Organic thermoelectric (OTE) materials that can convert waste heat to electricity have aroused interests due to their unique advantages over traditional inorganic TE materials, such as light weight, mechanical flexibility, low thermal conductivity, and solution processability[1−4]. In general, TE devices require both p-type and n-type semiconductors. The p-type polymers have been extensively studied, showing rapid advances, but there are few efficient n-type TE polymers[5, 6]. Therefore, the development of high-performance ndoped conjugated polymers is demanded. The TE performance is evaluated by the figure of merit, ZT = S2σT/κ, where S, σ, T, and κ are the Seebeck coefficient, electrical conductivity, absolute temperature, and thermal conductivity, respectively. As the κ values of polymers are much lower than that of inorganic materials, the TE performance of polymers can also be determined by the power factor (PF = S2σ)[7]. Thus, enhancing σ and S is the key to improve TE performance. The inferior performance for n-type OTE materials is mainly due to their low σ, so we focus on the σ issue in this article. To enhance the conductivity, some strategies can be applied, which will be discussed as follows. Lowering LUMO energy level is an effective approach to improve n-doping[8−10]. Introducing strong electron-withdrawing groups or atoms to the backbone can lower the LUMO level[11−13]. The D–A copolymer P(NDI2OD-T2) has deep-lying LUMO level (–3.80 eV). When doped with n-DMBI, a conductivity of ~10–3 S/cm was achieved[7]. To further down-shift LUMO level, Facchetti et al. designed polymer P(NDI2OD-Tz2) (Fig. 1)[14]. By introducing bithiazole unit, the polymer possesses a more planar backbone than N2200, resulting in a close π–π stacking. The electron-deficient nature of bithiazole enhances electron affinity of the polymer, yielding an enhanced σ of 0.1 S/cm and a reasonable PF of 1.5 μW/(m·K2) (Table 1). To reduce steric hindrance of NDI, thiophene-fused NDI derivative, naphtho[2,3-b:6,7-bʹ]dithiophenediimide (NDTI), was developed by Takimiya et al. Then they developed a polymer PNDTI-BBT-DP with strong electron affinity. It has a low LUMO level (~ –4.4 eV), which is sufficiently low for being doped by n-DMBI. The doped film offered a σ of 5.0 S/cm and a PF of 14 μW/(m·K2)[15]. Recently, Wang et al. reported PNB-TzDP that offered an excellent σ of 11.6 S/cm and a PF of 53.4 μW/(m·K2)[16]. Another strong electron-accepting unit BDOPV was developed by Pei et al., and the derivative polymers have low LUMO levels and have been investigated in various devices[17]. Among them, FBDPPV delivered a high σ of 14 S/cm and a PF of 28 μW/(m·K2). Subsequently, a σ over 90 S/cm was obtained from TBDPPV polymer doped with n-DMBI[18, 19]. Guo et al. synthesized thiazolothienyl imide dimer (DTzTI) unit by replacing thiophene with thiazole to further push down LUMO level. PDTzTI was studied in OTFT[20, 21]. When doped with TDAE, a σ of 4.6 S/cm and a PF of 7.6 μW/(m·K2) were obtained[22]. PCNI-BTI was developed, offering a σ of 23.3 S/cm and a PF of 10 μW/(m·K2)[23]. B←N coordination bonds show electron-withdrawing properties, gifting polymers with low LUMO levels[24]. Liu et al. reported a polymer PBN-19 with BNBP unit. After n-doping, PBN-19 exhibited a σ of 7.8 S/cm and a PF of 24.8 μW/(m·K2)[25]. Introducing polar triethylene glycol (TEG) side chains into polymers can improve the miscibility between dopant and polymer. Liu et al. found that the σ and PF of TEG-N2200 can be increased by a factor of 200 after replacing alkyl side chains of N2200 with TEG side chains[26]. It delivered a σ of 0.17 S/cm and a PF of 0.4 μW/(m·K2) (Table 1) after being doped with n-DMBI. They also designed polymer PNDI2TEG2Tz by replacing thiophene with thiazole unit, and the doped material showed a higher σ of 1.8 S/cm and a higher PF of 4.5 μW/(m·K2) as compared with N2200[27]. Similar methods were also used by other groups[28]. In short, we discussed the strategies of lowering LUMO energy level and incorporating polar side chains for making high-performance n-type OTE materials. More efforts should be focused on molecular engineering.

中文翻译：

---

提高n型有机热电材料的性能

可以将废热转化为电能的有机热电（OTE）材料由于其相对于传统无机热电材料的独特优势而引起了人们的兴趣，例如重量轻、机械柔性、导热率低和溶液可加工性等[1-4]。通常，TE 器件需要 p 型和 n 型半导体。p型聚合物已被广泛研究，显示出快速的进展，但很少有有效的n型TE聚合物[5, 6]。因此，需要开发高性能的n型掺杂共轭聚合物。TE 性能通过品质因数 ZT = S2σT/κ 来评估，其中 S、σ、T 和 κ 分别是塞贝克系数、电导率、绝对温度和热导率。由于聚合物的 κ 值远低于无机材料，聚合物的热电性能也可以通过功率因数（PF = S2σ）[7]来确定。因此，提高 σ 和 S 是提高 TE 性能的关键。n 型 OTE 材料的性能较差主要是由于它们的低 σ，因此我们在本文中关注 σ 问题。为了提高电导率，可以应用一些策略，这些策略将在下面讨论。降低 LUMO 能级是改善 n 掺杂的有效途径[8-10]。在主链中引入强吸电子基团或原子可以降低 LUMO 能级[11-13]。D-A 共聚物 P(NDI2OD-T2) 具有深层 LUMO 能级 (–3.80 eV)。当掺杂 n-DMBI 时，可实现约 10–3 S/cm 的电导率[7]。为了进一步降低 LUMO 水平，Facchetti 等人。设计的聚合物 P(NDI2OD-Tz2)（图 1）[14]。通过引入联噻唑单元，该聚合物具有比 N2200 更平坦的主链，从而导致紧密的 π-π 堆叠。联噻唑的缺电子特性增强了聚合物的电子亲和力，产生了 0.1 S/cm 的增强 σ 和 1.5 μW/(m·K2) 的合理 PF（表 1）。为了降低 NDI 的空间位阻，Takimiya 等人开发了噻吩稠合 NDI 衍生物萘并[2,3-b:6,7-bʹ]二噻吩二亚胺 (NDTI)。然后他们开发了一种具有强电子亲和力的聚合物 PNDTI-BBT-DP。它具有低 LUMO 能级 (~ –4.4 eV)，足以被 n-DMBI 掺杂。掺杂薄膜的 σ 为 5.0 S/cm，PF 为 14 μW/(m·K2)[15]。最近，王等人。报道了 PNB-TzDP 提供了 11.6 S/cm 的出色 σ 和 53.4 μW/(m·K2)[16] 的 PF。Pei 等人开发了另一种强电子接受单元 BDOPV。并且衍生聚合物具有低 LUMO 水平，并已在各种设备中进行了研究[17]。其中，FBDPPV 具有 14 S/cm 的高 σ 和 28 μW/(m·K2) 的 PF。随后，从掺杂n-DMBI的TBDPPV聚合物中获得了超过90 S/cm的σ[18, 19]。郭等人。通过用噻唑代替噻吩合成噻唑噻吩酰亚胺二聚体 (DTzTI) 单元以进一步降低 LUMO 水平。在 OTFT[20, 21] 中研究了 PDTzTI。掺杂 TDAE 后，σ 为 4.6 S/cm，PF 为 7.6 μW/(m·K2)[22]。开发了 PCNI-BTI，提供 23.3 S/cm 的 σ 和 10 μW/(m·K2)[23] 的 PF。B←N 配位键表现出吸电子特性，赋予聚合物低 LUMO 水平[24]。刘等人。报道了具有BNBP单元的聚合物PBN-19。在n掺杂后，PBN-19的σ为7.8 S/cm，PF为24.8 μW/(m·K2)[25]。在聚合物中引入极性三甘醇 (TEG) 侧链可以改善掺杂剂和聚合物之间的混溶性。刘等人。发现用 TEG 侧链代替 N2200 的烷基侧链后，TEG-N2200 的 σ 和 PF 可以提高 200 倍[26]。在掺杂 n-DMBI 后，它的 σ 为 0.17 S/cm，PF 为 0.4 μW/(m·K2)（表 1）。他们还通过用噻唑单元代替噻吩设计了聚合物PNDI2TEG2Tz，与N2200相比，掺杂材料的σ为1.8 S/cm，PF为4.5 μW/(m·K2)[27]。其他团体也使用了类似的方法[28]。简而言之，我们讨论了降低 LUMO 能级和结合极性侧链以制造高性能 n 型 OTE 材料的策略。更多的努力应该集中在分子工程上。