SEARCH

SEARCH BY CITATION

Keywords:

  • bandgap;
  • bulk heterojunction solar cells;
  • donor–acceptor polymer;
  • fullerene;
  • functionalization of polymers;
  • morphology

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information

This review describes the synthesis and photovoltaic performance of donor–acceptor (D–A) semiconducting polymers that have been reported during the last decade. 9,9-Dialkyl-2,7- fluorene, 2,7-carbazole, cyclopenta[2,1-b:3,4-b′]dithiophene, dithieno[3,2-b:2′,3′-d]silole, dithieno[3,2-b:2′,3′-d]pyrrole, benzo[1,2-b:4,5-b′]dithiophene, benzo[1,2 b:4,5 b′]difuran building blocks, and their D–A copolymers are described in this review. © 2012 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem, 2013


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information

Increasing energy consumption and the rising cost of energy are the driving forces to develop new technologies to harvest the solar energy. Organic solar cell technology is a promising candidate for the solar energy conversion compared to its inorganic counterparts due to its low cost, light weight, and potential use in flexible devices.1, 2 Following the first report of “two-layer organic photovoltaic cells” by Tang et al., polymer solar cells have undergone gradual evolution, and power conversion efficiencies above 7% and a highest efficiency of 8.37% have been achieved.3–8

The bulk heterojunction (BHJ) solar cell has become one of the most successful device structures developed to date.9 In this device structure, the donor is an electron rich conjugated polymer, and the acceptor is a soluble fullerene derivative. The donor and the acceptor are blended to form a bicontinuous interpenetrating network, which increases the donor–acceptor (D–A) interfacial area available for the exciton dissociation. The most used acceptor material in organic solar cells is [6,6]-phenyl-C61-butyric acid methyl ester (PC61BM) which is soluble in organic solvents. The morphology of the BHJ is critical to the performance of the solar cell. Excitons, which are coulombically bounded hole–electrons pair, form on the donor (p-type) semiconductor and diffuse. However, the excitons must reach the interface of the donor and acceptor where the charge separation occurs before they relax back into the ground state, or “quench,” through other processes. It has been demonstrated that the exciton diffusion length is on the order of 10–20 nm.10, 11 As a result, an ideal junction is a “BHJ” that is interpenetrated, but has connected phase-separated domains.11 This active layer blend is sandwiched between the high work function indium tin oxide (ITO) anode and the low-work function cathode electrode, such as aluminum.11 The anode is usually coated with poly(3,4-ethenedioxy thiophene):poly(styrenesulfonate) (PEDOT/PSS) hole injection layer.11 The hole injection layer planarizes the ITO and facilitates the collection of holes from the light-harvesting layer to the anode.11

The charge photogeneration process in BHJ solar cells takes place in four steps:12–14 (i) upon the irradiation of light on to the active area, the donor polymer absorbs a photon to generate excitons; (ii) the excitons subsequently migrate until they reach the D–A interface; (iii) the excitons dissociate at the D–A interface into electrons and holes; (iv) the electrons transfer from the lowest unoccupied molecular orbital (LUMO) of the donor to the LUMO of the acceptor and are collected at the cathode, whereas the holes are collected at the anode generating photocurrent and photovoltage.

The power conversion efficiency (PCE) (1) is defined as the ratio of the power produced by the solar cell (Pout) and the optical power of the incident light (Pin). The Pout of a solar cell is the product of three parameters, that is, short circuit current density (JSC), open circuit voltage (VOC), and fill factor (FF).15, 16 The typical current density (J)–voltage (V) curve of a BHJ solar cell device is shown in Figure 1. The JSC, or the device photocurrent at zero bias, depends on the photon absorbance of the active layer, the efficiency of free charge carrier generation, and the charge collection. The VOC (2), or the photovoltage at zero current density, depends on the off-set between the highest occupied molecular orbital (HOMO) of the donor material and the LUMO of the electron acceptor fullerene derivative.17, 18 The FF (3) depends on the balanced charge transport and recombination properties of the active layer.19

  • equation image(1)
  • equation image(2)
  • equation image(3)
thumbnail image

Figure 1. (a) Structure of a bulk heterojunction solar cell; (b) [6,6]-phenyl-C61-butyric acid methyl ester (PC61BM); (c) current–voltage characteristic for dark and light current in a bulk heterojunction (BHJ) solar cell.

Download figure to PowerPoint

Low Band Gap Donor–Acceptor Polymers for Bulk Heterojunction Solar Cells

Figure 2 shows the spectrum of the solar photon flux reaching the surface of the earth.20 The photon flux is distributed over a large range of wavelengths, with a maximum photon flux around 700 nm. To harvest the maximum amount of solar energy, the polymer active layer should absorb a larger number of incident photons.21 The absorption of the active layer is dependent primarily on the conjugated polymer. The mismatch between the absorbance spectra and the solar irradiance spectrum is one of the main reasons for the low PCEs of organic photovoltaics.22 The onset of the absorption spectra of a conjugated polymer is a measure of its bandgap. As shown in Figure 2, due to the relatively large band gap of poly(3-hexylthiophene) (P3HT), a widely used photovoltaic material, its absorbance only covers the wavelength range of 300–600 nm of the solar spectrum.2, 11, 16, 23–26 The absorption spectrum of P3HT matches poorly with the solar emission spectrum, resulting in a collection of a maximum of ∼20% of the solar photons.10, 11 Therefore, the ideal situation would be to design conjugated polymers with lower bandgaps that will absorb photons at wavelengths above 700 nm and thus higher current and higher PCEs could be obtained.22, 27 Two commonly used approaches to generate low bandgap polymers are (i) converting aromatic moieties to quinoid structures along the polymer backbone5, 28 and (ii) synthesizing D–A polymers.15, 22, 29

thumbnail image

Figure 2. The solar spectrum (green line) and the solid state UV–vis absorption spectra of (a) poly(3-hexylthiophene) (red line); (b) poly{4,8-bis(4-decylphenylethynyl)benzo[1,2-b:4,5-b′]dithiophene}(blue line).

Download figure to PowerPoint

In this review, we address only the D–A approach. As shown in Figure 3, the synthesis of D–A polymers results in a lower bandgap due to the orbital mixing of the donor and the acceptor units. The higher energy of the donor group and the lower energy of the acceptor unit results in a reduced bandgap (Fig. 3). The alternation of the electron donating and the electron withdrawing components also increases the double bond character between the units. Thus, the conjugated polymer backbone adopts a more planar configuration, facilitates π-electron delocalization along the polymer backbone, and decreases the bandgap.30 This class of D–A polymers, where the HOMO is mainly located on the donor unit and the LUMO is located on the acceptor unit, is susceptible to tuning of the HOMO and LUMO energy levels by attaching electron withdrawing groups on the acceptor and electron donating groups on the donor unit.31, 32 However, the tuning of the HOMO and LUMO energy levels, plays a major role not only in increasing the JSC but it also affects the VOC. A deeper HOMO level is needed to achieve higher VOC values (2).18 Thus, narrowing the bandgap without sacrificing the efficient charge separation and the VOC is a major hurdle in achieving high efficiency polymer solar cells.

thumbnail image

Figure 3. Orbital mixing in donor–acceptor semiconducting polymers.

Download figure to PowerPoint

The morphology of the active layer is also a crucial factor in achieving high PCEs in solar cells because an efficient charge transfer, transport, and collection strongly depend on the nanoscale morphology of the active layer.33–35 Therefore, if the size of the domains is too large, the excitons will be lost due to exciton decay. By contrast, if the size of the domains is too small, the recombination of charge carriers will be enhanced.15 Moreover, for an efficient charge collection, the donor polymer and acceptor PCBM domains should have a percolated pathway toward the anode and the cathode. To achieve the best morphology for a particular conjugated polymer, studies of the solvent effects, polymer:fullerene ratio effects, annealing effects, different types of additives, and the ratio of donor polymer/additive are required.36–38 It is apparent that JSC, VOC, and FF values of a BHJ solar cell are governed by the active layer donor:fullerene blend. Therefore, the components of the active layer, that is, donor polymer and fullerene acceptor, and the morphology of the active layer are crucial factors that influence BHJ solar cell device performance.

This review describes the D–A semiconducting polymers that have been reported in the last decade. Additionally, this review also summarizes the photovoltaic performance of the D–A polymers and the effect of the morphology on the device performances. Several reviews on band-gap engineering and D–A polymers for organic solar cells have been previously published.35, 39–48

The following donor building blocks will be described in this review: 9,9-dialkyl-2,7-fluorene (FL), 2,7-carbazole (CZ), cyclopenta[2,1-b:3,4-b′]dithiophene (CPDT), dithiene[3,2-b:2′,3′-d]silole (DTS), dithieno[3,2-b:2′,3′-d]pyrrole (DTP), benzo[1,2-b:4,5-b′]dithiophene (BDT), and benzo[1,2-b:4,5-b′]difuran (BDF) (Fig. 4).

thumbnail image

Figure 4. Donor building blocks for the synthesis of donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Fluorene Donor–Acceptor Polymers

FL has been incorporated as a weak donor unit in various D–A polymers for organic solar cells. In this tricyclic unit, the rigid biphenyl ring provides a planar backbone while the carbon at the C-9 position allows for the attachment of different substituents, which can provide enhanced solubility and a good packing in solid state. Polyfluorenes give rise to a low-lying HOMO level and can self-organize as anisotropic liquid crystalline structures at elevated temperatures, thus giving acceptable hole mobilities and provide good VOC and JSC.49, 50 However, an inherent disadvantage of polyfluorenes is the absence of absorption at short wavelengths due to their relatively large bandgaps (2.8–3.0 eV).51 Extensive research has been directed towards the lowering of the bandgap by synthesizing D–A copolymers. Benzothiadiazole, various quinoxalines, and pyrazine have been used as acceptors to generate D–A copolymers (Fig. 5). The optical and electronic properties of FL D–A copolymers are summarized in Table 1.

thumbnail image

Figure 5. Fluorene (FL) donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 1. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Fluorene (FL) D–A Polymers
inline image

In 2003, Anderson et al. reported a D–A copolymer containing FL donor and benzothiadiazole acceptor with a thiopene π-bridge (1a).51, 61 A variety of benzothiadiazole-FL D–A polymers have been synthesized by Suzuki cross-coupling polymerization with varying length of alkyl substituents attached on the FL donor unit (1a1f).51 The variation of the alkyl substituents from linear to branched improved the solubility of the final polymers, and molecular weight increased from Mn∼5K for polymers 1a1d to Mn ∼20K for the polymers 1g1f. However, the length of the alkyl substituents on the 9-position of the FL unit had a negligible influence on the conjugation of the polymer backbone. The HOMO/LUMO energy levels of the polymers were almost identical. All the polymers had a HOMO level of −5.50 eV and an optical bandgap of ∼1.9 eV. The photovoltaic properties of the copolymers 1a–1g (Fig. 5) were evaluated in BHJ solar cells using PC61BM acceptor. For the BHJ with 1g: PCBM = 1:4 weight ratio, a PCE of 2.2% was reported. For polymers 1a1d, PCE values of ∼1.5% were recorded. Further optimization for polymer 1c gave a PCE of 2.7% with 1c:PC71BM = 1:3 weight ratio, and JSC of 6.2 mA cm−2.52 Yang et al. reported a PCE of 4.5% for 1f at 1f: PC71BM = 1: 3 weight ratio (JSC = 9.1 mA cm−2, VOC = 0.97 V and a FF = 0.51).62 The high carrier transport in the active layer with a hole mobility of 3 × 10−5 cm2 V−1s−1 and the higher molecular weight are the factors that supported this relatively high PCE.62 The limited solubility, and hence the lower molecular weights of polymers 1a1d, was due to the unsubstituted thiophene–benzothiadiazole–thiophene unit. Li et al. synthesized three D–A FL polymers (2a2c) by varying the substituents on the thiophene–benzothiadiazole–thiophene unit, which allowed the synthesis of high molecular weight polymers (polymer 2a, Mn = 175 kg mol−1).52 The hexyl chains attached to the 4-position of the thiophene ring in polymer 2a lowered the HOMO energy level by 0.1 eV when compared with 1c. However, the photovoltaic performances were identical with a PCE of 2.2% recorded for the polymer 2a. Pei et al. synthesized D–A polymer 2d with octyloxy substituents at the 3-position of the thiophene unit, for which the HOMO increased by 0.2 eV when compared with 2a and the bandgap was lowered to 1.78 eV.53 This was attributed to the electron donating ability of the alkoxy side substituents. Changing the location of substituents from the thiophene to the benzothiadiazole units (2b2c) had little influence on the HOMO, absorbance, and the charge carrier transport properties of the D–A polymers.53 The polymer 2b at a weight ratio 2b: PCBM = 1:3 gave a PCE of 2.2% and a JSC of 4.4 mA cm−2. Upon changing the solvent from chloroform to dichlorobenzene at 2b: PCBM = 1:3 weight ratio the PCE increased to 3.1% with an increase of JSC to 6.7 mA cm−2. The lower PCE observed for the devices prepared from chloroform was due to the large-scale phase separation (high RMS values) which is not favorable for an efficient charge separation and gave lower current densities.52

A variety of D–A FL polymers have been synthesized by replacing the 2,1,3-benzothiadiazole with various heterocyclic acceptor units, such as [1,2,5]thiadiazolo[3,4-g]quinoxaline 3a3c,54 quinoxaline 4a4c,55, 63 5,56 6,57 2,2′-bi-1,3-thiazole 7,58 pyrazino[2,3-g]quinoxaline 8,59 and thieno[3,4-b]pyrazene 9a9c.56, 60 All of these FL D–A copolymers have been synthesized by Suzuki cross-coupling polymerization.

The use of stronger acceptor [1,2,5]thiadiazolo[3,4-g]quinoxaline (polymers 3a3c), increased the HOMO and the bandgap of the D–A polymers was reduced. However, these polymers exhibited very poor PCEs in BHJ with PC61BM acceptor.54 The energy difference between the LUMO of the D–A polymers (∼−3.90 eV) and the LUMO of PCBM (−3.8 eV) was too low for an efficient charge transfer.15, 54 Therefore, PCBM was substituted for a fullerene with a stronger electron affinity (BTPF60) a LUMO of ∼−4.1 eV, which gave PCE of ∼0.70%. for polymers 3a3c.64, 65

The HOMO energy levels were further lowered for the quinoxaline-FL D–A polymers 4a4c,55 and 5.56 Polymers 4a4c and 5 exhibited moderate PCEs of ∼2%. Lee et al. synthesized a D–A polymer (6) with a selenophene spacer instead of thiophene, which had a HOMO of ∼−5.43 eV and a bandgap of ∼1.88 eV, which is lower than that of 4a4c polymers.57 This was due to the insertion of selenophene spacer, which has a stronger electron donating ability when compared with thiophene. The PCE of polymer 6 was increased from ∼ 2.4% for a blend of 6:PC61BM = 1:4 to 3.3% for a blend 6:PC71BM 1:4. The AFM images showed that the 6:PC61BM showed a smoother morphology than the 6:PC71BM blends.57 D–A FL polymers with acceptors 2,2′-bi-1,3-thiazole (7),58 pyrazino[2,3-g]quinoxaline (8),59 and thieno[3,4-b]pyrazene (9a9c)56 showed similar HOMO and bandgaps when compared with 2,1,3-benzothiadiazole-fluorene polymers, and all showed moderate PCE values. The reported PCE for 7: PC61BM = 1:1 was 0.52% (JSC = 1.76 mA cm−2, VOC = 0.76 V, FF = 0.44), and the PCE of polymer 8: PC71BM = 1:3 was 2.30% (JSC = 6.50 mA cm−2, VOC = 0.81 V, FF = 0.44). The PCEs of 9a9c with PC71BM varied between 0.96% for 9a: PC61BM = 1:6 to 2.20% for 9c: PC61BM = 1:3. The increased PCE was attributed to the difference between the hole mobilities of polymer 9a (μ = 8 × 10−6 cm2 V−1s−1) and 9c (μ = 8 × 10−4 cm2 V−1s−1). This increase of two orders of magnitude in the hole mobilities significantly affected the JSC values and consequently, the PCE of the BHJ solar cells.56

2,7-Carbazole Donor–Acceptor Polymers

Another class of fused ring systems that has attracted much attention in the past few years is the CZ. CZ is an analog to the FL, in which a nitrogen atom bridges the two phenyl rings. The inherent advantages of the CZ polymers are their excellent air stability due to their deep lying HOMO levels and their propensity to form charge-transfer complexes.66 However, due to strong interchain interactions and the planar polymer backbone, the CZ polymers suffer from limited solubility causing inconveniences in device fabrication, which results in lower PCEs.66 A variety of CZ D–A polymers have been synthesized using Suzuki cross-coupling polymerization, among which a few selected D–A copolymers are shown in Figure 6. Optical bandgaps, the HOMO/LUMO energy levels, and the photovoltaic performances of CZ D–A polymers are summarized in Table 2.

thumbnail image

Figure 6. Carbazole (CZ) donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 2. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of 2,7-Carbazole (CZ) D–A Polymers
inline image

Leclerc et al. synthesized the first CZ D–A copolymer (10a) which had a low-lying HOMO of −5.45 eV, a low-lying LUMO of −3.60 eV, and a bandgap of 1.88 eV.67 A preliminary PCE of 3.6% was measured for the polymer 10a at a weight ratio of 10a:PC61BM = 1:4 (JSC = 6.8 mA cm−2, VOC = 0.86 V, FF = 0.56).66 Upon further optimization of the molecular weight of the polymer, the polymer/PCBM blend ratio, and the thickness of the active layer, a PCE of 4.35% was obtained (JSC = 9.42 mA cm−2, VOC = 0.90 V, FF = 0.51) for the polymer with Mn = 20 kg mol−1 (10a:PC61BM = 1:2; active layer thickness = 60 nm).76 Upon replacing PC61BM to PC71BM, for the polymer 10a:PC71BM 1:2 with an active layer thickness of 70 nm, the PCE was further increased to 4.57% (JSC = 10.22 mA cm−2, VOC = 0.89 V, FF = 0.51). Heeger et al. optimized the conditions with several adjustments, adding a titanium oxide as an optical spacer and a hole blocking layer sandwiched between the active layer and the aluminum layer; by increasing the blend ratio to polymer 10a:PC71BM = 1:4; and changing the solvent form chloroform to dichlorobenzene. These changes resulted in an increase of PCE to 6.2% (JSC = 11.9 mA cm−2, VOC = 0.88 V, FF of 0.66).77 Moreover, under these conditions, the BHJ solar cell device had an internal quantum efficiency of 100% indicating that nearly every absorbed photon led to a separated pair of charge carriers, all of which were collected at the electrodes.77 More recently, Cao reported a high PCE of 8.37% for D–A polymer 10a, by incorporating an alcohol/water soluble poly[(9,9-bis(3-(N,N-dimethylamino)propyl)-2,7-fluorene)-alt-2,7-(9,9-dioctylfluorene) as a cathode interlayer in a BHJ solar cell with PC71BM.8

A D–A polymer CZ-oxadiazolo[3,4-c]pyridine (10c), which is structurally similar to polymer 10a, showed a HOMO of −5.55 eV which is 0.10 eV lower than the polymer 10a, and had a smaller bandgap of 1.67 eV.66 Despite a lower HOMO, a low PCE of 0.7% was measured, most likely due to the lower hole mobility of the polymer 10c.66 The polymer 10b, in which the sulfur of [2,1,3]-benzothiadiazole was replaced by an oxygen atom, showed HOMO, LUMO, and optical bandgap almost identical to 10a. However, due to its lower hole mobility (1.0 × 10−4 cm2 V−1s−1) when compared with the polymer 10a (1.0 × 10−3cm2 V−1s−1) polymer 10b exhibited a PCE of 2.4% for a weight ratio of 10b:PC61BM = 1:4.66 The D–A polymer 10e, which is the pyridazine analog of polymer 10b, showed a deep-lying HOMO of −5.70 eV, a deep-lying LUMO of −4.15 eV, and a lower gap of 1.57 eV. The polymer 10e exhibited very poor performance in the BHJ solar cells (PCE = 0.47%). This was attributed to the low LUMO level, which prevented an efficient charge transfer to the PC61BM.68

Zhang et al. synthesized D–A polymer 10f with an electron donating alkoxy group attached to the acceptor moiety which increased the HOMO by 0.2 eV when compared with the polymer 10a.69 Polymer 10f exhibited an initial PCE of 2.75% (JSC = 8.0 mA cm−2, VOC = 0.74 V, FF = 0.55) which was increased to 5.4% upon addition of 1,8-diiodooctane (DIO) additive (JSC = 9.6 mA cm−2, VOC = 0.81 V, FF = 0.69).69 The reported high FF indicated that a balanced charge transport could be achieved by improving the morphology of the active layer blend upon the addition of DIO additive.69

D–A copolymers of CZ have been synthesized by Suzuki cross-coupling polymerization by varying the acceptor units, such as 4,7-bis(thiophen-2-yl)-benzo[c][1,2,5]thiadiazole (11),70 1,3-bis(thien-2-yl)thieno[3,4-c]pyrrole-4,6-dione (12a12b),71, 72 5,8-dithien-2-yl-2,3-diphenyl quinoxaline (13a13b),73, 74 3,6-dithien-2-yl-2,5-di(2-ethylhexyl)-pyrrolo[3,4-c]pyrrole-1,4-dione (14)75 and 5,8-bisthien-2-yl-pyrazino[2,3-d]pyridazine (15a15d).68 Among these D–A polymers, 12a and 13b showed good PCEs. The polymer 12a exhibited an initial PCE of 1.75% (JSC = 4.77 mA cm−2, VOC = 0.86 V, FF = 0.43) for a weight ratio of 12a:PC71BM = 1:2.72 After changing the solvent from chloroform to dichlorobenzene and addition of 3.0% DIO, the PCE of the polymer 12a increased to 3.41% (JSC = 7.29 mA cm−2, VOC = 0.86 V and FF = 0.54). The morphology of the polymer:PCBM blends varied from large phase separated domains to smooth films with no large phase separation obtained by changing the solvent for chloroform to dichlorobenzene and addition of DIO (Fig. 7). This change of morphology-generated differences in the photovoltaic performance with a smoother surface promoting an efficient charge transport between the donor and PC71BM acceptor, which enhanced both JSC and PCE.72

thumbnail image

Figure 7. AFM height images of polymer 12b with PC71BM blend spin-coated with different solutions (image size: 4 × 4 μm2). Copyright © 2011 American Chemical Society.72

Download figure to PowerPoint

For the polymer 13a, the PCE increased from 0.59 to 1.73% on changing the fullerene acceptor from PC61BM to PC71BM. This enhancement of the photovoltaic performances was attributed to the change in the blend morphology, as well as the improved light absorption of the active layer in the visible region due to the presence of the PC71BM fullerene acceptor.73, 74

Donor–Acceptor Copolymers of Bridged Bithiophene Donors

Cyclopenta[2,1-b:3,4-b′]dithiophene, dithieno[3,2-b:2′,3′-d′]silole, and dithieno[3,2-b:2′,3′-d] pyrrole are the bridged bithiophene donor building blocks described in this section. Because of the high bandgaps measured for the polymers with fused dibenzene analogs, such as FL and CZ, their corresponding D–A copolymers showed lower PCE values. These D–A polymers were not ideal for efficient light harvesting. To address this bandgap challenge, the thiophene analogs of the fused dibenzene units were developed.78 The fused thiophene analogs can produce D–A copolymers with higher planarity because the repulsion caused by the protruding hydrogen atoms of the phenyl groups with their neighboring units is relieved by replacing the phenyl with the thiophene units as demonstrated by DFT calculation (Fig. 8).79

thumbnail image

Figure 8. Torsion energy curves of the D–A polymers backbone units calculated with B3LYP/6-311G(d,p) for the monomer model (top). The minimum-energy structures indicated by the red arrows in the torsion energy curves. Color code: gray (C), black (H), blue (N), and yellow (S). Copyright © 2011 American Chemical Society.79

Download figure to PowerPoint

Cyclopenta[2,1-b:3,4-b′]dithiophene D–A Polymers

The thiophene analog of FL, cyclopenta[2,1-b:3,4-b′]dithiophene (CPDT), is obtained by fusing two thiophene rings with one sp3 C atom. CPDT is much more electron rich than FL and hence has a stronger orbital mixing with the acceptor units. The CPDT is a versatile building block for a variety of functional conjugated polymers, because the 4-carbon of the CPDT can be functionalized with a wide range of substituents to fine tune the electronic properties of the resulting copolymers.78 The structure of CPDT D–A reported polymers are shown in Figure 9. The optical and electronic properties of CPDT D–A polymers are summarized in Table 3.

thumbnail image

Figure 9. Cyclopenta[2,1-b:3,4-b′]dithiophene donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 3. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Cyclopenta[2,1-b:3,4-b′]dithiophene (CPDT) D–A Polymers
inline image

In 2005, Brabec et al. synthesized the first CPDT D–A copolymers by Stille cross-coupling polymerization to generate polymer 16a with a HOMO of −5.30 eV and bandgap of 1.40 eV, which exhibited a PCE of 3.5% (JSC = 11.8 mA cm−2 for a weight ratio of 16a:PC61BM = 1:3).91 However, due to the unfavorable phase separation of the polymer fullerene blend, high PCEs were not measured for this polymer.84 Bazan et al. further optimized the device performances of the polymer 16a with 16a:PC71BM = 1:3 with a processing additive 1,8-octanedithiol at 25 mg mL−1 and obtained a PCE of 5.5% with a JSC of 16.2 mA cm−2.38 One factor for the improvement of this PCE is that the alkane dithiols influence the physical interactions between the polymer chains and/or between polymer and fullerene phases (Fig. 10).38

thumbnail image

Figure 10. AFM topography images (10 × 10 mm2) of films cast from chlorobenzene with 24 mg mL−1 of progressively longer alkyl chains. (a) No additive; (b) 1,4-butanedithiol; (c) 1,6-hexanedithiol; (d) 1,8-octanedithiol; (e) 1,9-nonanedithiol. The height scale is 15 nm for all images. Reprinted with permission from Nature Materials.38

Download figure to PowerPoint

Neher et al. further improved the PCE of CPDT-2,1,3-benzothiadiazole D–A polymer (16b) to 6.16% by attaching a fluorine atom to the 2,1,3-benzothiadiazole acceptor unit.80 The incorporation of the fluorine atom simultaneously lowered the HOMO and LUMO of the D–A polymer, while no effect on the optical bandgap was observed.92–94 The HOMO and LUMO of the polymer 16a were −5.02 and −3.58 eV, whereas that of the polymer 16b were −5.15 and −3.71 eV, respectively. The photovoltaic measurements for the polymer 16b showed a VOC 130 mV higher than that of 16a, which was attributed to the increase in the polymer ionization energy upon fluorine attachment.80 JSC increased to 14.08 mA cm−2 for 16b:PC71BM = 1:3 with 2.5% DIO giving a PCE of 6.16%. The fluorine substituent promoted the formation of polymer rich phases and introduced fluorophobicity for the PC71BM, which led to an improvement of the phase separation to generate larger donor/PC71BM domains. Moreover, fluorination induced higher charge carrier generation and reduced geminate recombination, thereby increasing the JSC and FF of the polymer/PC71BM blends.80, 95

Yang et al. synthesized a CPDT-2,1,3-benzoselenodithiophene D–A polymer 16d which had a HOMO of 0.08 eV lower than that of 16a.78 The bandgap was also decreased to 1.35 eV, which is 0.1 eV lower than polymer 16a. Despite these improvements, a lower PCE of 0.66% was measured for the polymer 16d. The hole and electron mobilities of polymer 16d were 2.5 × 10−5 cm2 V−1s−1 and 2.1 × 10−4 cm2 V−1s−1, respectively. However, the electron mobility of the blend was one order of magnitude higher than the hole mobility, and this imbalance contributed to lowering of the PCE.82

A variety of CPDT D–A polymers have been synthesized by replacing the 2,1,3-benzothiadiazole acceptor, and depending on the acceptor strength, the HOMO and the LUMO energy levels and the optical bandgaps have been varied (Fig. 9). Thieno[3,4-c]pyrrole-4,6-dione (19),85 bis(bithienyl)-4,4′-dihexyl-2,2′-bithiazole (23)87 and 3,6-bis(4-hexylthienyl-2-yl)-s-tetrazine (24a24e)88, 89 acceptors have been used to generate CPDT D–A polymers that showed better performances in BHJ solar cells with PCEs of ∼3%. For example, polymer 19 reported by Guo et al. had a HOMO of −5.26 eV, which was 0.14 eV lower than the polymer 16a. Polymer 19 exhibited a PCE of 3.15% (JSC = 8.29 mA cm−2, VOC = 0.76 V, FF = 0.50) with PC71BM.85 Polymer 24e had a low HOMO level of −5.37 eV and a bandgap of 1.86 eV.89 This HOMO is 0.35 eV lower than that of 2,1,3-benzothiadiazole analog (16a). Polymer 24e:PC71BM = 1:3 gave a PCE of 1.03% (JSC = 4.83 mA cm−2, VOC = 0.60 V, FF of 0.35), which increased to 5.53% on addition of DIO additive (JSC = 12.5 mA cm−2, VOC = 0.75 V, FF = 0.59).88 The morphology of the polymer blend was demonstrated to play a pivotal role in the device performances. Upon the addition of the processing additive DIO, which has a higher boiling point than that of the host solvent and a good solubility only for the PC71BM, the tendency of PC71BM to crystallize was reduced due to the slower evaporation of DIO when compared with the host solvent.88

Dithieno[3,2-b:2′,3′-d]silole D–A Copolymers

Another fused-ring thiophene donor building block that attracted much attention in the past few years is DTS, in which the C atom at the 4-position has been replaced with a silicon atom. DTS D–A polymers have lower LUMO levels and have shown better performances in optoelectronic devices when compared with CPDT analogs. The changes in the electronic properties are due to the effective orbital mixing of the Si[BOND]C π* orbital with the π* orbital of the butadiene fragment allowing for a low-lying LUMO and a low bandgap.96, 97 Moreover, the introduction of silicon stabilizes the diene HOMO when compared with the carbon counterpart, enhancing the ambient stability of the silole polymers.98 A variety of DTS D–A copolymers with different acceptor units have been synthesized by Stille or Suzuki cross-coupling polymerizations (Fig. 11). The optical and electronic properties of DTS D–A copolymers are summarized in Table 4.

thumbnail image

Figure 11. Dithieno[3,2-b:2′,3′-d]silole (DTS) donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 4. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Dithieno[3,2-b:2′,3′-d]silole (DTS) D–A Polymers
inline image

In 2007, Tao reported the first DTS D–A copolymer containing 4,7-bis(2-thienyl)-2,1,3-benzothiadiazole as acceptor unit (29a).107 However, this polymer showed a very poor PCE of 0.18% which was due to the low molecular weight and the low solubility of the polymer.107 In 2008, Hou reported polymer 27 which had a HOMO of −5.05 eV and a LUMO of −3.27 eV.29 This polymer exhibited a PCE of 3.7% (JSC = 10.9 mA cm−2, FF = 0.50) for PC71BM 1:1 weight ratio.29 Upon annealing, the devices at 140 °C, the PCE of the polymer 27 increased to 5.1% with an increase in JSC to 13.06 mA cm−2 and FF to 0.62. Moreover, the charge carrier mobility of the polymer 27 was also increased upon annealing. The lengthening of the C[BOND]Si bond in comparison to the C[BOND]C bond placed the solubilizing alkyl chains further apart from the thiophene rings. This allowed a stronger π-stacking interaction, which enhanced the solid packing leading to a higher charge carrier mobility, and ultimately higher JSC and PCE.29, 108

Kim et al. synthesized polymers by varying the length of the alkyl chain substituents on the Si atom (28a, 28b).97 Regardless of the different lengths of the alkyl substituents, polymers 28a and 28b exhibited HOMO of ∼−5.55 eV and a low-lying LUMO of ∼−3.66 eV. The polymers had a high VOC of ∼0.90 V. By contrast, the JSC of the polymers showed a significant difference depending on the alkyl substituent. Polymer 28a with methyl substituents showed the highest JSC of 4.29 mA cm−2, a VOC of 0.94, and a FF of 0.45, resulting in a PCE of 1.84% (28a:P71BM = 1:4 weight ratio).97 Polymer 28b with butyl substituents showed a lower JSC of 2.03 mA cm−2, VOC of 0.81 V and a FF of 0.36 resulting in a PCE of 0.59% (28b:PC71BM = 1:4).97

Yang et al. reported polymers 29a29c, by varying the substituents on the silicon atom.98 As mentioned earlier by Tao et al., the polymer with two hexyl substituents (29a) had very low solubility in common organic solvents.107 The HOMO and LUMO levels of polymers 29a and 29b were quite similar, indicating that the length of the alkyl substituents does not significantly affect the HOMO/LUMO of the polymers. Therefore, both the polymers gave similar VOC values for the polymer: PCBM = 1:1 weight ratio. However, polymer 29a gave a PCE of 3.43% with a JSC of 10.67 mA cm−2, VOC of 0.62 V, and a FF of 0.51, while the polymer 29b gave a PCE of 2.95% with a JSC of 9.76 mA cm−2, VOC of 0.60 V, and a FF of 0.50.107 Tao et al. synthesized D–A polymer 30 with thieno[3,4-c]pyrrole-4,6-dione acceptor, which gave a high PCE of 7.3%.7 Polymer 30 had a deep HOMO of −5.57 eV and a deep-lying LUMO of −3.88 eV. This polymer achieved a preliminary PCE of 6.2% with a JSC of 10.95 mA cm−2, VOC of 0.90 V, and a FF of 0.63 for 30:PC71BM = 1:2 in dichlorobenzene with 3% DIO additive. Upon changing the solvent from dichlorobenzene to chlorobenzene, a PCE of 7.3% with a JSC of 12.2 mA cm−2, VOC of 0.88 V, and a FF of 0.68 were obtained.7

Li et al. synthesized DTS D–A copolymer containing thiazolothiazole as the acceptor (32a, 32b) and varied the position of the alkyl substituents on the thiophene spacer group.100, 101 Both polymers had similar molecular weights (Mn ∼ 6 kg mol−1) and comparable HOMO of −5.04 eV. However, polymer 32b with the alkyl chains substituted at the 4-position of the thiophene spacer group had a LUMO of −3.41 eV, which is 0.6 eV lower than polymer 32a, for which the alkyl group was attached at the 3-position of the thiophene unit. Polymer 32b had a higher JSC of 11.25 mA cm−2 and a PCE of 5.88%. Polymer 32a had a JSC of 11.20 mA cm−2 and a PCE of 5.59%.101 To investigate the effect of the substituent and the topology on the photovoltaic properties of the DTS-thiazolothiazole D–A polymers, Jenekhe et al. synthesized polymers 32c32e.102 Due to the electron donating nature of the alkoxy substituents of polymer 32e the HOMO increased by 0.1 eV when compared with 32c and 32d. The polymer 32d with the ethylhexyl substituents showed the highest PCE of 5.00% at weight ratio polymer:PC71BM = 1:2. This high PCE was attributed to the fact that the ethylhexyl substituted polymer 32d had face-on-orientation of the π−π stacking in polymer relative to the substrate, which increased the vertical hole transport.102 The photovoltaic properties of the polymers 3435 were examined by Li et al. in blends with PC71BM and the PCE of the polymer 35 (PCE = 2.10%) was higher than that of 34 (PCE = 0.58%).104 The higher PCE of 35 was explained by the smooth morphology of polymer PC71BM blend.104 A similar effect was observed for polymers 37 and 38 reported by Li et al. where the polymer with thiophene spacer exhibit higher PCE of 3.80% when compared with polymer 37 which gave a PCE of 1.64%.106

Dithieno[3,2-b:2′,3′-d]pyrrole Donor–Acceptor Copolymers

Another class of structurally related analogs of CPDT is dithieno[3,2-b:2′,3′-d]pyrrole (DTP), in which the nitrogen atom is substituted at the 4-position. As discussed earlier, the atom substituted at the 4-position plays a pivotal role in the electronic properties of the resulting polymer. When the nitrogen is substituted at the 4-position due to its stronger electron donating ability, the HOMO of the polymer is increased and the bandgap is reduced. However, due to the higher HOMO level, DTP D–A polymers are not stable in air and can be easily oxidized.109, 110 DTP D–A copolymers with various acceptor units have been synthesized by Stille cross-coupling polymerization and their structures are shown in Figure 12. The optical and electronic properties of DTP D–A copolymers are summarized in Table 5.

thumbnail image

Figure 12. Dithieno[3,2-b:2′,3′-d]pyrrole (DTP) donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 5. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Dithieno[3,2-b:2′,3′-d]pyrrole D–A Polymers
inline image

Wang et al. synthesized a series of benzothiadiazole DTP copolymers with varying alkyl chain substituents on the N atom. (39a39c).111 The three polymers have almost the same HOMO/LUMO energy levels, and optical bandgaps, suggesting that the length of the alkyl chain does not have a significant impact on the electrochemical properties of the polymers. However, the PCE of the polymers improved with increasing the alkyl chain length, due to the enhanced film morphology. The polymer with pentyl chain substituents (39c) gave the highest PCE of 2.80% with a JSC of 11.9 mA cm−2 (VOC = 0.54 V, FF = 0.44). Wang et al. reported that with the decrease in the length of the alkyl chains, the morphology of the thin films varied from larger domains (400–900 nm) with large phase separation to smaller domains (50 nm) with uniform phase separation.111

For the DTP D–A copolymers, the optical bandgaps are relatively low, but the LUMO and the bandgaps depend on the choice of the acceptor unit. However, the HOMO of the DTP D–A copolymers varied slightly with an average HOMO in the range of −4.90 eV. Because of this relatively high HOMO, the VOC of the DTP D–A copolymers in BHJ solar cells are low, resulting in lower PCE values.

Benzo[1,2-b:4,5-b′]dithiophene Donor–Acceptor Polymers

Benzo[1,2-b:4,5-b′]dithiophene (BDT) is the most used donor building block for the synthesis of semiconducting polymers used in OFETs and BHJ solar cells.45, 114–116 The BDT building block offers two advantages: the fused BDT allows the incorporation of substituents on the central benzene core while maintaining the planarity of the two thiophene units, and the symmetric nature of the BDT monomer eliminates the need of controlling regioregularity during the polymerization.114 Moreover, due to the large planar conjugated structure of the BDT unit, the polymer can easily forms π−π stacks, which renders high charge carrier mobility.78 Taking these advantages into account, many groups have synthesized a variety of BDT D–A copolymers with tunable bandgaps and HOMO/LUMO energy levels. The most common strategy is the structural variation at the 4- and 8-positions of the BDT by the addition of alkoxy, alkyl, alkylthiophene, and alkylbithiophene substituents. Moreover, copolymerization of BDT with different acceptor monomers generated a plethora of BDT D–A copolymers. The BDT D–A copolymers were synthesized by Stille and Suzuki cross-coupling polymerizations.

BDT Donor–Acceptor Polymers Containing Alkyl and Alkoxy Substituents

Hou et al. first introduced the BDT-conjugated polymers, by copolymerizing alkoxy-substituted BDT with electron rich and electron deficient aromatic units. Among these are two D–A copolymers, polymer 46a with benzo[1,2,5]thiadiazole, and 46b with benzo[1,2,5]selenadiazole acceptors (Fig. 13).78 The presence of the less electronegative selenium in polymer 46b increased the HOMO compared to 46a. However, both polymers exhibited low PCEs (0.90% for 46a; 0.90 and 0.18% for 46b).78 These polymers were tested without any post-treatment which indicated that BDT could be a promising candidate for BHJ solar cells. Some of the alkoxy and alkyl substituted-BDT D–A polymers are shown in Figure 13, and their electronic properties and photovoltaic performances are outlined in Table 6.

thumbnail image

Figure 13. Benzo[1,2-b:4,5-b′]dithiophene (BDT) donor–acceptor semiconducting polymers (alkyl and alkoxy substituents, 43a-48e).

Download figure to PowerPoint

Table 6. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Benzo[1,2-b:4,5-b′]dithiophene (BDT) D–A Polymers (Polymers 43a−48e)
inline image

Yu et al. synthesized a series of BDT copolymers containing thieno[3,4-b]thiophene carboxylate acceptor with varying alkyl and alkoxy substituents (43a43d).93, 95 The polymer 43a with a HOMO level of −5.04 eV and an optical bandgap of 1.62 eV showed a preliminary PCE of 4.76% (JSC = 12.5 mA cm−2, VOC = 0.58 V, FF = 0.51) for a weight ratio of 43a: PC61BM = 1:1.95 Upon further optimization by replacing the PC61BM with PC71BM for weight ratio 43a: PC71BM = 1:1.2, the PCE increased to 5.6% (JSC = 15.6 mA cm−2, VOC = 0.56 V, FF = 0.63).117 This enhancement of PCE was attributed to the following factors: (i) the higher absorption coefficient of PC71BM which improved the light absorption of the active layer in the visible region; (ii) the high hole mobility and the balanced carrier mobility in the blend led to a high FF; and (iii) the interpenetrating network in the blend film favored the exciton dissociation in the interfacial area and improved the charge transport.117 When the alkoxy groups on the 4- and 8-positions of BDT were replaced with alkyl substituents (43c), the HOMO was lowered by 0.14 eV and the LUMO was lowered by 0.07 eV, resulting in a higher bang gap of 1.75 eV when compared with polymer 43a.93 The deeper HOMO level was beneficial to obtain a large VOC of 0.72 V for the polymer 43c at weight ratio 43c:PC61BM = 1:1. A high PCE of 5.10% and a high JSC = 13.90 mA cm−2 were measured for 43c.93 The introduction of the electron deficient fluorine atom on the acceptor unit for polymer 43d further lowered the HOMO to −5.12 eV and gave a VOC of 0.76 V.93 Upon using a mixed solvent system of DIO and dichlorobenzene, the PCE of 43d was increased to 6.10% with an increase of JSC from 9.20 to 13.0 mA cm−2. The change in the morphology of the blends from aggregates with larger domains to smoother films improved the charge transport and hence increased the JSC and the PCE.93 The attachment of fluorine atoms on both the acceptor and BDT donor unit (44c) further decreased the HOMO to −5.48 eV and LUMO to −3.59 eV, without affecting the bandgap.95 The polymer 44c gave a lower PCE of 2.7% when compared with polymers 43a44b. The lower PCE was due to the fluorination which provided a driving force for the phase separation of the polymer/fullerene blend film with no bicontinuous network, resulting in a poor charge carrier mobility for polymer 44c (μ = 7 × 10−5 cm2 V−1s−1).95

Coffin et al. synthesized polymer 46c for which, due to the presence of the electronegative oxygen atom, a deeper HOMO was obtained when compared with 46a.119 A preliminary PCE of 1.41% was obtained for BHJ with PC71BM.119 A dramatic increase of PCE to 5.65% was observed upon the addition of 1-chloronaphthalene (2 wt %) as an additive.119 Several groups have synthesized D–A copolymers by bridging the BDT and the benzothiadiazole acceptor through a conjugated moiety such as thiophene, alkylthiophene, and alkoxythiophene.111–113 A broadening of the absorption spectrum and improved hole mobility were observed for copolymers 47c47i. Additionally, the alkyl/alkoxy groups enhanced the solubility of the D–A copolymers, resulting in higher molecular weights. The polymer 47b has a higher HOMO level than the polymers 47c47i. For polymers 47c–47i the HOMO and the LUMO have values of ∼−5.25 and ∼−3.53 eV, respectively.121–123 These polymers exhibited PCEs greater than 2% with polymer 47d having the highest PCE of 4.02%.122 You et al. reported BDT D–A copolymers from alkyl-substituted BDT and benzothiadiazole with thiophene bridging groups (48a48e).6, 124, 125 These polymers showed deep lying HOMO of ∼−5.40 eV, which are lower than that of the polymers 47a47i. These polymers gave high VOC and PCE of ∼5%. A high PCE of 7.2% (JSC = 12.91 mA cm−2, VOC = 0.91 V, FF = 0.61) was measured for polymer 48e for a weight ratio 48e:PC61BM = 1:1.6 The introduction of fluorine atom lowered the HOMO, enhanced the VOC, JSC, and the FF of the BHJ due to the balanced charge transport and the improved active layer morphology.6, 95

You et al. synthesized polymers 49a49b (Fig. 14 and Table 7) by replacing the benzothiadiazole with benzotriazole acceptor, which gave higher HOMO levels of ∼−5.30 eV when compared with 48a48e.4 Polymer 49b with fluorine substituent, gave a high PCE of 7.1% with PC61BM (JSC = 11.83 mA cm−2, VOC = 0.71 V, FF = 0.73). Zhan and Li reported BDT D–A copolymers with thiazole and thiazolothiazole acceptors (50, 51a51f).127, 128 Despite attaching various thiophene or alkylthiophene units on either side of the acceptor, the highest PCE achieved was only 3.20%.128

thumbnail image

Figure 14. Benzo[1,2-b:4,5-b′]dithiophene (BDT) donor–acceptor semiconducting polymers (alkyl and alkoxy substituents, 49a-54i).

Download figure to PowerPoint

Table 7. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Benzo[1,2-b:4,5-b′]dithiophene (BDT) D–A Polymers (Polymers 49a−54i)
inline image

Li et al. synthesized BDT and diketopyrrolopyrrole D–A copolymers 52a52c (Table 7).130 Polymer 52a gave a PCE of 4.10% with PC71BM.130 N-Alkylthienopyrroledione has also used as an acceptor for the synthesis of BDT D–A copolymers (53a53e, Table 7). Leclerc et al. synthesized polymer 53a which showed an initial PCE of 5.5% at weight ratio 53a:PC71BM = 1:1.131 Frechet et al. further optimized the device performance by using DIO additive and increased the PCE to 6.8% (JSC = 11.50 mA cm−2, VOC = 0.85 V, FF = 0.68).133 Several groups synthesized BDT D–A copolymers with thienopyrroledione as the acceptor unit and thiophene or alkylthiophene as the bridging conjugated moiety (54a54i).134, 135 The efficiencies did not improve with the additional conjugation along the polymer backbone and the highest reported PCE was 3.9% for polymer 54d.134

BDT Donor–Acceptor Polymers Containing Conjugated Substituents

The choice of the substituents attached to the 4- and 8- positions of the BDT moiety play a crucial role in controlling the solubility, molecular weight, optical, and electronic properties of conjugated polymers. Phenylethynyl, triisopropylsilylethnyl, alkylthienyl, and alkylbithienyl substituents were attached on the BDT and used as donors to generate D–A BDT copolymers.

Stefan et al. synthesized a series of BDT D–A copolymers with electron withdrawing phenylethynyl substituents attached to the BDT core (55a55d, Fig. 15 and Table 8).115 The bandgaps of the copolymers varied from 2.3 eV (55a) to 1 eV (55d) depending on the strength of the acceptor unit. However, the HOMO was relatively unchanged and the change in the bandgap was due to the change in the LUMO of the synthesized D–A copolymers.115 Among these polymers the lowest bandgap of 1eV was measured for 55d with benzobisthiadiazole as the acceptor unit.115 Cho et al. reported BDT D–A copolymers 56a56b with octyl phenylethynyl substituents on the BDT core.136 Both polymers had deep-lying HOMO of ∼-5.60 eV. These polymers also had a deep lying LUMO and generated a downhill driving force for the electron transport in BHJ solar cells.15, 136

thumbnail image

Figure 15. Benzo[1,2-b:4,5-b′]dithiophene (BDT) donor–acceptor semiconducting polymers (ethynyl substituents, 55a-57c).

Download figure to PowerPoint

Table 8. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Benzo[1,2-b:4,5-b′]dithiophene (BDT) D–A Polymers (Polymers 55a–57c)
inline image

Zhan et al. and Cho et al. synthesized BDT D–A copolymers with triisopropylsilylethynyl (TIPS) substituents and thiazolothiazole (57a),137 bithiazole (57b),136 and bisthienylbenzothiadiazole (57c) acceptors.136 For the polymer 57a, an optical bandgap of 1.94 eV with a HOMO level of −5.30 eV and a PCE of 4.33% for the 57a:PC71BM = 1:1 weight ratio (JSC = 9.77 mA cm−2, VOC = 0.89 V and FF = 0.50) was obtained. The polymer 57b showed a deep-lying HOMO of −5.91 eV, but this polymer showed lower PCE values due to the lower LUMO level (-3,52 eV). The polymer 57c showed a PCE of 4.61% for the BHJ device configuration of ITO/MoO3/polymer:PC71BM/ZnO/Al.136

Alkylthienyl and alkylbithienyl were attached to BDT and used as donor building blocks for the synthesis of BDT D–A copolymers (Fig. 16 and Table 9). Yang et al. reported the synthesis of polymer 58a with dialkylthienyl-substituted BDT and dithienylbenzothiadiazole acceptor.138 This polymer showed a broader absorption in the visible region, a HOMO of ∼−5.30 eV, and a bandgap of 1.87 eV. Because of the deep HOMO a high VOC of 0.90 V was obtained in a BHJ device for weight ratio 58a:PC71BM = 1:1 (JSC = 10.77 mA cm−2, VOC = 0.92 V, FF = 0.58), resulting in a PCE of 5.66%.138

thumbnail image

Figure 16. Benzo[1,2-b:4,5-b′]dithiophene (BDT) donor–acceptor semiconducting polymers (thienyl substituents, 58a-65).

Download figure to PowerPoint

Table 9. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Benzo[1,2-b:4,5-b′]dithiophene (BDT) D–A Polymers (Polymers 58a–67)
inline image

Dai et al. synthesized polymers 58b and 58c with fluorine and alkoxy substituents attached on the electron acceptor unit, respectively.139 The combination of electron rich alkylthienyl substituted BDT donor moiety and the electron deficient fluorine on the acceptor unit generated polymer 58b with HOMO of ∼−5.41 eV and a PCE of 6.21%. Since the presence of the fluorine atom also lowered the LUMO, a bandgap of 1.69 eV was measured for polymer 58b.139 Several groups reported the synthesis of alkylthienyl-substituted BDT with a variety of acceptors such as, thiazolothiazole (59),140 alkoxycarbonyl-substituted thieno[3,4-b]thiophene (60),141 alkylcarbonyl-substituted thieno[3,4-b]thiophene (61),141 naptho[1,2-c:5,6-c]bis[1,2,5]thiadiazole (62),142 benzotriazole (63),143 5,5′-dibenzo-[c][1,2,5]thiadiazole (64),144 and diketopyrrolopyrrole (65).145 Among these, polymer 61 gave the highest PCE of 7.59% (JSC = 17.48 mA cm−2, VOC = 0.74 V, FF = 0.58) for a weight ratio 61:PCBM = 1:1.5 with 3% DIO additive.141

Stefan et al. reported two BDT D–A copolymers with bithienyl-substituted BDT and benzothiadiazole and thienopyrroledione as acceptors (6667).146 Six alkyl substituents per each BDT repeating unit improved the solubility of polymers and allowed the synthesis of relatively high molecular weight polymers. Because of the conjugated bithienyl substituents, both the polymers showed a deep-lying HOMO of ∼-5.53 eV and a bandgap of 1.70 eV. Because of the deep-lying HOMO a very high VOC of 1.00 V was measured for polymer 66 with a PCE of 2.52% for a weight ratio of 66:PC61BM = 1:3 (JSC = 5.36 mA cm−2, FF = 0.47).146 However, for these polymers the addition of the DIO additive lowered the PCEs. This behavior was attributed to the difference in morphology of the active area of the polymer/PCBM blend, which changed from a smooth surface to a rougher surface with large domains upon the addition of DIO (Fig. 17).146

thumbnail image

Figure 17. 3D TMAFM phase images of solar cell devices: (a) polymer 66/PCBM (1:3 w/w) active layer without additives; (b) polymer 66/PCBM (1:3 w/w) active layer with 1,8-diiodooctane; (c) polymer 67/PCBM (1:4 w/w) active layer without additives; (d) polymer 67/PCBM (1:4 w/w) active layer with 1,8-diiodooctane; scan size 5 × 5 μm (spin-cast from chloroform). Copyright © 2012 American Chemical Society.146

Download figure to PowerPoint

Benzo[2,2-b:4,5-b′]difuran D–A Polymers

BDF is a structurally similar analog to BDT for which the thiophene is replaced by a furan ring. The smaller radii and the higher electronegativity of oxygen is expected to give BDF polymers a good packing, a more planar polymer backbone, and a deep-lying HOMO.147 The BDF homopolymers showed average hole mobilities.148 The reported BDF D–A copolymers are shown in Figure 18 and their optoelectronic properties are summarized in Table 10.

thumbnail image

Figure 18. Benzo[2,2-b:4,5-b′]difuran (BDF) donor–acceptor semiconducting polymers.

Download figure to PowerPoint

Table 10. Optoelectronic Properties, Hole Mobility, and the Photovoltaic Performance of Benzo[2,2-b:4,5-b′]difuran (BDF) D–A Polymers
inline image

Huo et al. reported the first BDF D–A copolymer (68a), which showed a HOMO of −5.10 eV and a LUMO of −3.24 eV.147 When compared with the BDT D–A copolymer, BDF polymers have a lower bandgap. A preliminary PCE of 4.66% (JSC = 10 .00 mA cm−2, VOC = 0.78 V, FF = 0.54) was obtained for a weight ratio of 68a:PC71BM = 1:1.5. Upon annealing the BHJ devices at 90 °C the JSC increased from 10.00 to 11.77 mA cm−2, and a PCE of 5.01% was obtained.147 The increase of JSC was due to the smoother morphology obtained after annealing the devices.147 Liu et al. synthesized a series of BDF D–A copolymers with bisthienylbenzothiadiazole (68b), bisthienylbenzooxazole (68c), and bisthienylbenzotriazole (69) acceptors.149 The polymer 68c showed the lowest HOMO of −5.19 eV with a low bandgap of 1.70 eV. Polymer 69 had a HOMO of −4.99 eV and a bandgap of 1.93 eV. The difference in the bandgaps of 68b and 69 was attributed to the weaker electron withdrawing benzotriazole unit when compared with the other acceptor unit.149 Among these polymers, 68c had the highest PCE.149

Two D–A copolymers with thienothiophene acceptor were synthesized by Huo et al. (70 and 71).150 The polymer with alkoxy substituted-BDF (70) showed an initial PCE of 4.52% and upon the addition of 3% DIO, the PCE value was increased to 5.22%. The lower PCE was attributed to the larger aggregation of PC71BM with domain sizes varying in the range of 100–350 nm for blends without DIO. Upon the addition of DIO, a smoother morphology with smaller domain sizes was obtained which resulted in a higher JSC. A similar effect was observed for the alkylthienyl-substituted BDF polymer (71) upon the addition of DIO, which resulted in an increase of PCE from 4.54 to 6.43%.150

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information

D–A alternating structures have been shown to generate semiconducting polymers that have a lower bandgap due to the orbital mixing of the donor and acceptor components. Various combinations of donor and acceptor building blocks have been used to generate D–A copolymers with some of them giving excellent performance in BHJ solar cells with fullerene electron acceptors. However, not all the D–A combinations generated the expected decrease in the bandgap. A more judicious selection of the D–A pairs is highly desirable but currently there are no well-established pairing rules. A potential guidance in selecting D–A pairs can be provided by DFT calculations that can give an estimate bandgap, which ideally should be compared with the experimental values obtained for the synthesized D–A copolymers.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information

Financial support for this project from NSF (Career DMR-0956116) and Welch Foundation (AT-1740) is gratefully acknowledged. The authors gratefully acknowledge the NSF-MRI grant (CHE-1126177) used to purchase the Bruker Advance III 500 NMR instrument.

REFERENCES AND NOTES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information
Thumbnail image of

Ruvini S. Kularatne received her B.Sc. in Chemistry from University of Peradeniya in Sri Lanka in 2008. She joined Prof. Mihaela Stefan's research group in 2009 as a Ph.D. student. She is currently a fourth year graduate student and her research targets the synthesis of novel benzodithiophene donor–acceptor polymers for organic solar cell applications.

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information
Thumbnail image of

Harsha D. Magurudeniya received his B.Sc. in Chemistry from University of Peradeniya in Sri Lanka in 2008. He joined Stefan's group in 2009 as a Ph.D. student at the University of Texas at Dallas. He is currently a fourth year graduate student and his research targets the synthesis of nickel catalysts for the polymerization of bulky monomers using Grignard metathesis (GRIM) polymerization.

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information
Thumbnail image of

Prakash Sista received an integrated M.Sc. in Chemistry (B.Sc. and M.Sc.) from the Indian Institute of Technology Bombay in 2003. He has received M.S. in Chemistry from University of San Francisco in 2008 working on electron transfer reactions of ruthenium complexes. He received his Ph.D. degree from University of Texas at Dallas in 2011 working in Stefan's group.

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information
Thumbnail image of

Michael C. Biewer is an Associate Professor of Chemistry at the University of Texas at Dallas. He received his Ph.D. degree from Yale University working in the group of J. Michael McBride. He subsequently worked as a postdoctoral researcher in the group of David M. Walba at the University of Colorado at Boulder before joining the faculty at the University of Texas at Dallas.

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. CONCLUSIONS
  5. Acknowledgements
  6. REFERENCES AND NOTES
  7. Biographical Information
  8. Biographical Information
  9. Biographical Information
  10. Biographical Information
  11. Biographical Information
Thumbnail image of

Mihaela C. Stefan received her Ph.D. degree in Chemistry from Politehnica University Bucharest. She joined the Department of Chemistry at the University of Texas at Dallas in 2007. She received the NSF Career Award in 2010, the NS&M Outstanding Teacher Award in 2009, and the Inclusive Teaching Diversity Award in 2012. Her research group is developing novel polymeric materials for organic electronics and drug delivery applications.