152. S. Vasimalla, S. P. Senanayak, M. Sharma, K. S. Narayan and P. K. Iyer, Chem. Mater., 2014, 26, 4030–4037.
153. G. Kim, A.-R. Han, H. R. Lee, J. Lee, J. H. Oh and C. Yang, Chem. Commun., 2014, 50, 2180–2183.
154. W. Yue, M. Nikolka, M. Xiao, A. Sadhanala, C. B. Nielsen, A. J. P. White, H. Y. Chen, A. Onwubiko, H. Sirringhaus and I. Mcculloch, J. Mater. Chem. C, 2016, 4, 9704–9710.
155. R. Stalder, J. Mei, J. Subbiah, C. Grand, L. A. Estrada, F. So and J. R. Reynolds, Macromolecules, 2011, 44, 6303–6310.
156. F. Grenier, P. Berrouard, J. R. Pouliot, H. R. Tseng, A. J. Heeger and M. Leclerc, Polym. Chem., 2013, 4, 1836–1841.
157. J. R. Pouliot, F. Grenier, J. T. Blaskovits, S. Beaupré and M. Leclerc, Chem. Rev., 2016, 116, 14225–14274.
158. Y. Fukutomi, M. Nakano, J.-Y. Hu, I. Osaka and K. Takimiya, J. Am. Chem. Soc., 2013, 135, 11445–11448.
159. M. Nakano, I. Osaka and K. Takimiya, Macromolecules, 2015, 48, 576–584.
160. W. Yang, M. Nakano, T. Michinobu, Y. Kiyota, T. Mori and K. Takimiya, Macromolecules, 2017, 50, 857–864.
161. X. Zhao, Y. Wen, L. Ren, L. Ma, Y. Liu and X. Zhan, J. Polym. Sci., Part A: Polym. Chem., 2012, 50, 4266–4271.
162. Y. Wang, H. Guo, S. Ling, I. Arrechea-Marcos, Y. Wang, J. T. López Navarrete, R. P. Ortiz and X. Guo, Angew. Chem., Int. Ed., 2017, 56, 9924–9929.
163. Y. Shi, H. Guo, M. Qin, J. Zhao, Y. Wang, H. Wang, Y. Wang, A. Facchetti, X. Lu and X. Guo, Adv. Mater., 2018, 30, 1705745.
164. D. Khim, A. Luzio, G. E. Bonacchini, G. Pace, M. J. Lee, Y. Y. Noh and M. Caironi, Adv. Mater., 2018, 30, 1705463.
165. T. Bura, J. T. Blaskovits and M. Leclerc, J. Am. Chem. Soc., 2016, 138, 10056.
166. H. Bohra and M. Wang, J. Mater. Chem. A, 2017, 5, 11550–11571.
167. L. G. Mercier and M. Leclerc, Acc. Chem. Res., 2013, 46, 1597–1605.

---