Conjugated Polymer Photovoltaic Materials

Abstract

During the past few decades, conjugated polymers with various molecular structures have been explored for applications in polymer solar cells (PSCs). In this chapter, an overview of conjugated polymer photovoltaic materials is given to provide insights for molecular design and fine-tuning of high-performance photovoltaic polymers. First, we briefly summarize and provide design considerations of conjugated polymer photovoltaic materials. Second, representative photovoltaic polymers are introduced. Third, representative conjugated polymer acceptor materials are briefly introduced and discussed.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

During the past few decades, conjugated polymers with various molecular structures have been explored for applications in polymer solar cells (PSCs). In order to obtain more efficient PSCs, great effort has been devoted to optimizing and synthesizing conjugated polymers with superior photovoltaic properties, which boost the power conversion efficiency (PCE) toward or even above 8–10 % in several research groups [1–13]. In this chapter, an overview of conjugated polymer photovoltaic materials is given to provide insights for molecular design and fine-tuning of high-performance photovoltaic polymers. First, we briefly summarize and provide design considerations of conjugated polymer photovoltaic materials. Second, representative photovoltaic polymers are introduced. Third, representative conjugated polymer acceptor materials are briefly introduced and discussed. Because photovoltaic properties of conjugated polymers are susceptible to device fabrication conditions and device structures, excluding the factors of optical and interfacial enhancement by device engineering, the ‘initial PCE’ of the PSCs used in simple conventional devices of ITO/PEDOT: PSS/Polymer: PCBM/Ca (or Mg, Ba, LiF)/Al (or Ag, Au), as depicted in Fig. 5.1, is discussed.

Fig. 5.1

Simple conventional device structure of a polymer solar cell

5.1.1 Brief Summary of Photovoltaic Polymers

As the key photovoltaic materials in PSCs, conjugated polymers can be classified into two types, namely donor and acceptor photovoltaic polymers. Following the pioneering works of Heeger et al. and Friend et al. in the 1990s [14, 15], thousands of polymer donors with different backbones and side groups have been developed, synthesized, and used in polymer solar cells during the past few decades, and have been deemed to be one of the driving forces of the development of PSCs. Conjugated polymer photovoltaic donor materials have been well discussed and reviewed recently [16–30]. However, to classify the polymer donors rationally is still quite difficult because of the rapid growth of the numbers of polymer donors. Herein, three types of the first used homopolymer donor materials, i.e., the derivatives of poly(1,4-phenylene vinylene) (PPVs), polythiophene (PTs), poly(thienylene vinylene) (PTVs), etc., are discussed as the first category of polymer donors (see Fig. 5.2). In recent years, the Donor–Acceptor copolymers (D–A copolymers) based on two or more conjugated building blocks have played important roles in promoting the development of the PSC field, so we provide more examples of conjugated D–A copolymers as the second category of polymer donors. For example, benzothiadiazole (BT)-based polymers, silole-containing polymers, diketopyrrolopyrrole (DPP)-based polymers, indacenodithiophene (IDT)-based polymers, Benzo[1,2-b:4,5-b′]dithiophene (BDT) based polymers, Thienopyrroledione (TPD) based polymers, etc., are introduced and discussed (see Fig. 5.2). In comparison with polymer donors, polymer acceptors have attracted less attention, although some recent work has shown that polymer acceptors have great potential in realizing highly efficient PSCs without using fullerene derivatives. Therefore, in the last section of this chapter, we give a brief introduction of polymer acceptors.

Fig. 5.2

Two categories of polymer donor materials and examples of representative polymers in each category

5.1.2 Design Considerations of Conjugated Polymer Photovoltaic Materials

The basic requirements for molecular design of high efficiency photovoltaic polymers have been discussed and summarized in several reviews [16–25]. It is known that the PCE of PSC can be calculated based on three photovoltaic parameters: the open circuit voltage (V _oc), the short circuit current (J _sc), and the fill factor (FF). For highly efficient photovoltaic polymers, important factors including solubility, light absorption, molecular energy level, mobility as well as morphology should be considered.

Appropriate solubility and good film-forming properties should first be considered in designing novel conjugated polymers. To get high quality thin films of conjugated polymers by solution coating processes, conjugated polymer donor materials must have good solubility in commonly used organic solvents such as chloroform (CF), chlorobenzene (CB), toluene, and dichlorobenzene (DCB). Because conjugated polymers have rigid backbones and intermolecular π–π interaction provides strong driving force for aggregation, the unsubstituted conjugated polymers are all insoluble in organic solvents. Therefore, long, flexible and/or branched side groups, such as alkyls or alkoxys, are introduced as functional groups onto their backbones to overcome the strong aggregation effect and hence to afford solution processability for conjugated polymers. For example, as shown in Fig. 5.3, the unsubstituted polythiophene is an insoluble polymer; however, when alkyls, such as hexyl, octyl, or alkyls with more carbon numbers, are introduced as side groups, the derivatives can be readily dissolved in many types of organic solvents such as CF, toluene, CB, o-DCB, and so on.

Fig. 5.3

Unsubstituted and substituted PTs

For a polymer photovoltaic material, a broader and stronger absorption, matching well with the solar radiation spectrum, is necessary to achieve high J _sc. As shown in Fig. 5.4a, solar irradiation has a very broad spectrum, which is mainly distributed at the visible and infrared regions with a peak at ca. 700 nm, so to harvest solar light the photovoltaic polymer should absorb the majority of the wavelength region from 400 to 900 nm. Clearly, from the point of view of the absorption spectrum, the PSCs based on P3HT cannot make good use of solar light. Therefore, conjugated polymers with low band gaps (LBG polymers) were developed and used in PSCs, leading to the possibility of realizing high J _sc. In fact, the development of LBG polymers has played a very important role in the rapid progress in the PSC field.

Fig. 5.4

a Spectral photon flux of AM 1.5 G and the UV-vis absorption of P3HT. b Schematic energy levels diagram of electron donors and acceptors in polymer solar cells

The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of photovoltaic polymers should be appropriate to maintain efficient charge separation and reduce energy loss. The schematic diagram of molecular energy and open-circuit voltage has been drawn in Fig. 5.4b. The value of V _oc for a PSC is directly proportional to the energy difference between the HOMO of the polymer donors and the LUMO of the acceptor materials [31]. As the LUMO of [6,6]-phenyl-C₆₁ (or C₇₁)-butyric acid methyl ester (PC₆₁BM or PC₇₁BM) is generally considered to be −3.9 eV, the V _oc of polymer/PCBM-based polymer solar cells should rely on the HOMO level of polymers. Therefore, lowering the HOMO level of the polymer donors and thus reducing the energy loss during the charge separation process have been seen as key to achieving high V _oc.

Furthermore, in order to facilitate charge transport in the BHJ active layer, a high mobility should also be required for the polymer donors. The hole and electron mobility (μ _h and μ _e) are important parameters to evaluate the photovoltaic properties of the donor and acceptor photovoltaic materials. Moreover, in order to reduce or avoid germinate and bimolecular recombination in the BHJ layers, the donor and acceptor materials used in a BHJ active layer in a PSC device should have balanced mobilities for holes and electrons. For example, considering that μ _e of PCBM is ca. 10⁻³ cm²/(V s), when PCBM is used as the acceptor, μ _h of the polymer donors should be kept at the same level or higher than that of PCBM.

Morphological properties of polymer/PCBM blends are also of great importance for the photovoltaic performance of polymers [32–35]. For example, because of the low dielectric constants, the exciton diffusion length in the bulk of conjugated polymers is less than 10 nm, so nanoscale phase separation in the BHJ active layer is needed. If the aggregations of the donor and/or the acceptor are too big, the excitons are not diffused to the D/A interface efficiently, so that strong geminate recombination is observed. Besides of the size of phase separation in the BHJ blend, the crystallinity of conjugated polymers is also an important issue. As is well known, the transport of π-electrons in polymer aggregations is anisotropic, i.e., the intramolecular charge transport is along the conjugated backbones, whereas the intermolecular charge transport is along the overlapped π-orbits, which are perpendicular to the conjugated backbones. Therefore, various methods have been developed to modulate morphologies of conjugated polymers in the BHJ blends. For example, for the photovoltaic system based on a LBG polymer called PDPP3T [34] and PCBM, when the blend was processed with CF, a very low PCE (<2 %) was achieved; using CF and 1,8-diiodooctane as binary solvent, a moderate PCE of 4.7 % was achieved. Recently, Ye et al. [35] introduced a ternary solvent system of DCB/CF/DIO to optimize the morphology as well as the overall performance of the PDPP3T/PCBM system, which increased the PCE to 6.71 %. Morphological studies of the photovoltaic system of PDPP3T and PCBM clearly revealed that the phase separation size and the crystallinity of the polymer in the blend can be tuned effectively by modulating the processing solvents of the films. On the other hand, morphologies of conjugated polymers can also be tuned by changing their molecular structures, which have been deemed as one of the main tasks for molecular design of conjugated polymers.

To meet the above requirements for highly efficient polymer donors, various molecular design strategies have been developed to modulate the photovoltaic properties of conjugated polymers and many new conjugated polymers have been designed and used in PSCs. In the following sections, the design strategies and some of the representative conjugated polymer photovoltaic materials are discussed in detail.

5.2 Conjugated Polymer Donor Materials

5.2.1 Three Important Types of Homopolymer

5.2.1.1 Poly(1,4-Phenylene Vinylene) Derivatives

In 1995, Yu et al. [14] pioneered the concept of the bulk-heterojunction (BHJ) in MEH-PPV:fullerene systems, which is considered to be the best PSC device architecture to date. Two poly(1,4-phenylene vinylene) (PPV) derivatives including poly[(2-methoxy-5-(2-ethyl-hexyloxy))-1,4-phenylenevinylene] MEH-PPV and poly[(2-methoxy-5-(3′,7′-dimethyloctyloxy))-1,4-phenylenevinylene] (MDMO-PPV) are very important donor polymers, which have played a very important role in the early stage of the development of the field of PSCs. Their molecular structure and corresponding photovoltaic results are shown in Fig. 5.5 and Table 5.1, respectively.

Fig. 5.5

Synthesis method and molecular structures of MEH-PPV and MDMO-PPV

Table 5.1 Photovoltaic results of some of the representative polymer donors of PPVs, PTVs, and the like

MEH-PPV and MDMO-PPV show excellent solubility in commonly used solvents such as tetrahydrofuran, CF, toluene, xylenes, CB, and 1, 2-DCB. The solutions of these two polymers show very good film forming abilities, and the dilute solution of MEH-PPV or MDMO-PPV in CB (1–2 mg/mL) can still form high quality films by the spin coating process, which gives more opportunity to utilize these polymers to carry out studies in device physics and device engineering. Actually, these two polymers can also be used as electroluminescent materials in polymer light emitting devices.

PPV and its derivatives can be synthesized via various methods, including the Wessling precursor method [36], Gilch method [37], Heck coupling reaction [38], Knoevenagel polycondensation [39], etc. With the Wessling method, the PPV precursor can be synthesized by the reaction of bis-(sulfonium halide) salts of p-xylene with a base (NaOH) in water or alcohol solution, and then the precursor solution is spin-cast on ITO substrate, PPV film being formed by heat treatment at 180–300 °C under vacuum. The PPVs prepared by the Wessling precursor method have a lot of defects and impurities because of the oxidation of the precursor polymer, the residual precursor moieties, and undesired side reactions during the thermal conversion.

By Heck coupling reaction [38], organic halides and vinylbenzene compounds can be coupled to generate a carbon–carbon bond under the catalysis of Pd(0). Many functional groups, such as aldehyde, ester, nitryl, hydroxy and carboxy, have no obviously negative effects on the coupling reaction, so that this reaction has been widely used for preparation of PPVs. Cyano-substituted PPV can be readily prepared by the Knoevenagel polycondensation reaction between equimolar amounts of a terephthaldehyde derivative and a 1,4-diacetonnitrile-benzene derivative [39]. The condensation reaction takes place upon addition of excess potassium tert-butoxide or tetrabutylammonium hydroxide in THF/tert-butanol mixture at 50 °C. In the Knoevenagel condensation reaction, tetrabutylammonium hydroxide is used as catalyst, and the solvent for the reaction can be THF, toluene, or DMF.

As shown in Fig. 5.5, the Gilch method is very convenient to prepare PPV and its derivatives, and, by this method, 1,4-bis-chloromethyl-benzenes were treated with potassium tert-butoxide in non-hydroxylic solvents such as tetrahydrofuran [37]. The temperature of the reactant, the concentration of the monomer and the base, and the speed of the base addition are all crucial conditions for molecular weight and PDI of the polymer. The molecular weight can also be controlled by using a benzylchloride derivative as the end capping reagent. Many PPV derivatives can be prepared with high molecular weight and high purity. Therefore, the Gilch method is the most successful method for the synthesis of PPVs, and the classic materials of PPVs, MEH-PPV and MDMO-PPV, are synthesized by this method.

A breakthrough in the PCE of the PSC devices based on PPV derivatives were presented by Brabec, Sariciftci, and coworkers [40]. In 2001, Shaheen et al. [40] introduced LiF as the n-type buffer layer in MDMO-PPV-based PSC devices and investigated the effect of processing solvent on the photovoltaic performance. By replacing toluene with CB, optimal phase separation and increased interactions between conjugated polymers were observed in MDMO-PPV-based PSC devices. As a result, a dramatically improved PCE up to 2.5 % was achieved, which was a nearly a threefold enhancement over previously reported values and also the world record at that time. Afterwards, Brabec et al. [41] systematically investigated the effect of the thickness of LiF on the photovoltaic performances of MDMO-PPV/PC₆₁BM-based PSC devices. Under optimal conditions, the PSC achieved an improved PCE of 3.3 %. Compared to PSCs without the LiF interfacial layer, the white light efficiencies of LiF-based PSC increased by over 20 %. Cao et al. [42] investigated the photovoltaic performance of MEH-PPV with a series of PCBM derivatives with different alkyl end groups on its side chain. The results revealed that the PC₆₁BM derivative appending the butyl end group performs best as acceptor blended with MEH-PPV and achieved a high PCE of 2.6 %. Wienk et al. [43] incorporated PC₇₁BM as the electron acceptor in thin-film polymer photovoltaic cells based on MDMO-PPV and provide a high J _sc because of the increased absorption in the visible region from PC₇₁BM and an ultrafast charge transfer upon photoexcitation of MDMO-PPV or PC₇₁BM. In 2008, Tajima et al. [44] investigated the effect of regioregularity on the photovoltaic properties of the MDMO-PPV-based PSC devices. Fully regioregular MDMO-PPV was utilized for polymer photovoltaic devices, and their performance was compared with that of regiorandom MDMO-PPVs. A PCE up to 3.1 % was recorded in regioregular MDMO-PPV, whereas a moderate PCE of 1.7 % was achieved with regiorandom MDMO-PPV. The higher PCE of regioregular MDMO-PPV originated from both higher hole mobility and better nano-morphology. Mikroyannidis et al. [45] synthesized a novel LBG PPV derivative in 2010, which generated a record PCE for the PPV/PCBM system so far. However, although MEH-PPV and MDMO-PPV played vital roles in the early years (1995–2003) of PSC study, the narrow absorption range (400–560 nm) and poor hole mobility (~10⁻⁷ cm² V⁻¹ s⁻¹) limited the photovoltaic performance of PPV derivatives.

5.2.1.2 Poly(thienylene vinylene) Derivatives and the Like

The likes of PPVs, poly(thienylene vinylene) derivatives (PTVs) were developed and applied in the PSC field because of their broad absorption and higher hole mobility. The molecular structure and corresponding photovoltaic results of a few PTV derivatives are provided in Fig. 5.6 and Table 5.1, respectively. Hexyl-substituted PTV, namely P3HTV, is the simplest soluble polymer among PTVs, although the photovoltaic performance of P3HTV is rather poor (PCE is ~0.2 %). The Stille coupling polycondensation reaction is widely used for synthesis of PTVs. As show in Fig. 5.6, PTVs can be easily prepared by using 2,5-dibromothiophenes and 1,2-bis-tributylstannylethylene as starting materials. In this reaction, aromatic solvents such as toluene, xylenes, and CB can be used as reaction solvents; the Pd(0) compound with appropriate ligand-like tetrakis(triphenylphosphine)palladium(0) can be used as the catalyst. Actually, the Stille coupling polycondensation reaction is also one of the most widely used methods for synthesis of other types of conjugated alternating copolymers. Therefore, although photovoltaic performance of the PSCs based on PTVs is quite low, the synthesis of PTVs paved the way for the study of molecular design and synthesis of highly efficient photovoltaic polymers.

Fig. 5.6

Molecular structures and synthesis routes of PTVs

Hou and Li et al. [46] introduced the conjugated side chains in the PTV backbone and synthesized a series of PTV derivatives with extended absorption in the UV-vis range from 350 to 740 nm. The PCE of the biTV-PTV-based PSC reached 0.32 %, which exhibited 52 % enhancement in comparison with that of the PSC device based on P3HTV under the same conditions. To overcome the intrinsic drawbacks of PTVs, for instance, high-lying HOMO levels and low V _oc, Li and coworkers [47] introduced an electron-deficient carboxylate group in P3HTV. The results indicated that the introduction of the carboxylate group in side chains can lower both LUMO and HOMO values of PTVs. Meanwhile, the photoluminescence of PTV can improved significantly. As a result, a higher V _oc up to 0.86 V was achieved. Under the D/A weight ratio of 1:2, a high PCE of 2.01 % was recorded. Notably, the PCE of the P3CTV-based PSC device is about 10 times higher than that of the device based on P3HTV. The PCE of 2.01 % is the highest efficiency for PSCs based on PTVs. The dramatic improvement of the photovoltaic performances of PTV by the carboxylate substitution demonstrated that PTVs might be promising photovoltaic polymers upon suitable structural modification.

In 2010, He et al. [48] reported a new vinylene-based polymer, PBDTV, which was copolymerized by vinylene and BDT via Pd-catalyzed Stille-coupling method. The PBDTV film depicted a broad absorption range covering from 350 to 618 nm and high hole mobility of 4.84 × 10⁻³ cm² V⁻¹ s⁻¹. At the optimal conditions (the weight ratio of PBDTV:PC₇₁BM of 1:4 and the active layer thickness of 65 nm), the PCE of the PBDTV-based PSC device reached 2.63 % with V _oc of 0.71 V, J _sc of 6.46 mA/cm², and FF of 57 % under the illumination of AM 1.5G, 100 mW/cm².

5.2.1.3 Polythiophene Derivatives

Polythiophenes (PTs), particularly regioregular poly(3-alkylthiophene)s (P3ATs), are a widely used class of polymer donors because of their excellent thermal and chemical stability as well as good charge transport properties. The photovoltaic properties of representative polythiophene derivatives (see Fig. 5.7) are listed in Table 5.2. Because PTs do not dissolve in most of the common solvents, alkyl substitution is a useful method to help with the solubility of P3ATs. The length of alkyl group in P3ATs plays an important role in determining the solubility, crystallinity, and morphology.

Fig. 5.7

Molecular structures of several PT derivatives

Table 5.2 Photovoltaic performances of some PT derivatives

Generally, the polymerization of thiophenes is carried out at their 2- and 5-positions. For many PTs, such as poly(3-alkylthiophene)s, the repeated units are asymmetric, so there are three relative orientations available when two thiophene rings are coupled between the 2- and 5-positions. Usually, the 2-position is called the head, and the 5-position the tail. As shown in Fig. 5.8, this leads to a mixture of four regioisomers when 3-substituted (or asymmetric) thiophene monomers are employed [49]. The HT-HT structure of PTs is seen as regioregular polymers, and the HT-HT isomer proportion in the polymer is known as regioregularity. Regioregular poly(3-substituted thiophene) can easily access a low energy planar conformation, so the regioregularity is an important factor in characterization of poly(3-substituted thiophene).

Fig. 5.8

Four regioisomers of poly(3-substituted thiophene)s

¹H and ¹³C NMR can be used to determine the structure and the regioregularity of PTs [50]. In a regioregular PT (HT-coupling ≈ 100 %), the proton at the 4-position of thiophene exhibits a neat peak at δ = 6.98. There are four chemically distinct triad regioisomers in regioirregular PATs, as shown in Fig. 5.9. In ¹H NMR spectra, the TT-HT isomer has a peak at δ = 7.00, HH-TT isomer has a peak at δ = 7.05, and the HT-HH isomer has a peak at δ = 7.02. Therefore, using the integral area of the peaks, the relative ratio of HT-HT couplings to non-HT-HT couplings can be determined. In the ¹³C NMR spectrum, regioregular PTs exhibits four resonances in the aromatic region (δ = 128.5, 130.5, 134.0, and 140.0 ppm), but regioirregular PTs show many resonances from 120 to 150 ppm.

Fig. 5.9

Synthesis methods of regioregular poly(3-alkylthiophene)s by the McCullough and GRIM method, Rieke method, and the Stille and Suzuki coupling reactions

PT derivatives can easily be synthesized by chemical oxidation methods [51]. In typical oxidation polymerization of PTs, 2,5-unsubstituted thiophenes, such as thiophene, 3-alkylthiophenes or 3-phenylthiophenes, etc., can be dissolved in CF, and, under an inert gas, the excessive oxidant (FeCl₃, MoCl₅, or RuCl₃) is added. The polymerization can take several hours. Generally, the FeCl₃ oxidation method has been widely used. The molecular weights of PTs prepared from this method range from 30 to 300 K, and the polydispersities range from 1.5 to 5.0. By this method, the regioregularities of poly(3-alkyl thiophene)s range from 70 to 80 %, whereas the regioregularity of poly(3-phenyl thiophene)s (P3PTs) can reach 90–95 %. The poor reproducibility is one of the major problems for oxidation polymerization reactions. As reported by Pomerantz et al. [52], the polymerization of 3-octylthiophene with FeCl₃ was repeated under identical reaction conditions five times, and the molecular weights of the five samples of poly(3-octylthiophene) ranged from 54 to 122 K with PDIs ranging from 1.6 to 2.7. Additionally, by using the identical preparation process, the polymer samples obtained from varied batches contain different levels of Fe impurities. Therefore, the oxidation method is seldom used to prepare PTs for applications in PSCs.

As shown in Fig. 5.9, regioregular poly(3-alkylthiophene)s can be prepared by various methods, such as the McCullough and GRIM method [53], the Rieke method [54], and the Stille [55] and Suzuki [56] coupling reactions. In 1992, McCullough et al. [53] reported the first synthesis method of head-to-tail coupled poly(3-alkylthiophene)s, and, from this method, close to 100 % HT-HT couplings. In this method, the monomer, 2-bromo-5-(bromomagnesio)-3-alkylthiophene, is obtained from 2-bromo-5-alkylthiophene at cryogenic temperature, and is then polymerized with catalytic amounts of Ni(dppp)Cl₂ (dppp is diphenylphosphinopropane). In this method, regioregular PTs were obtained in yields of 44–69 %. The Rieke method can also be used to synthesize regioregular poly(3-alkylthiophene) [54]. Regioselective control was realized on the basis of steric congestion at the reductive elimination step in the catalytic cycle. When the Ni(dppe)Cl₂ or Ni(dppp)Cl₂ was used as catalyst, the HT-couplings were more than 98.5 %. Other catalysts with less bulky, labile ligands such as PPh₃ combined with larger metal centers such as Pd lead to regiorandom poly(3-alkylthiophenes). The Stille [55] and Suzuki [56] coupling methods are two convenient approaches to synthesize regioregular PTs, and these two methods exhibit great advantages in synthesis of multi-functional PTs because of the compatibilities of a large number of organic functional groups.

Regioregular poly(3-hexyl) thiophene (P3HT) is the most widely studied polymers in the field of PSC. The photovoltaic performance of P3HT/PCBM has been widely studied by numerous groups. It should be noted that the photovoltaic properties of P3HT/PCBM-based PSC devices are strongly dependent on the molecular weight and regioregularity of P3HT as well as the device fabrication methods [57–59]. Various strategies have therefore been developed to optimize the performance of P3HT-based polymer solar cells. In 2003, Sariciftci et al. [60] developed a postproduction treatment, namely, applying external voltage, which considerably improved the photovoltaic performance of solar cells based on the P3HT/PCBM system. Using this method, an enhancement of J _sc and an increase in external quantum efficiency (EQE) of 70 % are demonstrated. The breakthrough in the photovoltaic performance of P3HT/PCBM was realized by Li et al. [61] and Ma et al. [62] in 2005. Li et al. [61] developed a method of slow growth to optimize the morphology of P3HT/PCBM and an extremely high PCE of 4.4 % was realized, which was the highest value in the past decade. Alternatively, Ma et al. [62] applied post-production annealing at 150 °C, and P3HT/PCBM-based PSC devices with PCE approaching 5 % were achieved. In addition, these devices exhibited remarkable thermal stability. The improved performance was ascribed to the improved nanoscale morphologies, the increased crystallinities of P3HT, and the improved interfacial contact to the cathode, thereby enhancing the overall device efficiency. In 2010, He and Li et al. [63] introduced a novel fullerene acceptor, namely ICBA with a higher-lying LUMO level for the P3HT based polymers, which promoted the performance of P3HT to a new height by greatly increasing V _oc. After that, solvent additives and device engineering methods were utilized and increased the PCE of P3HT/ICBA to over 6.5 % [64, 65]. Clearly, the success of P3HT is largely associated with the active layer morphology [66, 67], such as using different casting solvents and film-forming speed, solvent and thermal annealing, etc.

Other P3HT-analogue polymers, for example poly(3-butylthiophene) (P3BT) and poly(3-pentylthiophene) (P3PT), were also explored in the device fabrications because of their similar crystalline and absorption characteristics. Nguyen et al. [68] investigated the effect of different alkyl lengths such as butyl, hexyl, octyl, decyl, and dodecyl on the photovoltaic properties of P3ATs-based PSC devices. Results revealed that longer alkyl chains (number of carbon atoms over 8) give poor efficiency and larger scale of phase separation. Jenekhe et al. [69] demonstrated that P3PT/PCBM-based BHJ PSC devices also give similar performances to those of P3HT/PCBM under the same conditions. Jenekhe and coworkers [70] also prepared P3BT nanowires and P3PT nanowires by solution-phase self-assembly, which were used to construct highly efficient P3AT/PCBM PSC devices. The fullerene/P3AT nanocomposite films showed an electrically bicontinuous nanoscale morphology and desirable PCE up to 3.3 %, which were identical with those of P3HT/PCBM-based photovoltaic cells. Afterwards, Gadisa et al. [71] optimized the performance of P3PT/PCBM and a high PCE of 4.6 % was observed. Later, Li et al. [72] optimized the performance of P3PT with different fullerene acceptors. A PCE of 3.1 % and a PCE up to 5.4 % were achieved in P3PT/PCBM- and P3PT/ICBA-based PSC devices, respectively.

Two structurally related polymers of P3HT with high performance were developed by Hou and coworkers [73, 74]. Hou and Li et al. introduced conjugated side chains in PTs and developed a novel class of two-dimension (2D) conjugated polymers. A typical example among these 2D polymers is PTs with bi(thienylenevinylene) side chains. Three 2D conjugated PTs with bi(thienylenevinylene) side chains biTV-PTs (see PbTV in Fig. 5.7) were designed and synthesized by Hou and Li et al. in 2006 [73]. Compared with the properties of P3HT, the biTV-PTs show broad absorption bands (350–650 nm), much stronger absorbance, and lower HOMO levels. The PCE of the biTV-PT-based PSC devices reached ~3.2 %, which is ~40 % increased relative to that (~2.4 %) of the devices based on P3HT under the same conditions. This discovery expanded the scope of designing 2D conjugated polymers for efficient PSC devices.

Although photovoltaic performance as well as the photocurrent of P3HT can be improved via morphological optimizations, the low voltage (~0.6 V) is still the limiting factor of P3HT. In 2009, Hou et al. [74] also developed an easy and effective way to promote the V _oc of a poly(3-alkylthiophene) by reducing the number of alkyl chains of P3HT, and a novel P3HT-analogue polymer, P3HDTTT (see Fig. 5.7) with a V _oc up to 0.82 V was obtained. In 2011, Li et al. [75] also introduced carboxylate substituent in polythiophene derivatives to tune downward the HOMO of polythiophene. A novel polythiophene derivative of PT-C3 (see Fig. 5.7) was also synthesized and characterized. The PSCs based on PT-C3/PC₇₁BM exhibited relatively high V _oc of ~0.8 V. The PCE of the PSCs based on PT-C3 reached 3.87 % with V _oc = 0.78 V, J _sc of 9.68 mA cm⁻², and FF of 51.2 % under the illumination of AM1.5G, 100 mW/cm². These studies indicated that, in respect of further enhancing the J _sc, P3HDTTT would be a potential polymer to replace P3HT as blue absorber in tandem PSC devices.

5.2.2 Donor–Acceptor Copolymers

5.2.2.1 Benzothiadiazole (BT)-Based Polymers

D–A copolymers based on BT and its derivatives (see the molecular structures and photovoltaic parameters in Fig. 5.10 and Table 5.3) were explored because the HOMO and LUMO energy levels as well as band gaps can been effectively tuned by choosing donor (D) and acceptor (A) building blocks with appropriate electron-donating or -accepting natures. In 2003, the fluorene (FL) and benzothiadiazole (DTBT) copolymer, PFDTBT, was synthesized by Andersson and coworkers [76], which was generally considered as one of the first reports of donor–acceptor (D–A) conjugated polymers. In this work, PFDTBT exhibited poor solubility but reached a moderate PCE of 2.2 %. Although the PCE was not high, the D–A copolymer exhibited potentials such as LBGs and tunable energy levels. Following the pioneer work of Andersson et al., numerous novel D–A copolymers based on different donor units and acceptor units were synthesized and utilized in PSC devices. It should be noted that D–A copolymers have been the most successful class of polymer photovoltaic materials for PSCs in recent years.

Fig. 5.10

Molecular structures of D–A copolymers based on benzothiadiazole

Table 5.3 Photovoltaic results of D–A copolymers based on benzothiadiazole

Considering the V _oc of this polymer is as high as 1 V, the performance of PFBTBT still has a large promotion space. In 2008, Hou et al. [77] altered the alkyl chains in the classical PFDTBT backbone. Two novel polymers, bisEH-PFDTBT and bisDMO-PFDTBT, employing the same polymer backbone as PFDTBT but different side chains, were studied to investigate the side-chain effects. After carefully optimizing the side chains, the PCE of bisDMO-PFDTBT was increased to 4.5 %. Notably, the saturated alkyl chains have little influence on the molecular energy levels of PFDTBT. From quantum-chemical calculations, both the ethyl groups are in the proximity of the conjugated backbone, and thus decreased the probability of π–π stacking in bisEH-PFDTBT originating from the steric effect. In contrast, there are two small methyl groups located on the third and seventh carbons of the bisDMO-PFDTBT which not only decrease the steric effect but also increase the solubility and hole mobility.

The cyclopentadithiophene and DTBT copolymer, PCPDTBT, is also a well-known LBG D–A copolymer, which was developed by Brabec et al. [78] in 2006. PCPDTBT is the first LBG polymer with high efficient photovoltaic activity in the IR spectral region. In 2008, Bazan et al. [79] introduced solvent additive 1,8-octanedithiol (OT) to optimize the nanomorphology of this polymer, and a twofold enhancement was observed. A high PCE of 5.5 % was recorded and regarded as a milestone of novel photovoltaic polymers. It should be noted that PCPDTBT is an amorphous polymer. The approach provided a feasible tool for modulating the heterojunction morphology in donor/acceptor systems where thermal annealing is not effective.

Another representative D–A copolymer is PCDTBT, which was copolymerized by carbazole (CZ) and DTBT by the Leclerc group [80]. PCDTBT had a low-lying HOMO level of −5.45 eV and a moderate band gap of 1.88 eV. A preliminary PCE of 3.6 % was measured at a donor/acceptor weight ratio of 1:4. Various methods were proposed to optimize the photovoltaic properties of PCDTBT. Notably, PCDTBT is among the most efficient, stable, and low-cost photovoltaic materials for PSC devices with a high V _oc (0.85–0.90 V), high PCE (6–7 %), and long lifetime (~7 years), which is now considered to be one of the new benchmarks for the development of highly efficient BHJ solar cells [81, 82]. Afterwards, Bo et al. [83] optimized the side chains of DTBT; a novel polymer, HXS-1 with planar configuration, achieved a high PCE over 5 % and excellent FF approaching 70 %.

A platinum metallopolyyne (herein called Pt-DTBT) with a LBG of 1.85 eV was also reported by Wong et al. [84]. The PSCs based on the metallated polymer exhibited an average PCE of 4.1 % without annealing. It is noteworthy that the devices based on the Pt-DTBT/PCBM system exhibited a very high J _sc of ~15 mA/cm² and relatively high V _oc of ~0.80 V. This is the first time that a metallated polymer was applied in PSC devices to get such a high PCE.

5.2.2.2 Silole-Containing Polymers

Fused coplanar thiophene-based heterocycles, such as dithieno[3,2-b:2′,3′-d]silole (DTS) and IDT, have been actively utilized as donor units to construct novel D–A copolymers. To optimize the photovoltaic performance of D–A copolymers as discussed above, such as PFDTBT and PCPDTBT, silole-containing building blocks, such as silafluorene (SiF) and dithienosilole (DTS), have been widely utilized in efficient D–A polymers. In the following we overview the representative polymers in the third generation. The majority of these polymers exhibited high PCE ranging from 5 to 7 %.

Cao and coworkers [85] successfully introduced the silicon atom into FL units and copolymerized the novel SiF unit and 4,7-di(2′-thienyl)-2,1,3-benzothiadiazole (DTBT). High-performance polymer solar cells composed of an alternating copolymer PSiF-DBT as the electron donor and PC₇₁BM as the electron acceptor were investigated. A high PCE up to 5.4 % with a high V _oc of 0.90 V, a J _sc of 9.5 mA cm⁻², and a FF of 50.7 % was achieved under the illumination of AM 1.5G 80 mW/cm². Moreover, PSiF-DBT also showed a high hole mobility of ∼1 × 10⁻³ cm² V⁻¹ s⁻¹. Bo and coworkers [86] optimized the side chains in DTBT of the PSiFDBT, and obtained an improved PCE of up to 6.05 %. Driven by the high performance of PCPDTBT and PSiFDBT, Hou et al. [87] designed and synthesized a LBG DTS-containing polymer, PSBTBT. The preliminary PCE obtained with thermal annealing (140 °C, 5 min) was 5.1 % and then was increased to ~5.6 %. Chen et al. [88] further revealed the origin of the high performance of PSBTBT relative to PCPDTBT. Striking morphological changes were observed in polymer: fullerene bulk heterojunctions upon the substitution of the bridging atom. GIXRD investigations indicated increased π–π stacking in silole-based polymers compared to the carbon-bridged analogue [89, 90]. More importantly, the response range of the PSBTBT covered the whole visible range from 380 to 800 nm, which indicated that PSBTBT is an efficient red-absorbing polymer for tandem PSCs. Benefitting from the broad absorption (300–800 nm), PCE values up to 7 % were achieved in tandem PSC devices by the Yang group and others.

Afterwards, various units were copolymerized with these silole-containing building blocks and a class of low-band gap polymers was developed. Their molecular structure and photovoltaic parameters are depicted in Fig. 5.11 and Table 5.4. In particular, DTS-based copolymers show a broad absorption, relatively lower HOMO energy level, and higher hole mobility, which are attractive for researchers. Considering that the thiazolothiazole (TTz) unit has a rigid and coplanar configuration and thereby ensures a highly extended π-electron system and strong π–π stacking, Zhang et al. [91] copolymerized DTS with the TTz acceptor unit. The PCE of the PSC based on PDTSTTz/PC₇₁BM (1:1, wt/wt) reached 5.59 % with V _oc = 0.77 V, J _sc = 11.9 mA/cm², and FF = 61 % under optimized conditions (thermal annealing at 100 °C for 15 min). Cui et al. [92] designed a strong electron-withdrawing unit, naphtho[2,3-c]thiophene-4,9-dione, which was copolymerized with DTS to construct a D–A copolymer, PDTSNTDO, with a narrow band gap and lower lying HOMO level. The PCE of the PDTSNTDO-based device reached 5.21 %, with a high V _oc of 0.88 V.

Fig. 5.11

Molecular structures of silole-containing polymers

Table 5.4 Photovoltaic results of silole-containing polymers

Marks et al. [93] synthesized a new series of bithiopheneimide (BTI)-based donor–acceptor copolymers for efficient PSCs. Among these, PSC featuring BTI and DTS copolymer as donor and PC₇₁BM as acceptor exhibited promising device performance with PCE up to 6.41 % and high V _oc over 0.80 V. The BTI analogue, TPD-based device exhibited 0.08 V higher V _oc with an enhanced PCE of 6.83 %, which is mainly attributed to the lower-lying HOMO induced by the higher imide group density in the backbone. Lu and Tao et al. [94] recently synthesized a new D–A copolymer PDTSTPD of DTS and thienopyrrole-4,6-dione (TPD) obtaining both a LBG (1.73 eV) and a deep HOMO level of −5.57 eV. When blended with PC₇₁BM, PDTSTPD exhibited an excellent PCE of 7.3 % on the photovoltaic devices with an active area of ~1 cm². These results demonstrate the great potential of DTS-based polymers for high-performance solar cells, and provide valuable insights into structure-property relationship of atom substitution.

5.2.2.3 Diketopyrrolopyrrole (DPP)-Based Polymers

As one of the high-performance pigments, the DPP unit was developed in the past few decades for constructing LBG (1.5 eV) conjugated polymers produced by its strong electron-withdrawing properties. Another attractive property of DPP is its excellent charge carrier mobility for both holes and electrons. Most of the photovoltaic performance of DPP-based polymers is likely to depend on its solvent-induced morphology [32].

Janssen et al. [95] introduced DPP unit in D–A copolymers in 2008. Dozens of newly designed DPP-based LBG polymers have been frequently reported since 2009. The molecular structure and photovoltaic parameters of these DPP based polymers are depicted in Fig. 5.12 and Table 5.5. Hou et al. [96] copolymerized DPP with various electron-donating monomers and obtained a series of new LBG polymers based on the DPP unit. Among these DPP-based polymers, benzodithiophene and DPP copolymer, PBDT-DPP showed a small band gap of 1.34 eV and a moderate PCE of 4.45 % was achieved.

Fig. 5.12

Molecular structures of representative DPP-based efficient polymers

Table 5.5 Photovoltaic performances of the representative D–A copolymers based on DPP units

When copolymerizing DPP with thiophene and the like, ultra-LBG (E _g ~1.3 eV) polymers such as PDPP3T [34] and PDPP3MT [97] can be obtained. The first highly performing DPP-containing polymer, PDPP3T, was synthesized [34] and applied in organic photovoltaic and field-effect transistor devices by Janssen et al. in 2009. After optimizing processing solvent and molecular weight, the moderate PCE of 4.7 % was increased to 6.7 % by Ye et al. [35] and 7.0 % by Janssen et al. [97]. To tune the coplanarity of PDPP3T, Janssen et al. [97] introduced methyl into the thiophene group of DPP unit, and a higher performance ultra-LBG polymer PDPP3MT was achieved with PCE of 6.8 % in the classic device configuration. Bronstein et al. [98, 99] reported the synthesis and polymerization of a novel thieno[3,2-b]thiophene-DPP-based building block in recent years. Copolymerization with thiophene afforded the resulting polymer, PTT-DPP-T, with a high hole mobility of 1.95 cm² V⁻¹ s⁻¹. PSC devices comprised of PTT-DPP-T and PC₇₁BM also exhibited an excellent PCE of 5.4 % and high J _sc up to 15 mA/cm² [98]. Later, Li et al. [100] designed a high-molecular-weight conjugated polymer based on an alternating electron-rich TT unit and an electron-deficient DPP unit, which also provided efficient polymer solar cells with PCE up to 6.9 %. The optimal morphology of the new polymer/PCBM blend reduced bimolecular recombination and thereby allowed a high FF up to 70 % and high PCEs over 6 % with film thickness up to 300 nm to be achieved. Because of the lower π-electron density of the benzene unit compared with thiophene, the DPP and benzene copolymer, PDPPTPT [101], showed a relatively higher band gap of 1.53 eV and lower HOMO level (−5.35 eV) together with a broad photo-response range up to 800 nm. When blended with PC₇₁BM, higher PCE of 5.5 % with J _sc = 10.8 mA/cm², V _oc = 0.80 V, and FF = 65 % was achieved. Clearly, PDPPTPT is a suitable photovoltaic polymer for multi-junction devices because of the balance between the broad absorption range and high voltage.

Frechet et al. [102, 103] developed a series of furan-containing DPP-based polymers, such as PDPP2FT and PDPP3F, with substantial power conversion efficiencies. Inserting furan into the backbone of the conjugated polymers enables the utility of relatively small solubilizing alkyl chains because of the significant contribution of the furan to overall polymer solubility in common organic solvents. PSC devices fabricated from PDPP2FT and PC₇₁BM as active layers showed a high PCE reaching 6.5 % [102]. The design and synthesis of the successful examples of furan-containing LBG polymers paved the way to developing environmental photovoltaic polymers. Interestingly, it is also noted that furan-containing LBG polymers with high side-chain tunability also provide insights into molecular order in efficient PSCs. In recent years, the PCE of DPP based polymers was also increased to exceed 7 % by the Janssen group [104] and the Yang group [9–11] because of the molecular weight and side chain optimization. These highly efficient polymers are introduced in the fourth generation. Moreover, these red absorber materials exhibited potential applications in tandem and triple junction PSCs.

5.2.2.4 Indacenodithiophene (IDT)-Based Polymers

Ladder type units such as IDT, also known as thiophene-phenylene-thiophene (TPT), constitute a class of efficient D–A copolymers, which have been emerging as efficient building blocks since 2008 [105, 106]. Typically, the IDT unit is extremely versatile with a coplanar aromatic ring structure, and the electron density can be manipulated by the choice of the bridging group between the rings. The coplanarity of the IDT unit could enhance interchain interaction of the polymers and is expected to afford higher hole mobility. Ting et al. pioneered the IDT-based random and alternating copolymers for photovoltaic applications [107, 108]. In 2008, two LBG IDT-based polymers with high hole mobility (3.4 × 10⁻³ cm² V⁻¹ s⁻¹) were designed and synthesized for application in PSCs by Ting and coworkers [106]. High-performance PCE of 4.4 % was obtained, which was superior to that of the analogous P3HT based PSC device under the same device fabrication condition. In 2010, Ting and collaborators [108] also synthesized an alternating copolymer (a-PTPTBT) based on IDT and BT and the highest PCE reached 6.4 % of the corresponding PSC under the optimal condition of solvent vapor annealing. The molecular structure and photovoltaic parameters of the representative IDT based polymers are depicted in Fig. 5.13 and Table 5.6.

Fig. 5.13

Molecular structures of the representative IDT-based efficient polymers

Table 5.6 Photovoltaic results of the representative IDT-based polymers

Jen et al. [109] first applied aryl side chain substituted IDT in quinoxaline-based conjugated polymers and combined IDT and two quinoxaline derivatives to form novel polymers (PIDT-diphQ and PIDT-phanQ; Fig. 5.13). Because of the enhanced planarity of phenanthrenequinoxaline (phanQ), PIDTphanQ/PC₇₁BM-based PSC device exhibited an improved PCE of 6.24 % compared to the PCE of 5.69 % in PIDT-diphQ/PC₇₁BM-based device. Jen et al. also developed several high-performance IDT-based D–A copolymers such as PIDT-DFBT [110, 111]. Huang, Cao and coworkers [112] synthesized three low band-gap conjugated polymers via Stille copolymerization of IDT and naphtho[1,2-c:5,6-c]bis(1,2,5-thiadiazole) (NT) based monomers. The energy levels, absorption spectra, and band gaps of the target polymers were well tuned by utilizing different thiophene derivatives as spacer between IDT and NT units, and polymer PIDT-C12NT which employed bithiophene attached with dodecyl side chain as spacer exhibited superior properties compared with the other two copolymers. All polymers exhibited deep HOMO levels and subsequently led to high open circuit voltages of the fabricated PSC devices. The best performance (5.05 %) was achieved with PIDT-C12NT as donor polymer, which can be ascribed to its higher hole mobility, the optimal interpenetrating network, as well as enhanced absorption coefficient with respect to the other two polymers. The photovoltaic results demonstrated that the combination of IDT and NT with appropriate spacers might be a promising approach for the application of solar cells.

Considering potentials of the alkyl-substituted IDT, such as good planarity, good solubility, and high hole mobility, Zhang et al. [113] copolymerized alkyl-substituted IDT with different acceptor units including bithiazole (BTz), thiazolothiazole (TTz), tetrazine (TZ), and benzothiadiazole (DTBT). Among these copolymers, PIDTTTz has the highest hole mobility of 4.99 × 10⁻³ cm² V⁻¹ s⁻¹ and the PCE of the PSC based on PDTSTTz/PC₇₁BM reached 5.79 % with a V _oc of 0.89 V, a J _sc of 13.3 mA/cm², and a FF of 48.9 %, under the donor/acceptor component ratio of 1:2 (wt/wt). In comparison with PIDT-TTz, PIDT-DTBT has a medium band gap of 1.68 eV and a similar hole mobility of 2.24 × 10⁻³ cm² v⁻¹ s⁻¹. The PSC based on PIDT-DTBT/PC₇₁BM reached an even higher PCE of 6.17 %. These results indicated that the D–A copolymers based on the alkyl-substituted IDT unit are promising photovoltaic polymer materials because of the excellent hole mobility and solubility as well as deeper HOMO levels. The selenophene and other analogues of IDT were developed in recent years. For instance, Jen et al. [114] improved the molecular weight of IDSe-based polymers and a high PCE up to 6.8 % was recorded for a PIDSe-DFBT-based PSC device, which exhibited over 10 % enhancement compared to that of a PIDT-DFBT-based PSC device. This work demonstrated that selenium substitution on the IDT is an effective method to reduce the band gap and improve the photovoltaic performance of IDT-based polymers with higher molecular weight.

Recently, several IDT-based polymers realized high PCEs over 7 % and exhibited unique charge transport properties. For instance, Jen et al. [111] incorporated TT moiety in IDT unit and designed an IDTT unit. Utilizing IDTT unit in DFBT and TT bridged DFBT-based polymers, corresponding copolymers PIDTT-DFBT and PIDTT-DFBT-TT with PCE up to 7 % as well as V _oc over 0.95 V were observed. Hou et al. [115] incorporated IDT with DTQx and produced a highly efficient polymer PIDT-DTQx with a PCE up to 7.5 %, which was the highest value in IDT-based copolymers. Interestingly, PIDT-DTQx exhibited the best performance under the donor/acceptor weigh ratio of 1:4.

5.2.2.5 Benzodithiophene (BDT)-Based Polymers

Although various efficient photovoltaic polymers have been developed, the polymers discussed above (PCE < 7 %) still could not meet the need of commercial requirements. In this part, we overview the representative highly efficient photovoltaic polymers. In particular, the rapid progress of benzo[1,2-b:4,5-b′]dithiophene (BDT) and two dimensional BDT-based polymers are introduced. These polymers with PCE over 7 % draw lots of attention from both the industrial and the academic communities.

Benzodithiophene (BDT) has a large planar conjugated structure and easily forms π–π stacking, and thereby improves the hole mobility [116]. In 2008, Hou et al. [117] first introduced the BDT unit in the synthesis and application of photovoltaic polymers. In the work, alkoxy-substituted BDT was copolymerized with seven different units. The band gap and absorption spectrum of the BDT-based polymers could be effectively tuned within a wide range. In 2009, Yu and coworkers copolymerized BDT with thieno[3,4-b]thiophene (TT) and designed a series of BDT-TT copolymers, namely the PTB series [118–120]. Among these PTB series polymers, PTB7 [120] is the best-performing photovoltaic polymer because of the optimal side chain and functional substituents. Hou et al. also designed a series of BDT and TT copolymers, such as PBDTTT-C [121], PBDTTT-CF [122], and PBDTTT-S [123]. It should be mentioned that PBDTTT-CF and PTB7 are the first two polymers with PCE exceeding 7 %, which significantly pushes the PSC research to a new height. Following this pioneering work, PCEs of more than 7 % were frequently reported in various groups, as depicted in Fig. 5.14. The photovoltaic results of the corresponding polymers are listed in Table 5.7.

Fig. 5.14

Molecular structures of BDT-based highly efficient polymers

Table 5.7 Photovoltaic performance of the representative D–A copolymers based on BDT and BDT analogues

Besides TT, numerous building blocks including DPP, thieno[3,4-c]pyrrole-4,6-dione (TPD), benzothiadiazole (BT), and benzoxadiazole (BO) have been successfully combined with the BDT unit to constitute highly efficient photovoltaic polymers.

You and collaborators [124] reported the first successful application of fluorinated benzothiadiazole in BDT-based polymers. The resulting polymer PBDTDTffBT exhibited down-shifted HOMO and LUMO energy levels and a similar band gap relative to its non-fluorinated analogue. As a result, the PSC device employing PBDTDTffBT achieved a greatly improved PCE of 7.2 %. Similarly, fluorine was incorporated into 2-alkyl-benzotriazoles (TAZ) by You et al. [125] and a novel polymer PBnDT–FTAZ was designed. Interestingly, although the band gap of fluorinated polymer PBnDT–FTAZ was ~2.0 eV, the copolymer exhibited a PCE above 7 % when mixed with PC₆₁BM, and a PCE above 6 % was still attained even for thickness up to 1 μm. The superior performance originated from their high hole mobility and low HOMO and LUMO levels. Very recently, Wang et al. [126] successfully optimized the similar fluorinated benzothiadiazole-based conjugated copolymers, PBDTDTBTff, with different side chains. A PCE up to 8.30 % was achieved in PBDTDTBTff-3-based PSC devices with ~100 nm thickness active layers without any processing additives or post-treatments, which was the highest value for the conventional single-junction polymer solar cells via simple fabrication architecture. In addition, it is noteworthy that PBDTDTBTff-3 could afford high PCEs of 7.27 % at ~200 nm thickness active layers and 6.56 %, even for thicknesses up to ~300 nm. Therefore, the results demonstrated that BDT and fluorinated benzothiadiazole polymers should be promising candidates for developing high-performance large-scale roll-to-roll fabrication of PSCs.

Peng et al. [127] further applied dialkylthiol-substituted BDT in the synthesis of conjugated copolymers based on monofluorinated benzotriazole (TAZ) acceptor block, and a high-performance PBDTFBZS was developed. As expected, wide band-gaps and deep HOMO and LUMO energy levels were observed in the resulting copolymers. A PCE up to 7.74 % was achieved from the regular single device based on PBDTFBZS with a V _oc = 0.88 V, a J _sc = 12.36 mA/cm² as well as a high FF of 71.2 %. The enhanced V _oc can be ascribed to a low-lying HOMO energy level by the introduction of dialkylthiol and fluorine substituents on the PBDT-DTBT polymer backbone. The improvements in J _sc FF are probably because of high hole mobility, suppressed charge recombination, and optimal blend morphology. Because of the excellent performance of polymers, tandem PSC devices featuring PBDTFBZS as blue absorber material and DPP-based polymer as red absorber material exhibited high PCE up to 9.40 %. Compared to BT-based polymers, benzoxadiazole (BO)-based polymers exhibited a rather disappointed performance. Very recently, to overcome the relatively poor performance and low solubility of BO-based polymers, Li and coworkers [128] designed a TT-bridged polymer, namely, PBDT-TT-BO, which was copolymerized by BDT unit and TT-bridged BO acceptor unit. The PCE of the PSC device featuring PBDT-TT-BO as donor polymer reached 7.05 %, which was the champion result in BO containing conjugated polymers and comparable to that of its BT counterparts. A medium-band gap fluorinated quinoxaline-based conjugated polymer of PBDT-TFQ was designed and synthesized by Chou and coworkers [129]. With an optimized blend ratio of PBDT-TFQ:PC₇₁BM (1:1, wt/wt), a high PCE of 8.0 % was obtained, with a V _oc of 0.76 V, a J _sc of 18.2 mA/cm², and a FF of 58.1 %. The resulting copolymer achieved an extremely high J _sc, which was probably caused by the higher hole mobility of PBDT-TFQ together with the better morphology for efficient exciton dissociation and charge transport.

In recent years, a large class of donor–acceptor copolymers based on the two-dimensional conjugated BDT units (collectively called 2D-BDT units) was developed by Hou and coworkers [130–136]. In 2010, Huo et al. [130] copolymerized thiophene bridged 2,1,3-benzothiadiazole (BT) with an alkylthienyl-substituted BDT unit, and the first 2D-BDT-based polymer, PBDTTBT with V _oc up to 0.92 V and PCE over 5.6 % was achieved. By introducing the 2D-BDT units such as alkylthienyl substituted BDT, the PCEs of several novel polymers, including PBDTTT-C-T, PBDTTDTT-S-T, PBDTP-DTBT, and PBT-3F, have been increased to 8–9 % in Hou’s group. A typical example is PBDTTT-C-T. Huo et al. [131] replaced the alkyl chain with an alkylthienyl side chain in the BDT unit of PBDTTT-C, which resulted in a high-performance polymer PBDTTT-C-T. Notably, PBDTTT-C-T has been widely utilized in versatile photovoltaic devices such as inverted devices because of its excellent properties. The 2D-BDT design rules were also successfully applied to more than three pairs of BDT-based systems [132–136]. Two donor–acceptor conjugated polymers, PBDT-DTBT and PBDT-DTNT, based on 2,1,3-benzothiadiazole (BT) and naphtho[1,2-c:5,6-c]bis [1, 2, 5] thiadiazole (NT), have been designed, synthesized, and characterized by Huang and collaborators [137]. Compared with BT, NT contains two fused 1,2,5-thiadiazole rings which narrow the band gap, enhance the interchain packing, and improve the charge mobility of the resulting polymer. Consequently, the NT-based polymer PBDT-DTNT exhibited considerably better photovoltaic performance with a PCE of 6.00 %, which is significantly higher than that of the BT-based polymer PBDT-DTBT under identical conditions. Huo et al. [138] also synthesized a wide band gap BDTT-based polymer, PBDTTTZ, and achieved a desirable PCE of 5.21 %, which is one of the highest among wide band gap (>2 eV) polymers. Considering that the absorption edge of the PBDTTTZ is 620 nm, which is ∼20 nm shorter than that of P3HT, PBDTTTZ should be an excellent blue absorber for tandem devices relative to P3HT. Clearly, the impressive performance of BDT-based polymers has shown its obvious potential for achieving high performance in PSCs. The photovoltaic results demonstrated that photovoltaic polymers based on 2D-BDT units exhibited improved hole mobilities, and significantly improved photovoltaic performance relative to those of their corresponding alkoxy-substituted BDT-based photovoltaic polymers. Therefore, replacing BDT with 2D-BDT units should be a method to enhance the efficiency with wide applicability.

Similarly, Chen et al. [12] recently introduced the alkylthienyl side chain in the BDT unit of PTB7-Th, which produced a significantly improved PCE (~9.35 %) relative to PTB7 under the same conditions. Yang et al. [9–11] incorporated the 2D-BDT unit in the DPP based polymers, and high-performance low band-gap polymers such as PBDTT-DPP [9] and, PBDTT-SeDPP [10] were designed and synthesized. These PBDTT-DPP materials played vital roles in versatile highly efficient semi-transparent and tandem PSC devices. It is worth mentioning that the tandem devices incorporating PBDTT-DPP as LBG absorbing polymers have increased the PCE to a new height (PCE > 9 %) and have drawn worldwide attention.

Recently, alkylphenyl substituted BDT (BDT-P) has attracted much attention as a weak electron-donating unit with a large π-conjugated area and good planarity. Yang and coworkers [11] reported a series of copolymers based on BDT-P for photovoltaic application. Results revealed that BDT-P-based polymers exhibit similar photovoltaic performance and relatively higher V _oc in comparison with polymers based on alkylthienyl substituted BDT. Afterwards, Zhang et al. [139] designed and synthesized a novel copolymer PBDTP-DTBT based on benzothiadiazole and BDT-P. The best-performing PSC device based on PBDTP-DTBT/PC₇₁BM (1:1.5, wt/wt) reached a PCE up to 8.07 % with a V _oc = 0.88 V, a J _sc = 12.94 mA/cm², and a FF = 70.9 % under the irradiation of AM 1.5G, 100 mW/cm². Interestingly, with only 0.5 vol.% DIO, the PBDTP-DTBT-based D/A blends exhibited the best performance because of the well-tuned morphology. Although the DIO volume was higher than 1 %, the PCE was greatly reduced to a moderate value (<7 %). To ameliorate the relatively low voltage in alkylthienyl-substituted BDT and thiophene-bridged TT-based copolymers, Zhang et al. [140] further demonstrated the synergistic effect of introducing fluorine (F) atoms of lowering the molecular energy levels, HOMO and LUMO levels of copolymers of alkylthienyl substituted BDT and thiophene-bridged TT. When three F atoms were introduced in both the donor and acceptor units, the PSC device based on the trifluorinated polymer (PBT-3F) showed an extremely high PCE of 8.6 % and a significantly improved V _oc of 0.78 V, which was the efficiency record for PSCs utilizing the classic device configuration. This work demonstrated that introducing F onto the appropriate positions of the donor units in D–A polymers, especially BDT-based polymers, is a promising method to modulate effectively the molecular energy levels for better applications in PSCs.

Driven by the great success of BDT, various BDT analogues were designed and synthesized for producing high-performance D–A copolymers [141–144]. For instance, Hou et al. [142] designed a novel BDT analogue, namely DTBDT, and synthesized a series of DTBDT-based copolymers. Among these polymers, PDT-S-T [143] exhibited the best performance with PCE up to 7.79 % because of favorable and ordered molecular packing originating from the linear conformation of the backbone. Yu et al. [144] copolymerized DTBDT and TT, and a highly efficient polymer, PTDBD2, with PCE over 7.5 % was obtained. The V _oc of the PTDBD2-based PSC device is as high as 0.89 V, which is remarkably higher than that (0.74 V) of the PTB7-based PSC device. These high results indicated that DTBDT should be a promising candidate building block for highly efficient photovoltaic materials.

From the photovoltaic results listed in Table 5.7, BDT and its analogues have been shown to be the most successful building block for highly efficient photovoltaic polymers over 7 %. Notably, the photovoltaic performance of BDT-based polymers such as PTB7, PBDTTT-C-T, and PBDT-DTNT was further increased to 9 % by several groups via device innovations [2–8]. These device optimizations are not presented in this chapter in detail.

5.2.2.6 Thienopyrroledione (TPD)-Based Polymers

The thieno[3,4-c]pyrrole-4,6-dione (TPD) family of conjugated polymers has shown promise for PSC applications (see Fig. 5.15 and Table 5.8). The advantage of TPD is relatively cheap and easy synthesis. Incorporation of appropriate alkyl chains on TPD not only enables the preparation of soluble polymers but also greatly tunes the molecular packing as well as blend morphology. TPD-based push-pull polymers for photovoltaic applications were proposed in 2010 by Leclerc and co-workers [145, 146].

Fig. 5.15

Molecular structures of TPD-based highly efficient polymers

Table 5.8 Photovoltaic performance of the representative D–A copolymers based on BDT analogues

In the search for novel polymers suitable for PSCs, Leclerc [146], Jen [147], Xie [148], and Frechet [149] independently synthesized PBDTTPD, which included the benzodithiophene (BDT) unit as donor moiety and TPD as acceptor moiety. The preliminary PCE of the PBDTTPD/PC₇₁BM-based PSC device was as high as 5.5 % without additive optimization [146]. After utilizing coadditives of chloronaphthalene (CN) and diiodooctane (DIO), a PCE up to 7.1 % was achieved by Tao and coworkers [150]. Very recently, alkyl chain engineering was successfully utilized in PBDT-TPD, which dramatically promoted the efficiency of PBDT-TPD by Frechet and coworkers [151]. In their work, replacing branched side chains by linear ones in the BDT motifs induced a critical change in polymer self-assembly and backbone orientation in thin films which correlates with a dramatic drop in solar cell efficiency. In contrast, for polymers with branched alkyl-substituted BDT motifs, controlling the number of carbon atoms in the linear alkyl-substituted TPD motifs could effectively improve material performance. Optimized through this approach, PBDTTPD polymer-based PSC devices acquired high PCE of 8.5 % and high V _oc of 0.97 V, making PBDTTPD one of the best polymer candidates for the wide band gap subcell of tandem PSCs. This report emphasized the determining role that linear side-chain substituents play in the device performance of PBDTTPD.

In the previous part, a TPD-silole copolymer, PDTS-TPD with high PCE up to 7 %, was introduced [94]. In order to improve the intermolecular interactions of silole-based polymers further, Reynolds and co-workers [152] substituted silicon by the larger germanium atom and prepared the first dithienogermole (DTG)-containing conjugated polymer. The dithienogermole-thienopyrrolodione copolymer, PDTG-TPD, displayed an absorption shift to 735 nm, and a higher HOMO level than the analogous copolymer containing the commonly utilized DTS heterocycle. When PDTG-TPD was utilized in inverted PSCs, the cells displayed an average PCE of 7.3 %, relative to 6.6 % for the DTS-containing PSCs prepared under identical conditions. Notably, Reynolds et al. [153] also fabricated highly efficient PSC devices based on PDTG-TPD with a certified PCE up to 7.4 %, which is among the highest PCEs reported for photovoltaic polymers compatible with the roll-to-roll process. Followed by PDTS-TPD and PDTG-TPD, dithienogermolodithiophene (DTTG) was also incorporated as building blocks in TPD-based polymers because of the potentials of extended conjugation length and improved coplanarity. Very recently, Heeney et al. [154] reported the first synthesis of a novel ladder-type fused ring donor, DTG, in which two thieno[3,2-b]thiophene units are held coplanar by a bridging dialkyl germanium (Ge). Polymerization of DTTG with TPD afforded a polymer, PDTTG-TPD, with an optical band gap of 1.75 eV combined with a HOMO level of −5.68 eV. Bulk heterojunction PSC devices based on PDTTG-TPD/PC₇₁BM afforded a PCE up to 7.2 % without the need for thermal annealing or processing additives. The preliminary results indicated that DTTG-based polymers are promising candidates for high-performance PSC devices. It should also be noted that the synthetic route provides considerable synthetic scope to promote the performance further by altering the bridging atoms.

Another interesting example of TPD-based copolymer was published by Wei and co-workers [155]. Wei et al. designed a crystalline polymer PDTTPD, which was constructed by bithiophene and TPD. A preliminary PCE of 5.0 % (V _oc = 0.94 V, J _sc = 9.1 mA/cm²) was achieved for the PBTTPD/PC₇₁BM system. Incorporating a small amount of diiodohexane (DIH) in the blend resulted in the formation of substantially enhanced polymer crystallinity and smaller as well as better dispersed PC₇₁BM domains. As anticipated, an improved J _sc as high as 12.1 mA/cm² and a PCE of 7.3 % were recorded. Following the work of Wei et al., Marks and co-workers [156, 157] developed a series of TPD copolymers bearing thiophene, bithiophene, terthiophene, and quaterthiophene derivatives, respectively, as electron-donating moieties. Among these, a high-performance polymer PTPD3T [157] exhibited a high PCE of 7.9 % and extremely high FF up to 79.6 %. It is important to emphasize that the synthesis of monomers and polymers is both easy and versatile, wheres that of DTS or DTG derivatives is not.

5.2.2.7 Other Highly Efficient Polymers

As shown in Fig. 5.16 and Table 5.9, besides the above-mentioned polymers, other efficient polymers based on monomers such as quinoxaline (QX), thiazolo[5,4-d]thiazole (TTZ), thiophene bridged 2,1,3-benzothiadiazole, or naphtho[1,2-c:5,6-c]bis [1, 2, 5] thiadiazole also exhibited a high PCE over 6 %. QX has been widely implemented as an electron-deficient monomer of LBG polymers in PSCs. In 2010, Wang et al. [158] developed an easily synthesized donor–acceptor polymer (TQ-1) which showed PCE up to 6 %, with a high V _oc of 0.89 V, indicating that this polymer is a particularly promising candidate for high-efficiency low-cost polymer solar cells. Similar to DPP, isoindigo (IID) is also a high-performance pigment, which was developed in recent years for constructing LBG (<1.5 eV) conjugated polymers because of its strong electron-withdrawing character. By choosing the appropriate electron-rich unit terthiophene as the donor and IID as the acceptor, an easily accessible IID-based high-performance photovoltaic polymer (P3TI) was also reported by Wang and coworkers [159]. P3TI/PC₇₁BM-based PSC devices have an IQE of ∼ 87 % and a V _oc of 0.7 V with an optimized device PCE up to 6.3 %. Qian et al. [160] designed a novel polymer, PBT1, which is copolymerized with benzodithiophene-4,8-dione (BDD) and α-quaterthiophene units. Interestingly, a PCE up to 6.88 % was recorded in a PBT1-based PSC under optimal condition of high donor:acceptor ratio (1.5:1, wt:wt) and small thickness (75 nm). This work provided a successful example of using molecular structure as a tool to realize optimal photovoltaic performance with high polymer content, thereby enabling the realization of efficient absorption in thin films.

Fig. 5.16

Molecular structures of other highly efficient polymers with PCE over 6 %

Table 5.9 The photovoltaic parameters of other highly efficient polymers with PCE over 6 %

A recent breakthrough in several building units was made. Geng and coworkers [161] designed a five-ring-fused aromatic unit, namely dithieno[3,2-b;6,7-b]carbazole (DTC), which is structurally related to FL. Optimized by inverted device structures, PIIDDTC achieved an extremely high PCE of 8.2 %. Notably, PIIDDTC is the first IID-based polymer with a PCE beyond 7 %, and amorphous polymer with a PCE beyond 8 %. Naphthodithiophene (NDT) also emerged as an efficient building block in highly efficient photovoltaic polymers. Osaka et al. [162] recently developed photovoltaic copolymers based on naphthodithiophene (NDT). Introducing linear alkyl chains improved the solubility as well as gave rise to a change in the orientation without any alteration of the energy levels, which resulted in quite an impressive PCE over 8 % in a conventional single-junction PSC device. Surprisingly, the introduction of linear alkyl chains led to a drastic change in polymer orientation into the face-on motif, which was beneficial for the charge transport in solar cells and promotion of the photovoltaic performance. In addition, PSC devices based on the NNT-BT yielded PCEs as high as 8.2 % with thickness up to 300 nm. These results indicated that this polymer platform is of particular interest in the understanding of molecular packing and carrier transport in high-performance photoelectronic polymers.

A regular alternating terpolymer design strategy was proposed and applied to produce a photovoltaic polymer with tailored energy levels and optical band gap by Janssen and collaborators [163]. High-molecular-weight photovoltaic materials the terpolymer PDPP3TaltDPP were obtained with high efficiencies up to 8.0 % in PSCs employing PC₇₁BM as acceptors. Relative to the PCE of control copolymers PDPP3T (7.1 %) and PDPPTPT (7.4 %), the PCE of PDPP3TaltTPT exhibited more than 5 % improvement, which demonstrated that terpolymer could outperform the two parent copolymers when the design strategy was applied.

Similar to PTPD3T, Facchetti [26, 170] also developed a high-performance polymer PBTI3T, which demonstrated excellent PCE over 8.6 % and exceptionally high FF approaching 80 %. The high FF of PBTI3T is comparable to that of their inorganic counterparts. The extremely high FFs originated from the highly ordered, closely packed, and properly oriented active-layer microstructures with optimal horizontal phase separation and vertical phase gradation, which is beneficial for efficient charge collection and eliminated bulk as well as interfacial bimolecular recombination. This work depicted a comprehensive example to produce high FF in PSC device by integrating complementary materials design, synthesis, processing, and device engineering strategies.

Recently, a dithieno[3,2-b:2′,3′-d]pyran (DTPy)-containing polymer, PDTP-DFBT, was reported by Yang and coworkers [164]. The electron-donating property of the DTPy unit was found to be the strongest among the most frequently used donor units such as benzodithiophene (BDT) or cyclopentadithiophene (CPDT) units. When DTPy unit was polymerized with the strongly electron-deficient difluorobenzothiadiazole (DFBT) unit, a LBG (E _g = 1.38 eV) polymer PDTP-DFBT was obtained. In comparison with BDT or CPDT units, the DTP-based polymer PDTP-DFBT showed significantly improved solubility and processability as well as PCE. Excellent performance in single and double junction solar cells was obtained with the PCEs reaching 8.0 and 10.6 %, respectively, which demonstrated that DTPy unit is a promising building block for high-performance photovoltaic materials.

5.3 Conjugated Polymer Acceptor Materials

Although fullerene acceptors, particularly the well-known PCBM, bis-PCBM, and ICBA have been widely used and performed well with photovoltaic donor polymers in BHJ PSC devices, these materials are still not the best choices for photovoltaic industry due to the high cost and narrow absorption coverage [165]. Because of the high absorption coefficient, broader absorption, and high electron mobilities as well as suitable affinities, non-fullerene acceptors such as perylene bisimide small molecules [166–168], as well as polymers [169, 170], have attracted substantial interest as alternative acceptor materials to fullerene acceptors.

LBG conjugated polymers can also be used as an acceptor in a PSC if its LUMO energy level is low enough. As an electron-deficient building block, PDI can be copolymerized easily with a wide variety of electron-rich units to tailor the molecular levels as well as absorption properties of the resulting D–A copolymers. The molecular structures and photovoltaic results of all-polymer solar cells with PCE over 1 % are summarized in Table 5.10 and Fig. 5.17. In 2007, Zhan et al. [171] reported the synthesis of the first soluble rylene containing polymer based on alternating perylene diimide (PDI) and dithienothiophene (DTT), which exhibited good solution processability, broad absorption, excellent thermal stability, and high electron affinity. Electron mobilities as high as 1.3 × 10⁻² cm² V⁻¹ s⁻¹ have been measured by the OFET method. All-polymer solar cells using this polymer as acceptor polymer and a polythiophene derivative as donor polymer achieved a high PCE of over 1 % under AM 1.5G at 100 mW/cm². Similarly, a novel D–A copolymer of perylene diimide (PDI) and dithienothiophene, PPDIDTT-2, was also synthesized by Zhan et al. [172], showing a LBG of 1.46 eV and high electron affinity with a LUMO level of 3.9 eV. The copolymer exhibited broad absorption throughout the visible and into the near-IR region. All PSCs were fabricated with the blend PPDIDTT-2 as acceptor and PT-2 as donor (1:1 wt/wt), showing a PCE over 1 %. Soon after, Tan et al. [173] optimized the all-polymer solar cells based on the narrow band gap alternating copolymer of perylene diimide and bis(dithienothiophene) (PPDIDTT-2). In their work, a polythiophene derivative substituted by a tris(thienylenevinylene) conjugated side chain (PT-2) are used as donor polymer. The optimized device based on the blend of PT-2 and PPDIDTT-2 in the ratio 3:1 (wt/wt) achieved a J _sc of 5.02mA cm⁻² and a PCE of 1.48 %, under AM 1.5 G illumination at 100mW cm⁻². In 2010, Zhou et al. [174] systematically investigated all-polymer solar cells based on six perylene diimide-containing polymers (PX-PDI) as acceptor polymers and two polythiophene derivatives (P3HT and PT-2) as donor polymers. The highest PCE of 2.23 % was obtained in all-PSCs. The photovoltaic results also depicted that the PSC devices based on the PT1/PX-PDI systems have higher V _oc (0.58–0.76 V) than those of the devices based on the P3HT/PX-PDI blends (0.44–0.58 V). Owing to the application of solvent mixtures (toluene/CF, 9:1), the highest PCE of all-PSCs based on PT-2/PC-PDI reached 2.23 % because of the optimal phase separation. This work demonstrated that the solvent strategy and two-dimensional conjugated polymer donors are two effective methods to promote the performance of all-polymer solar cells. Pei et al. [175] developed two perylene diimide (PDI)-based acceptor polymers, r-PDI-diTh and i-PDI-diTh, which were synthesized by introducing a bulky side chain and thereby suppressing the π–π interactions between PDI units in the backbones of acceptor polymers. Therefore, more effective phase segregation of these acceptors with P3HT was realized. When regio-random polymer i-PDI-diTh was blended with P3HT, the PSC device gave a low PCE of 0.45 %. By using a similarly structured but regioregular polymer, r-PDI-diTh, the PCE further increased to 0.94 % because of the limited defect. By employing the inverted device configuration to match the vertical phase separation of donor polymer/acceptor polymer system better, a desirable PCE up to 2.17 % was achieved from the regioregular acceptor polymer, r-PDI-diTh-based PSC devices.

Table 5.10 The photovoltaic results of representative acceptor polymer-based PSC devices

Fig. 5.17

Typical examples of highly efficient acceptor polymers with PCE over 1 %

In 2007, McNeill and coworkers [176] reported efficient photovoltaic diodes which used ambipolar poly((9,9-dioctylfluorene)-2,7-diyl-alt-[4,7-bis(3-hexylthien-5-yl) -2,1,3-benzothiadiazole]-2′,2′′-diyl) (F8TBT) both as acceptor polymer blending with P3HT and as donor polymer blending with PC₆₁BM. In both cases, external quantum efficiencies of over 25 % were achieved. In particular, a PCE up to 1.8 % and extremely high V _oc over 1 V were recorded for the optimized F8TBT/P3HT-based PSC device. Further studies by McNeill et al. [177] demonstrated that the relatively low efficiency of the P3HT/F8BT system can be attributed to poor charge generation and separation efficiencies which result from the failure of P3HT reorganization.

McNeill et al. [178] recently applied PTB7 as the donor polymer in the all-polymer solar cells. A relatively low efficiency of 1.1 % was observed. Clearly, the lack of suitable acceptor polymers has limited the photocurrent and efficiency of polymer/polymer bulk heterojunction solar cells. To overcome the problem, Jenekhe et al. [179] evaluated three naphthalene diimide (NDI) copolymers as acceptor materials in BHJ solar cells. Relatively poor performance (~1.3 %) was observed in the NDI and thiophene copolymer, PNDIT based all-polymer solar cells. PSCs based on an NDI-selenophene copolymer (PNDIS-HD) acceptor and a thiazolothiazole copolymer (PSEHTT) donor exhibited a high PCE of 3.3 %. The observed FF values of 55–60 % were impressively high among all-polymer solar cells and comparable to typical values observed in polymer/PCBM systems. Amazingly, this efficiency was comparable to the performance of PSEHTT/PCBM-based PSC devices. The lamellar crystalline morphology of PNDIS-HD, leading to balanced electron and hole transport in the polymer/polymer blend solar cells, should account for the good performance.

Cheng et al. [180] recently introduced two-dimensional polymer PBDTTT-C-T as donor polymer and proposed a binary additive approach to optimize the performance of PPDIDTT-based all-polymer solar cells, and a dramatically improved PCE up to 3.45 % was observed. Zhou et al. [181] recently reported a high PCE of 3.68 % for the all-polymer solar cell utilizing TTV as the donor polymer and PC-NDI as the acceptor polymer. By introducing a conjugated side chain in the TT unit of donor polymer PTB7, the miscibility of the polymer/polymer blend was greatly improved. In addition, adding a small amount (1 vol.%) of 1,8-diiodooctane (DIO) could increase the aggregation of acceptor polymer, PC-NDI. For the PTB7/PC-NDI-based all-PSCs, the PCE decreased to 1.12 % with V _oc = 0.86 V, J _sc = 3.83 mA/cm², and FF = 34 %. This work demonstrated that two-dimensional conjugated polymer might be ideal donor polymers for all-polymer solar cells because of the superior properties. More recently, Mori et al. [182] fabricated highly efficient LBG donor/acceptor polymer blend solar cells by utilizing TQ-1 as donor polymer and N2200 as acceptor polymer. The device performance was optimized at a donor/acceptor blending ratio of 7:3 (wt:wt) to result in a high J _sc of 8.85 mA/cm², a FF of 55 %, and a V _oc of 0.84 V. The high PCE exceeded 4 %, which was among the highest values reported for all-polymer solar cells. The high photovoltaic performance demonstrated the great potential of polymer/polymer blend solar cells as a promising alternative to polymer/fullerene solar cells.

Because the typical exciton diffusion length of polymer/polymer blends is approximately 10 nm, the larger phase separation length and smaller donor/acceptor interfacial area in the polymer/polymer systems might be the origin of inefficient exciton dissociation and relatively low performance in these systems. To improve the performance of all-polymer solar cells, two strategies could be utilized. One strategy is morphology tuning, for instance, altering the processing solvent to modulate the scale and degree of phase separation of D/A blends. The other strategy is to rationally select and design superior donor polymers.

5.4 Summary and Outlook

In this chapter, an overview of conjugated polymer photovoltaic materials developed during the past two decades are summarized and commented. The basic design considerations of the conjugated polymer photovoltaic materials for the application in polymer solar cells are also discussed along with examples. The rapid progress in novel conjugated photovoltaic polymers demonstrates that there is plenty of room to bring the cost of PSCs down by developing low-cost and highly efficient polymeric photovoltaic materials [183, 184]. Guided by computational calculations and screening [185], further improvements will be realized in the near future by fine optimization of integrated backbones, side chains, and molecular weight [186] based on well-defined highly efficient polymers.