Introduction

Conjugated polymers are expected to serve as the main materials in optoelectronic devices such as organic photovoltaics (OPVs) [1], organic light-emitting diodes (OLEDs) [2], and organic field-effect transistors (OFETs) [3,4,5]. High-performance polymers have been developed by logical molecular designs [6, 7] and the development of conjugated polymeric materials will allow practical applications in organic optoelectronic devices. For use in wide practical applications, mass production of conjugated polymeric materials is required, while also satisfying demands concerning cost and environmental issues [8, 9]. As a practical method for the production of conjugated polymeric materials, polycondensation using C–H direct arylation [10,11,12] is a promising candidate, because this method does not require the use of organometallic monomers that are essential in polycondensation methods that use conventional cross-coupling reactions. The elimination of organometallic monomers reduces the number of synthetic steps for monomer preparation and undesirable metal-containing waste after polymerization. Having fewer synthetic steps, the process is less expensive and the reduction in waste reduces the environmental burden and increases the purity of the product [8, 9]. To take advantage of this approach in material production, direct arylation polycondensation must fulfill the following three requirements [13]:

1. High molecular weight: A synthetic method should afford a high-molecular-weight polymer that, in general, exhibits better performance regarding carrier mobility and photoelectric conversion [14,15,16,17]. To achieve a high molecular weight, the conversion of the coupling reaction must be high.

2. High selectivity: Selectivity in bond formation must be nearly perfect, because structural defects in polymers cannot be removed, even by purification. The structural defects act as carrier trapping sites, thereby lowering device performance [18, 19].

3. High purity: Impurities from metal catalysts or byproducts should be easily removed from the polymeric materials, because impurities lower the initial performance of the polymer and long-term stability [20, 21]. A reduced amount of catalyst and easily separable byproducts are required for obtaining high-purity materials through simple purification processes. Incorporation of decomposed phosphine ligands at chain ends [22] should be avoided, because the terminal defect may act as an impurity, which cannot be removed in purification steps.

In addition to satisfying these demands, the synthetic method must be broadly applicable for preparation of state-of-the-art polymeric materials. As several review papers on direct arylation polycondensation have been published in recent years [23,24,25], this review describes the recent results of the author and the author’s collaborators.

Initial development of direct arylation polycondensation

Direct arylation polycondensation of alkylthiophenes was reported in 1999 [26]. Although the idea was quite innovative, the polymerization results did not satisfy the demands mentioned above; the molecular weight and selectivity of the bond formation (head-to-tail content) were somewhat lower than those obtained using conventional methods. In 2010, Ozawa and colleagues [27] reported the successful synthesis of poly(3-hexylthiophene) (P3HT) via Pd-catalyzed direct arylation (Scheme 1). This method afforded high-molecular-weight P3HT and a high head-to-tail content, as well as a good yield. Mori and colleagues [28, 29] reported on the synthesis of P3HT using the same monomer and a Knochel–Hauser base by formation of a Grignard reagent, which afforded high head-to-tail content, a controlled molecular weight, and a narrow molecular weight distribution.

Scheme 1
scheme 1

The first successful direct arylation polycondensation for the preparation of P3HT

Shortly after Ozawa and Mori’s report, we reported on the direct arylation polycondensation of tetrafluorobenzene with a dibromofluorene derivative (Scheme 2) [30]. The reaction with Pd(OAc)2 and PtBu2Me·HBF4 produced the corresponding polymer with a high molecular weight in a good yield. The molecular weight and the yield were higher than those obtained for the polymer prepared by polycondensation using the conventional Suzuki–Miyaura coupling reaction (Mn = 16500, 74% yield) [31]. In addition, the high purity of the obtained polymer was confirmed by elemental analysis. The high purity was presumably due to the easily separable byproducts (H2O, CO2, and KBr) in the direct arylation polycondensation reaction. In addition, the pure polymer acted as a hole-blocking material in OLEDs [32]. These initial examples demonstrate the high potential of direct arylation polycondensation for the preparation of optoelectronic materials [33].

Scheme 2
scheme 2

Direct arylation polycondensation of tetrafluorobenzene

Direct arylation polycondensation of bithiophenes

We have investigated the direct arylation polycondensation of thiophene, bithiophene, and their analogs, because thiophene is the most promising unit for preparing excellent semiconducting materials. An initial assessment of catalytic systems revealed that a Pd precatalyst (Pd(OAc)2) with no phosphine ligand is the most effective catalyst for the direct arylation polycondensation of alkylated bithiophenes (Scheme 3). This highly active catalytic system allows the preparation of the corresponding polymer with a high molecular weight (Mn = 31800) in 3 h [34]. In addition to bithiophene derivatives, this reaction system can be used for the polycondensation of bithiazole, thienothiophene, monothiophene derivatives, and naphthalene diimide-based monomers (Scheme 3, Table S1 in Supplementary Information) [35,36,37,38,39,40,41,42,43]. This simple catalytic system has been used for the preparation of a variety of conjugated polymers by other groups [44,45,46,47], presumably because of its high reactivity and ease of use. It is important to note that this highly reactive system can induce side reactions that cause structural defects; a direct arylation reaction can occur at undesired C–H bonds [34, 35, 45]. An appropriate choice of monomers and the optimization of reaction conditions are essential in order to avoid side reactions [43, 48].

Scheme 3
scheme 3

Direct arylation polycondensation of various thiophene derivatives with a phosphine-free catalytic system

Evaluation of polymers prepared by direct arylation polycondensation

For the further development of direct arylation polycondensation, the reaction conditions were investigated in a model reaction using 3,4-ethylenedioxythiophene (EDOT) [49], which possesses highly reactive C–H bonds for direct arylation (Scheme 4) [37, 50, 51]. The target polymer can be prepared using a conventional cross-coupling method (Scheme 4a, S-PEDOTF) [51]. An investigation of the reaction conditions for the direct arylation polycondensation of EDOT revealed that 1-adamantanecarboxylic acid (1-AdCOOH) is an effective additive that assists the C–H bond cleavage process. The reaction with 1-AdCOOH afforded the corresponding polymer (Scheme 4b, L-PEDOTF) in the presence of only 1 mol% of the Pd precatalyst [37]. The molecular weight of L-PEDOTF (Mn = 47500) is higher than that of S-PEDOTF (Mn = 17100), which was prepared using the conventional method. In addition, direct arylation polycondensation under microwave heating produced the corresponding polymer with an extraordinarily high molecular weight (Scheme 4c, H-PEDOTF) [50]. It is possible that the uniform heating produced under microwave irradiation promoted efficient coupling. These results demonstrate that optimized direct arylation polycondensation can produce higher molecular weight polymers than those obtained using conventional methods. One reason for the high molecular weights of L- and H-PEDOTF is the high tolerance of the C–H bonds to the reaction conditions. The lack of degradation of the reactive point is advantageous, especially compared with the monomers used in the conventional method. For example, the C-Br moiety in the brominated EDOT in Scheme 4a is known to be unstable [52] and the boronate ester moiety can be decomposed via protodeboronation [53].

Scheme 4
scheme 4

Three methods for the synthesis of EDOT-based polymers (PEDOTF)

The purity and semiconducting properties of PEDOTFs were investigated to confirm the advantages of the direct arylation method (Table 1). The results of the elemental analysis of H-PEDOTF matched well to the values calculated from the formula of the repeating unit. Bromine was not detected by elemental analysis, even though the fluorene terminal units may, in principle, possess a Br moiety. The absence of Br termini may be attributed to the small number of the terminal units as well as minor debromination reactions in the later stage of polymerization. In contrast, the analytical results indicate that L- and S-PEDOTF contain impurities and Br termini. Inductively coupled plasma atomic emission spectrometry (ICP-AES) measurements reveal that direct arylation polycondensation achieved lower levels of Pd residues in the polymers than the conventional method, which is due to the low loading of Pd (1 mol%) in the direct arylation method. It should be noted that high-molecular-weight and high-purity H-PEDOTF was obtained through simple purification steps, requiring only washing with several solvents and re-precipitation. Soxhlet extraction and high-performance liquid chromatography purification were not required. H-PEDOTF exhibited better semiconducting properties in OPVs and OFETs than did the other samples that were evaluated (Table 1) [51]. Bulk heterojunction (BHJ) solar cells with H-PEDOTF and PC70BM reached 4% power conversion efficiency (PCE), although that with S-PEDOTF was only 0.48%. As the PCE correlates with hole mobility in the polymers in OFETs, the better performance of H-PEDOTF in the OPV may be due to the high hole mobility in the polymer. As H-PEDOT has a higher molecular weight and purity than the other samples, the dominant factor responsible for the improvement is unknown. Therefore, the effects on terminal structure, the residual amount of Pd, and the molecular weight were investigated for the model polymers. Thus, PEDOTF samples with different terminal structures (Br or H), the amount of residual Pd, and the molecular weights were tabulated for comparison (Table 2) [54]. P1 is the same sample as L-PEDOTF. P2 has very little Br termini because of the treatment of the C-Br moiety at the end of polymerization. The amount of Pd residue was reduced to 3 p.p.m. in P3 by washing the polymer with an aqueous solution of sodium N,N-diethyldithiocarbamate. P4 has a purity that is similar to P3, but with a higher molecular weight. Regarding both the PCE and hole mobility, P1 has lower values than those of P2–P4, indicating that the terminal structure is the most dominant factor in this case. The Br terminals may act as hole-trapping sites [55]. In contrast, the Pd residue and molecular weight had a negligible effect on the initial performance of the OPVs and OFETs. The stability tests of the OPV devices showed that Pd residue and the molecular weight of the polymer affect the lifetime of the device; the device containing P4 has the longest lifetime among the P2–P4 group, although the initial performances were very similar [54].

Table 1 Purity and material properties of the polymers
Table 2 Effects of terminal units, Pd residue, and molecular weight on material properties

Synthesis of donor–acceptor polymer

Direct arylation polycondensation for the synthesis of donor–acceptor (D–A) polymers was investigated after the establishment of the model reactions because D–A polymers often exhibit high levels of performance in OPVs; the D–A structure possesses a narrow highest energy occupied molecular orbital/lowest unoccupied molecular orbital gap, enabling wide light absorption and high hole mobility because of the strong interchain interactions [1]. For the synthesis of D–A polymers, a new reaction system was required, because the conditions mentioned above are not applicable to C–H bonds in the acceptor monomers, although donor monomers, such as EDOT, can be readily polymerized. An investigation of the reaction conditions for acceptor C–H monomers revealed that the addition of PCy3 ligands and the selection of a low polar solvent (toluene) are effective for the smooth polycondensation of acceptor monomers [56] such as thienopyrroledione derivatives [57, 58]. In addition, a Pd(0) precatalyst (Pd(PCy3)2) has been found to be a suitable precatalyst by mechanistic studies of direct arylation, thereby avoiding side reactions, such as the homocoupling reaction [42, 59]. Based on these findings, the polycondensation of thienopyrroledione and dibromocyclopentadithiophene derivatives was conducted using Pd(PCy3)2 as the catalyst and toluene as the solvent (Scheme 5) [60]. The reaction produced the corresponding D–A polymer with a molecular weight of 25,000 in 82% yield, which was higher than previously reported values for the Migita–Kosugi–Stille coupling polycondensation [61]. Nuclear magnetic resonance and mass spectrometry analyses indicated that the polymer possessed no structural defects and no Br termini, which might be due to a minor debromination reaction that occurred during the polycondensation reaction. Elemental analysis demonstrated the high purity of the polymer. This D–A polymer served as the p-type semiconducting material in BHJ solar cells with PC70BM. The maximum PCE of the solar cell was 6.8% after optimization of the device’s structure. The PCEs of the devices were comparable to those of the polymer obtained from the Migita–Kosugi–Stille coupling polycondensation [61], thus confirming that the direct arylation polycondensation yields high-performance semiconducting materials.

Scheme 5
scheme 5

Synthesis of D–A polymer by direct arylation polycondensation

In-depth studies by other groups have also enabled the synthesis of high-performance polymers without structural defects [62,63,64]. Further development of this method will enable the synthesis of state-of-the-art materials that can lead to practical applications.

Polycondensation using C–H/C–H coupling reactions has been investigated to develop next-generation methods [65,66,67,68]. One example is the aerobic oxidation polycondensation of thiazole monomers (Scheme 6) [69]. This polymerization proceeds with a catalytic amount of Cu(OAc)2 and oxygen from the air and yields a bithiazole-based conjugated polymer along with the formation of H2O, which is a green by-product. The thiophene-thiazole polymer served as a semiconducting material in OFETs. As the monomer is synthesized by direct arylation of thiazole at the 5-position, this strategy skips the preparation of an organometallic reagent and reduces the overall number of reaction steps.

Scheme 6
scheme 6

Combination of direct arylation and aerobic oxidative coupling polycondensation