Benzothiadiazole-based Conjugated Polymers for Organic Solar Cells

Benzothiadiazole (BT) is an electron-deficient unit with fused aromatic core, which can be used to construct conjugated polymers for application in organic solar cells (OSCs). In the past twenty years, huge numbers of conjugated polymers based on BT unit have been developed, focusing on the backbone engineering (such as by using different copolymerized building blocks), side chain engineering (such as by using linear or branch side units), using heteroatoms (such as F, O and S atoms, and CN group), etc. These modifications enable BT-polymers to exhibit distinct absorption spectra (with onset varied from 600 nm to 1000 nm), different frontier energy levels and crystallinities. As a consequence, BT-polymers have gained much attention in recent years, and can be simultaneously used as electron donor and electron acceptor in OSCs, providing the power conversion efficiencies (PCEs) over 18% and 14% in non-fullerene and all-polymer OSCs. In this article, we provide an overview of BT-polymers for OSCs, from donor to acceptor, via selecting some typical BT-polymers in different periods. We hope that the summary in this article can invoke the interest to study the BT-polymers toward high performance OSCs, especially with thick active layers that can be potentially used in large-area devices.


INTRODUCTION
Solar cells, which are capable of converting solar energy into electrical energy, are considered as one of the most promising and efficient technologies to alleviate the global energy crisis. [1] In recent years, organic solar cells (OSCs) have attracted extensive attention in the photovoltaic field due to their advantages, such as flexibility, light weight, facile fabrication technique and low cost compared to the conventional inorganic solar cells. [2−5] The photoactive layers, which consist of donor and acceptor materials, play important roles in OSCs. [6−13] Fullerene derivatives were usually used as electron acceptor due to their high electron mobility and three-dimensional electron transport properties. Although the power conversion efficiencies (PCEs) of fullerene-based OSCs have been enhanced to over 11%, the inherent disadvantages of fullerene derivatives, such as the limited chemical structures and the difficulty to tune the energy levels, hamper the further improvement of photovoltaic properties. For comparison, non-fullerene acceptors have been demonstrated to out-perform fullerenes in OSCs, [14−19] in which the PCEs have reached up to 18%. [20] Additionally, the donor materials also play important roles in OSCs, since they are responsible for matching with the electron acceptor to convert photons into free charges. [21−26] Nowadays, the donor polymers are usually constructed via donor-acceptor (D-A) motif, in which the electron-donating and electron-deficient units are incorporated into one polymer. Therefore, the intramolecular charge transfer between donor and acceptor segments can be generated, resulting in the lower optical band gap. Simultaneously, the other properties of the polymers, such as energy levels, crystallinity and charge transport properties, can also be effectively adjusted. Therefore, D-A conjugated polymers have dominated the development of OSCs. Among them, benzothiadiazole (BT) as electron-deficient unit has been widely used to design conjugated polymers for high performance OSCs. [27−29] BT-based monomers can be simply prepared and modified with F atom or different side chains to tune the optoelectronic properties and solubility (Scheme 1). [30−34] By choosing copolymerized units with different properties, BT-based copolymers can be simply synthesized via Suzuki or Stille coupling polymerization. BT-based polymers usually own high crystallinity, good charge transport characteristics, low band gaps and excellent optoelectronic properties. Based on those excellent properties, researchers can construct BT-polymers with tunable physical and optoelectrical properties toward efficient OSCs.
Since Dhanabalan et al. reported the first BT-based conjugated polymer for application in OSCs in 2001, [35] BT-based  unit with two alkyloxy side chains to improve the solubility of the polymer. The corresponding polymer HXS-1 (named as P4) provided a PCE of 5.4% in solar cells. [30] This design motif was then widely used in developing BT-polymers with prominent photovoltaic performance. [52] Benzodithiophene (BDT) is another copolymerized unit that is widely used in BT-polymers. For example, in 2011, You et al. firstly introduced fluorine atoms into the BT core to construct the BT-polymer P7 with BDT as comonomer (Fig. 2). [32] The introduction of fluorine atoms can effectively lower the frontier energy levels and improve the crystallinity of the polymer, resulting in high open-circuit voltage (V oc ), short-circuit current density (J sc ) and fill factor (FF) in solar cells. Therefore, P7-based OSCs exhibited a PCE of 7.2%, while the corresponding polymer without fluorine atoms only showed a PCE of 5.0%. The PCEs of BT-BDT based polymers could be further enhanced to over 9% when enhancing the molecular weight and using a fullerene derivative as interfacial layer. [53] BT-polymers with oligothiophene are the representative conjugated polymers. [54−56] Researchers introduced long branched alkyl side units into the 3-position of thiophene, so that two alkyl side units in the monomer containing BT and four thiophenes can ensure the enough solubility for the polymers. This design can also significantly reduce the steric hinderance between alkyl side units and conjugated backbones, and therefore the advantages of polythiophene can be maintained in BT-polymers. For example, the BT-polymer P8 (Fig. 2) was found to show near-infrared absorption property (onset close to 800 nm), high hole mobility (1.92 cm 2 ·V −1 ·s −1 in organic field-effect transistors) and high PCEs (7.64% at 230 nm) at thick films due to high crystallinity. [57] The PCE of this polymer was further enhanced to over 10% after controlling the aggregation behavior at high temperature. [55] These polymers also exhibited promising photovoltaic performance in non-fullerene OSCs (NFOSCs), which will be discussed in the following section.
In addition to the wide (<700 nm) and medium (<800 nm) band gaps, BT units can also be used to design narrow band gap conjugated polymers with absorption onset over 800 nm. Yang et al. used the silicon-contained fused bithiophene as comonomer into the BT-polymer P9 (Fig. 2), providing an absorption onset at 800 nm. OSCs based on P9:PCBM blends exhibited a PCE of up to 5.1%. [58] Later, they inserted a strong electron-donating oxygen atom into the cyclopentadithiophene unit to form the dithienopyran unit. The resulting polymer P10 based on dithienopyran and fluoro-substituted BT units exhibited absorption up to 900 nm. Single-junction OSCs based on P10 as donor and PC 71 BM as acceptor provided a PCE of 7.9% with the spectral response from 300 nm to 900 nm. [59] The narrow band gap property enabled P10 to be applied into tandem solar cells, providing a high PCE of 10.6%. Guo et al. developed a novel electron-donating unit 3alkoxy-3′-alkyl-2,2′-bithiophene with the head-to-head linked alkyl and alkyoxy units for the BT-polymer P11 (Fig. 2). The new polymer exhibited an absorption spectrum over 900 nm and showed a high PCE of 9.76% when using PC 71 BM as the acceptor in solar cells. [60] These results reveal that, by using the rational selection of copolymerized units, BT-polymers can provide distinct absorption properties, which can facilitate their application in non-fullerene solar cells.

BT-based Donor Polymers in Non-fullerene OSCs
Before discussing BT-polymers for application in non-fullerene OSCs, we intend to briefly summarize the characteristics of BTpolymers. Firstly, as mentioned above, the chemical structures of BT-polymers can be tailored to show distinct absorption spectra from visible light to near-infrared region, indicating that they can perfectly match with non-fullerene acceptors with complementary absorption. Secondly, BT-polymers always have  high crystallinity, especially for the polymers with oligothiophenes as comonomer (such as P8). Therefore, we could observe highly-ordered diffraction peaks in grazing incidence wide angle X-ray scattering (GIWAXS) images ( Fig. 3a) [55] and fibrous structures in blended films from transmission electron microscope (TEM) images (Fig. 3b). [61] This characteristic is quite similar to diketopyrrolopyrrole-based conjugated polymers. [62] Thirdly, BT-polymers exhibited high hole mobilities, in which the mobilities over 1 cm 2 ·V −1 ·s −1 could be obtained in organic fieldeffect transistor devices. [63] High crystallinity and hole mobilities enable BT-polymers to show good charge transport properties, exhibiting high performance in thick photoactive layers due to the reduced charge recombination. [55,64] In addition, from the aspect of synthesis and chemical structure, BT units can be easily prepared and there are also several modification cites. All these merits demonstrate the great potential application of BTpolymers in OSCs.
Although the PCEs of BT-polymer based fullerene OSCs have been enhanced to 11.7%, [36] the drawbacks of fullerene acceptors, such as weak absorption in the visible region and difficulty to tuning chemical structures and purification, limited the further improvement of PCEs. Fortunately, these drawbacks have been successfully overcome by developing new non-fullerene acceptors (NFAs) with excellent properties. With the development of those new NFAs, the PCE of BT-based OSCs was further improved up to 18.22%. [20]

BT-based donor polymers in fused-ring electron acceptor based NFOSCs
Since Zhan et al. reported the first fused-ring electron acceptor (FREA) ITIC with 2-(3-oxo-2,3-dihydroinden-1-ylidene)malononitrile as end groups, [15] ITIC and its derivatives have attracted much attention in recent years. [19,65−69] In general, wide band gap conjugated polymers with absorption onset below 700 nm are used as electron donor to match with these near-infrared FREAs in order to realize complementary absorption spectra, [70] in which the electron-deficient unit benzo[1,2-c:4,5-c′]dithiophene-4,8-dione (BDD) is the most widely reported building block to construct donor polymers toward high performance NFOSCs with PCEs over 17%. [24] BT-polymers have also been applied into FREA based OSCs, and it provides a record PCE of 18.22%. [20] When ITIC was published in the year of 2015, Yan et al. used an ITIC derivative ITIC-Th to work with the BT-polymer P8 for application in NFOSCs, in which the donor and acceptor exhibited complementary absorption spectra (Fig. 4). [71] A PCE of 6.6% with spectra response from 300 nm to 800 nm could be obtained. Zhu et al. developed a series of non-fullerene acceptors (ZITI-N-EH) to improve the miscibility with P12 that has the similar structure to P8. [72] As a consequence, the blend thin films could be fabricated at room temperature, and also provided the high PCE of 13.07%. It is also worthy to mention that the PCE could be maintained at 12.35% at a 200 nm thick film. Huang et al. also used a random copolymer design strategy to develop a BT-polymer P13 (Fig. 4). The blend thin films based on P13 could be fabricated from nonhalogenated solvents, and the solar cells exhibited PCEs over 10% with the thickness of photoactive layers from 100 nm to 300 nm. [73] These merits indicate the potential application of BT-polymers for large-area OSCs.
In addition to oligothiophene-based BT-polymers, BDTbased BT-polymers have also been applied into IC-based NFOSCs. Zhan et al. reported a BT-polymer P14 with alkyloxy side chains at BT backbones and fluoro-benzene side units attached to BDT (Fig. 4). [74] They used the acceptor IDIC to work with P14, providing a PCE of 11.03% in OSCs. Bo et al. developed the asymmetric BT monomers with alkyloxyl/alkylthio unit and fluorine atom, in which oxygen and sulfur atoms significantly influenced the photovoltaic performance (Fig. 4). [75] Oxygen-based BT polymer P15 showed the PCEs of 7.28%, while the PCE was reduced to 1.55% in sulfur-based polymer P16. This example revealed the importance of finely tunable chemical structures in designing BT-polymers for OSCs.
Sun et al. used the narrow band gap polymer P10 as donor to work with a near-infrared acceptor FOIC (Fig. 4). [76] The blend exhibited strong photo-response in the NIR region, and meanwhile the blend could be transparent in the visible light region. Therefore, the corresponding solar cells could be used as semitransparent solar cells by using the thin metal contact. They found that a PCE of 4.2% could be obtained with a high transparency in the visible light region.
Recently, Ding et al. reported a wide band gap BT-based donor polymer for NFOSC with an amazing result. [20] They constructed BT-based donor polymer P17 with a fused-ring   (Fig. 5). The high planarity of DTBT backbone gifted P17 a higher hole mobility (1.59× 10 −3 cm 2 ·V −1 ·s −1 ). When P17 was used as donor material to construct Y6 (Fig. 5)-based NFOSC, it provided an exciting PCE of 18.22%, which was the highest PCE in the stated-of-the-art NFOSCs.

BT-based donor polymers with other types of electron acceptors in NFOSCs
Besides fused-ring electron acceptors, there are also several kinds of small molecules as electron acceptor in BT-polymer based NFOSCs. [77,78] For example, Yan et al. used the rhodamine end-capped acceptor O-IDTBR combined with the BT-polymer P18 as electron donor to fabricate NFOSCs (Fig. 6). [79] P18 and O-IDTBR exhibited similar absorption spectra with onset at ~800 nm. NFOSCs based on their blends provided a high PCE of 10.4% with a high V oc of 1.08 V and a very low energy loss (E loss ) of 0.55 eV. Yan et al. also reported a series of work focusing on NFOSCs based on the BT-polymers as donor and perylene bisimide (PBI) derivatives as acceptor. A typical work was that they developed a BT-polymer P19 containing fluoro-substituted BT units and alkylthiophene units to simultaneously obtain small band gap and deep frontier energy levels (Fig. 6). [80] Additionally, a bis-PBI molecule SF-PDI 2 with spirofluorene as core was developed as a wide band gap acceptor to match with P19 to show complementary absorption spectra. A high PCE of 9.5% was obtained with the high V oc of 1.11 V and hence a low E loss of 0.61 eV. They also provided systematical studies about the origination of low E loss , mainly due to the low non-radiative recombination. The E loss could be further reduced to 0.53 eV together with a high PCE of 10.58% when using a new fused and star-shaped PBI-based acceptor FTTB-PDI4. [81] Bo et al. also used a series of 1,8-naphthalimide (NI) based electron acceptor for BT-based NFOSCs. [82−84] NI-based acceptors exhibit wide band gap absorption spectra and highlying energy levels, which can enable the high V oc in solar cells. For example, they used the coplanar BT-polymer P20 (Fig. 6) as donor and NI-based molecule NI-AA-NI with ethynyl units as acceptor to fabricate OSCs, in which a high V oc of 1.07 V and a PCE of 3.71% could be obtained. [84]

BT-based donor polymers in all-polymer solar cells
All-polymer solar cells (all-PSCs) as one type of OSCs have also attracted tremendous attention, not only for their tunable structural, optical and electrochemical properties, but also for their general merits over small molecules, such as good thermal stability and robust mechanical property. However, the performance of all-PSCs still lags behind fullerene-based or fused-ring electron acceptor-based OSCs (Fig. 7).
In 2015, in order to study the relationship between chemical structures and photovoltaic properties, Kim Sci. 2021, 39, 525-536 veloped a series of BT-polymers with different number-average molecular weights (P21 L , M n =12 kg/mol; P21 M , M n = 24 kg/mol; P21 H , M n =40 kg/mol) to investigate their photovoltaic performances (Fig. 7). [85] The P21 H with high M n im-proved the miscibility with the polymer acceptor P(NDI2OD-T2) and hence inhibited the formation of crystalline region. Therefore, the hole and electron mobilities of the optimized all-PSCs based on P21 H :P(NDI2OD-T2) are significantly higher    (Fig. 7). [86] They found that when enhancing the content of fluorine atoms, the polymers exhibited enhanced dipole moment difference between the ground and excited states ( μ ge ). The large μ ge of P24 was helpful for exciton dissociation and meanwhile suppressed charge recombination in solar cells when using P(NDI2HD-T2) as electron acceptor. As a consequence, P24 based all-PSCs provided a high PCE of 6.42% with simultaneously enhanced V oc , J sc and FF, while P22 only showed a low PCE of 2.65%. Liu et al. also introduced different numbers of fluorine atoms into BT-polymers to tune the crystallinity so as to match with the B→N polymer rr-PBN that has been developed in their group. [87] P25 with four fluorine atoms on each monomer exhibited a high PCE of 6.45% in all-PSCs (Fig. 7). Apparently, all-PSCs based on BT-polymers as donor are far to be explored.

BT-BASED POLYMER ACCEPTORS IN ALL-POLYMER SOLAR CELLS
The electron-deficient nature of BT indicates that BT-polymers can be used as electron acceptor for all-PSCs when providing rational design of chemical structures. Indeed, in recent years, there are many BT-polymers that have been developed for this purpose. [88−93] As early as 2007, the BT-polymer named as F8TBT (P26 here) that had been previously used as electron donor to work with PCBM in OSCs was also found to exhibit electron transport property, [94] so that P26 could also be used as electron acceptor in solar cells (Fig. 8). All-PSCs based on P3HT:P26 exhibited PCEs up to 1.2% when using a thermal annealed process. [95] This P3HT:P26 blend was then used as model system to study the crystallinity, morphology control and device physics in all-PSCs. [95−97] The deficient nature of P26 can also be clearly observed, that is, it has the high-lying frontier energy levels so that only P3HT can be used as electron donor in OSCs. In order to lower the frontier energy levels of BT-polymers, Guo et al. creatively introduced two cyano units into BT core. [98] Cyano unit as strong electron-deficient group can effectively lower the frontier energy levels, while the absorption spectra can still be maintained at the near-infrared region. The corresponding polymer P27 (Fig. 8) with indacenodithiophene as comonomer showed a narrow band gap of 1.43 eV and suitable energy levels of −3.75 and −5.59 eV as the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) levels. The optical and electrochemical properties of P27 are similar to those of the polymer acceptor based on NDI unit. More importantly, P27 exhibited high extinction coefficient in the near-infrared region, which perfectly solved the problem of low extinction coefficient of NDI-polymers. Therefore, P27 was used as electron acceptor in all-PSCs, providing a PCE of 8.32% with a broad photo-response from 300 nm to 850 nm. The solar cells also exhibited very high V oc , resulting in a low E loss of 0.53 eV. In a recent work, they introduced a strong electron-donating comonomer into the polymer P28 (Fig. 8) to further broaden the absorption spectra, and a very low band gap of 1. 28  polymer. [99] All-PSCs based on P28 as electron acceptor exhibited a PCE of 10.2% with a high J sc of 22.52 mA·cm −2 . These excellent polymer designs and their consequence undoubtedly prove that BT-polymers as electron acceptor have the great potential application in all-PSCs. Another interesting type of BT-polymer acceptors is based on the well-known near-infrared acceptor Y6 developed by Zou et al. [100] Y6 and its derivatives have been widely used as electron acceptor for high performance OSCs, in which BT is usually used as core. Inspired by the pioneer work that used ITIC as core to construct polymer acceptors by Li et al., several works focused on using Y6 to construct new polymers (for example, P29 with one thiophene as copolymerized unit). [37,91,101,102] These polymers exhibited near-infrared absorption up to 900 nm, well-aligned frontier energy levels and good electron transport properties. Therefore, the PCEs based on these polymers as electron acceptors could realize PCE over 14% (Fig. 8) when providing rational control of molecular weight, copolymerized units and so on. It can be envisioned that further development of Y6-based polymer acceptor can enable the PCEs of all-PSCs to be over 17%.

CONCLUSIONS
In this review, we provided an overview of BT-polymers for OSCs, from the aspect of electron donor and electron acceptor. With the judicious molecular design and device engineering, the PCEs based on BT-polymers have achieved over 18% and 14% when using as donor and acceptor. These outstanding achievements reveal the great potential of BT-polymers in OSCs, and the high PCE (18.22%) was one of the highest PCEs at thestated-of-the-art NFOSCs. However, if looking into the studies of NFOSCs, we will find that BT-polymers are far to be explored, so that it also leaves many tasks to be explored.
Firstly, it is necessary to explore BT-polymers as donor for NFOSCs, such as by using the known BT-polymers that have been used in fullerene-based OSCs and also developing new BT-polymers. The structural engineering includes BT core and comonomers, from the aspect of tuning the frontier energy levels and crystallinity, such as by introducing heteroatoms, different side units and the molecular weight. It is worth mentioning that direct arylation polymerization method as a green synthetic method is an important topic in the future, which should be considered in the synthesis of BTpolymers. [103] Secondly, BT-polymers usually show strong crystallinity as diketopyrrolopyrrole-polymers, which is different from BDDpolymers with strong aggregation. [104] Apparently, fused-ring electron acceptors tend to work with BDD-polymers possibly due to the well-organized phase separation, but would be incompatible with BT-polymers. To solve this problem, it would be the possible way to design BT-polymers with low crystallinity or develop suitable fused-ring acceptors. Additionally, it is very important to reveal the impact of crystallinity of BT-polymers on the morphology and hence the photovoltaic performance in solar cells.  Thirdly, it is important to focus on the photovoltaic devices based on BT-polymers, such as thick films, large-area and flexible devices, semitransparent cells and stability. The high crystallinity of BT-polymers enables to show high carrier mobilities, so that they still can exhibit high performance in thick films. This is in particular important for large-area devices. It seems that BT unit is very stable under thermal stress and heat, but there are still very few studies focusing on the stability of BT-polymers.
In conclusion, BT-polymers exhibit many characteristics in organic electronics, which can act as the important conjugated materials for OSCs. Further studies, from material design, morphology control to device engineering, are urgently needed to endow BT-polymers based OSCs with high stability as well as high performance even under thick films, which will be the promising system for industry application.