Synthesis of 2,2′-[2,2′-(arenediyl)bis(anthra[2,3-b]thiophene-5,10-diylidene)]tetrapropanedinitriles and their performance as non-fullerene acceptors in organic photovoltaics

Synthetic Metals 255 (2019) 116097 Contents lists available at ScienceDirect Synthetic Metals journal homepage: www.elsevier.com/locate/synmet Synthesis of 2,2′-[2,2′-(arenediyl)bis(anthra[2,3-b]thiophene-5,10diylidene)]tetrapropanedinitriles and their performance as non-fullerene acceptors in organic photovoltaics T Denis S. Baranova,b, , Olga L. Krivenkoa, Maxim S. Kazantsevb,c, Danil A. Nevostrueva, Elena S. Kobelevaa,b, Vladimir A. Zinovievd, Alexey A. Dmitrieva,b, Nina P. Gritsana,b, Leonid V. Kulika,b ⁎ a V.V. Voevodsky Institute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences, 630090 Novosibirsk, Russia Novosibirsk State University, 630090 Novosibirsk, Russia N.N. Vorozhtsov Novosibirsk Institute of Organic Chemistry, Siberian Branch of the Russian Academy of Sciences, 630090 Novosibirsk, Russia d A.V. Rzhanov Institute of Semiconductor Physics, Siberian Branch of the Russian Academy of Sciences, 630090 Novosibirsk, Russia b c A R T I C LE I N FO A B S T R A C T Keywords: Non-fullerene acceptors π-Extended TCNQ-based molecules Functionalized anthra[2,3-b]thiophenes Synthesis Photovoltaic properties Three novel molecules based on thiophene-fused 11,11,12,12-tetracyano-9,10-anthraquinodimethane (TCAQ) units linked through functionalized π-conjugated arene bridges are synthesized by three-step sequence of Sonagashira coupling, double thioannulation and Knoevenagel condensation. The energy of their frontier orbitals shows that these molecules are suitable for replacing fullerenes as electron acceptors in composites with conjugated donor polymers used in organic photovoltaics (OPV). The composite of polymer donor PCDTBT with fluorene-linked thiophene-fused TCAQs demonstrates good solubility in organic solvents and bulk heterojunction morphology with characteristic donor and acceptor domain size of several tens of nanometers, favourable for photoelectric conversion. This composite also exhibits prominent light-induced EPR signal indicating efficient photoinduced free charge generation. OPV devices based on this composite were fabricated by vacuum-free method. They show photoelectric conversion with maximum PCE of 0.71%. 1. Introduction 7,7,8,8-Tetracyanoquinodimethanes (TCNQs) have been extensively studied as strong electron acceptors in charge-transfer systems [1–3]. Fused aromatic TCNQs, mainly 11,11,12,12-tetracyano-9,10-anthraquinodimethane (TCAQs) were widely used as strong electron-acceptor building blocks for many organic functional materials [4,5] such as functionalized conducting layers [6,7], copolymers [8,9], solvatochromic compounds [10], chromophores [11,12], luminophores [13], molecular receptors [14] and liquid crystals [15]. However, the TCAQ or other fused aromatic TCNQ derivatives are very rarely used as acceptors and exhibit poor performance in organic photovoltaic OPV devices compared to conventional fullerene acceptor PCBM. For example, OPV with TCAQ-pyrrolidino[3,4:1,2][60]fullerene derivative as electron acceptor and poly[2-methoxy-5-(3,7-dimethyloctyloxy)-1,4phenylenevinylene] as the electron donor showed power conversion efficiency (PCE) of the order of 0.03% [16]. Meanwhile TCNQ structural fragment is promising for design of stable aceno-like organic semiconductors with high electronic conductivity [17]. Introduction of strong electron-withdrawing TCNQ groups into polycyclic conjugated backbone leads to lowering of the LUMO energy levels [18]. This strategy is an alternative, for example, to the synthesis of polyhalogenides [19,20], aza-analogs [21–23], and diimide acenes [24,25], which are also used for modification of acenolike n-type semiconductors. In addition, polycyclic TCNQs have a butterfly-shaped structure in which the central TCNQ ring adopts a boattype conformation whereas the lateral rings remain planar [26,27]. As a result, TCNQ fragments increase solubility of the polycyclic molecules [5] and also improve their stability [28], for example, they prevent photooxidation and Diels-Alder transformations, what is especially important problem for expanded acenes [29]. Note that effect of heteroatoms (S, N, O) on the properties of the polycondenced framework is similar. For example, thiophene-fused polycyclic systems such as anthra [2,3-b]thiophene (AT) and anthra[2,3-b:6,7-b′]dithiophene (ADT) are ⁎ Corresponding author at: V.V. Voevodsky Institute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences, 630090 Novosibirsk, Russia. E-mail address: baranov@ns.kinetics.nsc.ru (D.S. Baranov). https://doi.org/10.1016/j.synthmet.2019.06.013 Received 3 April 2019; Received in revised form 6 June 2019; Accepted 12 June 2019 0379-6779/ © 2019 Elsevier B.V. All rights reserved. Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. =7.0 Hz, 6H, 2Me), 1.01–1.26 (m, 20H, 10CH2), 2.05 (m, 4H, 2CH2), 7.61 (s, 2H, HAr), 7.65 (dd, J = 1.3, 7.8 Hz, 2H, HAr), 7.76 (d, J =7.8 Hz, 2H, HAr), 7.84 (m, 4H, HAr), 8.33 (m, 2H, HAr), 8.37 (s, 2H, HAr), 8.51 (s, 2H, HAr). 13C NMR (CDCl3, 100 MHz) δ: 14.22, 22.75, 23.88, 29.37, 29.38, 30.09, 31.92, 40.39 (2C8H17), 55.64 (C), 86.21, 100.80 (2 C≡C), 120.59 (2CH), 121.16 (2C), 126.54 (2CH), 127.59 (4CH), 128.09 (2CH), 129.42 (2C), 131.60 (2CH), 131.66 (2C), 132.31 (2CH), 132.94 (2C), 133.42 (2C), 133.54 (2C), 134.64 (4CH), 141.82 (2C), 141.91 (2C), 151.63 (2C), 181.76, 182.03 (4C = O). IR (v/cm−1): 2924, 2851 (C8H17), 2201 (C≡C), 1676 (C = O). Found (%): C, 79.54; H, 5.58; Cl, 7.68. Calc. for C61H52Cl2O4 (%): C, 79.64; H, 5.70; Cl, 7.71. more stable than the corresponding acenes [30]. Therefore, the idea of combining a heteroacene framework with TCNQ in one molecule for synthesis of new functional materials is attractive. Notably those thienoacene-like building blocks in combination with strong electron withdrawing terminal groups, such as dicyanomethylene, are successfully used for synthesis of non-fullerene acceptors, which show record PCE in OPV [31–33]. In this paper, we report three novel thiophene-fused TCAQ-based molecules and their performance as non-fullerene acceptors in organic photovoltaics. The peculiarity of these π-conjugated molecules is the presence of TCNQ fragment in linear AT backbones, which are connected through different arene bridges containing solibilizing substituents. The differences in the structure and properties of the synthesized compounds are given by the central linker. Double thioannulation of arenediyl-bis(ethynylanthraquinone) substrates with sodium sulfide was employed to build polycyclic framework. TCNQ moieties were formed by Knoevenagel condensation of diquinones with malodinitrile. Interestingly, the linker structure together with the steric effect of the substituents, did not significantly affect synthetic transformations. Optical, morphological and photovoltaic properties of the composites of polymer donor PCDTBT with novel thiophene-fused TCAQ-based molecules were tested. 2.2.1.3. 3,3’-(2-Octyloxy-5-tert-pentylbenzene-1,3-diyl)bis(ethyne-2,1diyl)bis(2-chloroanthracene-9,10-dione) (1c). Yield 1.35 g (83%), mp 184–185 °C (toluene). 1H NMR (CDCl3, 400 MHz) δ: 0.76 (m, 6H, 2Me), 1.17 (m, 6H, 3CH2), 1.31 (m, 2H, CH2), 1.35 (s, 6H, 2Me), 1.53 (m, 2H, CH2), 1.70 (m, 2H, CH2), 1.93 (m, 2H, CH2), 4.43 (t, J =6.7 Hz, 2H, CH2O), 7.58 (s, 2H, HAr), 7.84 (m, 4H, HAr), 8.33 (m, 4H, HAr), 8.37 (s, 2H, HAr), 8.51 (s, 2H, HAr). 13C NMR (CDCl3, 100 MHz) δ: 9.33, 14.20, 22.79, 26.33, 28.48, 29.45, 29.71, 30.66, 31.96, 36.85, 37.86 (OC8H17, Et, 2Me), 75.40 (C), 89.29, 96.28 (C≡C), 116.04 (2C), 127.59 (4CH), 128.10 (2CH), 129.34 (2C), 131.65 (2C), 132.47 (2CH), 133.08 (2C), 133.38 (2C), 133.40 (2C), 133.49 (2C), 134.62 (2CH), 134.67 (2CH), 141.96 (2CH), 145.06 (C), 159.73 (C), 181.76, 181.91 (4C = O). IR (v/cm−1): 2955, 2922, 2855 (Alk), 2202 (C≡C), 1672 (C = O). Found (%): C, 76.24; H, 5.21; Cl, 8.76. Calc. for C51H42Cl2O5 (%): C, 76.02; H, 5.25; Cl, 8.80. 2. Experimental 2.1. Measurements and materials Characterization of the synthesized molecules was carried out by instruments and procedures used our previously [37]. Starting materials, such as 2-chloro-3-iodoanthracene-9,10-dione [34], 1,4-diethynyl-2,5-bis(octyloxy)benzene [35], 2,7-diethynyl-9,9-dioctyl-9Hfluorene [36], 1,3-diethynyl-2-octyloxy-5-tert-pentylbenzene [37], were prepared following the appropriate literature procedures. Glass/ITO substrates with Ω = 20 ohm/cm2 (S101), poly(3,4-ethylenedioxythiophene) polystyrene sulfonate PEDOT:PSS (Al 4083), PCDTBT (M1311) and electrical connection legs (E242) were purchased from Ossila Ltd. [6,6]-Phenyl-C61-butyric acid methyl ester PCBM was purchased from Solenne b. v. Solvents were received from Sigma-Aldrich and were distilled before use. Bi, In, Sn for Field’s Metal Alloy (FM) [38] were purchased from Rotometals. 2.2.2. General procedure for cyclization of bis(ethynylanthraquinones) 1 with Na2S·9H2O A mixture of bis(ethynylanthraquinone) 1 (0.9 mmol) and Na2S·9H2O (1.3 g, 5.4 mmol) in pyridine (45 mL) was heated under reflux for 3 − 10 h. After the completion of reaction the mixture was diluted with H2O (15 mL), the crude product was filtrated, washed with ethanol. The residue was purified by silica gel chromatography (eluted with hot toluene). 2.2.2.1. 2,2’-(2,5-Dioctyloxybenzene-1,4-diyl)bis(anthra[2,3-b]thiophene5,10-dione) (2a). Yield 390 mg (50%), mp 285–287 °C (toluene). 1H NMR (CDCl3, 400 MHz) δ: 0.93 (br.s, 6H, 2Me), 1.32–1.65 (br.m, 20H, 10CH2), 2.05 (br.s, 4H, 2CH2), 4.18 (br.s, 4H, 2CH2O), 7.18 (br.s, 2H, HAr), 7.65 (br.m, 4H, HAr), 7.81 (br.s, 2H, Hthiophene), 8.23 (br.m, 4H, HAr), 8.52 (br.s, 2H, HAr), 8.65 (br.s, 2H, HAr). IR (v/cm−1): 2922, 2853 (OC8H17), 1667 (C = O). Found (%): C, 75.73; H, 5.58; S, 7.56. Calc. for C54H50O6S2 (%): C, 75.49; H, 5.87; S, 7.46. 2.2. Synthesis 2.2.1. General procedure for preparation of bis(ethynylanthraquinone) substrates (1) 2-Chloro-3-iodoanthracene-9,10-dione (1.5 g, 4.0 mmol), diethynylarene (2.0 mmol), CuI (15 mg, 0.078 mmol), PdCl2(PPh3)2 (30 mg, 0.042 mmol) and pyridine (30 mL) were to a flask, and mixture was stirred and heated to 60 °C under argon. Then, a hot aqueous solution of 1.2 M Na2CO3 was added, and the reaction mixture was stirred at 95 °C for 1 h. After the completion of reaction, mixture was cooled, the crude product was filtrated, washed with water. The residue was purified by silica gel chromatography (eluted with toluene). 2.2.2.2. 2,2’-(9,9-Dioctyl-9H-fluorene-2,7-diyl)bis(anthra[2,3-b] thiophene-5,10-dione) (2b). Yield 495 mg (60%), mp 192–193 °C (toluene). 1H NMR (CDCl3, 400 MHz) δ: 0.73 (br.s, 4H, 2CH2), 0.77 (t, J =7.0 Hz, 6H, 2Me), 1.05–1.21 (m, 20H, 10CH2), 2.13 (m, 4H, 2CH2), 7.75 (s, 2H, Hthiophene), 7.83 (m, 10H, HAr), 8.38 (m, 4H, HAr), 8.72 (s, 2H, HAr), 8.80 (s, 2H, HAr). 13C NMR (CDCl3, 100 MHz) δ: 14.19, 23.95, 29.31, 29.32, 30.04, 31.90, 40.43 (2C8H17), 55.82 (C), 120.34 (2CH), 121.00 (2C), 121.15 (2C), 122.43 (2CH), 123.06 (2CH), 126.33 (2C), 127.50 (2CH), 127.52 (4CH), 129.11 (2C), 130.44 (2C), 132.65 (2C), 134.18 (2CH), 134.22 (2CH), 134.29 (2CH), 141.87 (2C), 144.51 (2C), 144.76 (2CH), 151.29 (2C), 152.44 (2C), 183.01, 183.47 (4C = O). IR (v/cm−1): 2951, 2920, 2850 (C8H17), 1666 (C = O). Found (%): C, 80.13; H, 5.78; S, 6.88. Calc. for C61H54O4S2 (%): C, 80.05; H, 5.95; S, 7.01. 2.2.1.1. 3,3’-(2,5-Dioctyloxybenzene-1,4-diyl)bis(ethyne-2,1-diyl)bis(2chloroanthracene-9,10-dione) (1a). Yield 1.5 g (86%), mp 249–250 °C (toluene). 1H NMR (CDCl3, 400 MHz) δ: 0.86 (t, J =7.0 Hz, 6H, 2Me), 1.25–1.44 (m, 20H, 10CH2), 1.89 (m, 4H, 2CH2), 4.09 (t, J =6.6 Hz, 4H, 2CH2O), 7.09 (s, 2H, HAr), 7.83 (m, 4H, HAr), 8.32 (m, 4H, HAr), 8.35 (c, 2H, HAr), 8.48 (s, 2H, HAr). IR (v/cm−1): 2922, 2851 (OC8H17), 2204 (C≡C), 1672 (C = O). HMRS, m/z: 862.2817 (calc. for C54H48Cl2O6: 862.2823[M]+). 2.2.2.3. 2,2’-(2-Octyloxy-5-tert-pentylbenzene-1,3-diyl)bis(anthra[2,3-b] thiophene-5,10-dione) (2c). Yield 510 mg (71%), mp 245–246 °C (toluene - petroleum ether). 1H NMR (CDCl3, 400 MHz) δ: 0.70 (t, J =7.0 Hz, 3H, Me), 0.81 (t, J =7.3 Hz, 3H, Me), 1.04 (m, 8H, 4CH2), 1.16 (m, 2H, CH2), 1.42 (s, 6H, 2Me), 1.71 (m, 2H, CH2), 1.77 (m, 2H, 2.2.1.2. 3,3’-(9,9-Dioctyl-9H-fluorene-2,7-diyl)bis(ethyne-2,1-diyl)bis(2chloroanthracene-9,10-dione) (1b). Yield 1.5 g (81%), mp 191–192 °C (toluene). 1H NMR (CDCl3, 400 MHz) δ: 0.64 (br.s, 4H, 2CH2), 0.80 (t, J 2 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. CH2), 3.62 (t, J =6.6 Hz, 2H, CH2O), 7.69 (s, 2H, HAr), 7.83 (m, 4H, HAr), 7.97 (s, 2H, Hthiophene), 8.39 (m, 4H, HAr), 8.78 (s, 2H, HAr), 8.86 (s, 2H HAr). 13C NMR (CDCl3, 100 MHz) δ: 9.42, 14.17, 22.70, 25.93, 28.60, 29.31, 30.27, 31.51, 31.85, 36.96, 38.14 (C8H17, Et, 2Me), 75.03 (C), 122.36 (2CH), 123.25 (2CH), 123.94 (2CH), 127.50 (2CH), 127.52 (2CH), 127.79 (2C), 129.09 (2C), 129.14 (2C), 129.17 (2C), 130.19 (2C), 134.18 (2CH), 134.23 (2CH), 134.28 (2CH), 143.72 (2C), 145.50 (2C), 146.49 (2C), 146.68 (C), 151.98 (C), 183.44, 183.07 (4C = O). IR (v/cm−1): 2959, 2922, 2856 (Alk), 1663 (C = O). HMRS, m/z: 800.2617 (calc. for C51H44O6S2: 800.2625[M]+). 2.4. Device fabrication and characterization The architecture of devices was ITO/PEDOT:PSS/Active layer/FM. ITO substrates were cleaned mechanically in detergent solution, after that subsequently ultrasonicated in distilled water, acetone, ethanol and deionized water. In the end, substrates were dried by clean air, baked at 80 °C for 10 min in air atmosphere and UV-ozone treated for 30 min. PEDOT:PSS was filtered through PTFE hydrophilic 0.45 μm pore-size filter and spin-coated on the top of pre-cleaned ITO substrates at 5000 rpm for 20 s, followed by annealing at 120 °C for 10 min under Ar atmosphere. Active layer was spin-coated at 2000 rpm for 30 s from blend of donor (PCDTBT) and acceptor (PCBM or 3b) chlorobenzene solution (w/w ratio was 1:1, total concentration 20 mg/ ml) on the top of PEDOT:PSS layer, followed by annealing at 120 °C for 10 min under Ar atmosphere. Finally, FM was used as cathode. For this, substrates and aluminum melting pot with FM inside were placed on hot plate and heated in air to 100 °C. The melted alloy was deposited by dripping onto active layer and the amount of the metallic alloy determined final its thickness, typically in the range of 0.5–1.5 mm. The active area was in range of 5.2 – 8.4 mm2. As the light source light-emitting diode CREE XM-L U3 with color temperature of 5000 K was used. The light intensity was calibrated using standard Si solar cell (100 mW/cm2). In order to measure current density-voltage characteristics potentiostat P-20X (Electrochemical instruments, Russia) was used. 2.2.3. General Procedure for TiCl4/pyridine-catalyzed condensation of diquinone 2 with malononitrile A mixture diquinone 2 (0.25 mmol) and malononitrile (150 mg, 2.25 mmol) in CH2Cl2 (30 mL) was stirred under argon at r.t. for 10 min. Then, TiCl4 (0.27 mL, 2.5 mmol) and dry pyridine (0.52 mL, 6.5 mmol) were added. After stirring the reaction mixture for 1 h, the reaction mixture was chromatographed on a silica gel with CH2Cl2. The solvent was evaporated under vacuum, crude product 3 was crystallized. 2.2.3.1. 2,2′-[2,2′-(2,5-Dioctyloxybenzene-1,4-diyl)bis(anthra[2,3-b] thiophene-5,10-diylidene)]tetrapropanedinitrile (3a). Yield 194 mg (74%), mp 355–356 °C (toluene - ethyl acetate). :1H NMR (CDCl3, 400 MHz) δ: 0.87 (t, J =6.6 Hz, 6H, 2Me), 1.30 (m, 8H, 4CH2), 1.36 (m, 4H, 2CH2), 1.44 (m, 4H, 2CH2), 1.58 (m, 4H, 2CH2), 2.02 (m, 4H, 2CH2), 4.25 (t, J =6.5 Hz, 4H, 2OCH2), 7.43 (s, 2H, HAr), 7.74 (m, 4H, HAr), 7.98 (s, 2H, Hthiophene), 8.27 (m, 4H, HAr), 8.63 (s, 2H, HAr), 8.71 (s, 2H, HAr). IR (v/cm−1): 2924, 2854 (OC8H17), 2222 (C≡N). Found (%): C, 75.44; H, 5.04; N, 10.69; S, 6.15. Calc. for C66H50N8O2S2 (%): C, 75.40; H, 4.79; N, 10.66; S, 6.10. 2.5. Details of quantum chemical calculations Geometry optimization and the HOMO-LUMO energy gap calculation for 2 and 3 were performed at the B3LYP/def2-TZVP level [39–41] using ORCA suit of programs [42]. Positions and oscillator strengths of the electronic transitions in the electronic absorption spectra of 2 and 3 were calculated using time-dependent (TD) DFT [43] at the TD-B3LYP/ def2-TZVP level using the same software. 2.2.3.2. 2,2′-[2,2′-(9,9-Dioctyl-9H-fluorene-2,7-diyl)bis(anthra[2,3-b] thiophene-5,10-diylidene)]tetrapropanedinitrile (3b). Yield 224 mg (81%), mp 348–349 °C (ethyl acetate). 1H NMR (CDCl3, 400 MHz) δ: 0.67-0.83 (m, 10H, 2CH2+2Me), 1.03–1.22 (m, 20H, 10CH2), 2.10 (m, 4H, 2CH2), 7.75 (m, 12H, HAr+Hthiophene), 8.28 (m, 4H, HAr), 8.65 (d, J =35.4 Hz, 2H, HAr), 8.74 (d, J =29.1 Hz, 2H, HAr). IR (v/cm−1): 2924, 2855 (C8H17), 2222 (C≡N). Found (%): C, 79.44; H, 4.95; N, 10.16; S, 6.03. Calc. for C73H54N8S2 (%): C, 79.18; H, 4.92; N, 10.12; S, 5.79. 3. Results and discussion 3.1. Synthesis The approach to the assembly of TCAQ-based molecules is shown in Scheme 1. Previously, we used a similar route for the synthesis of 2,2′[2,2′-arenediylbis(11-oxoanthra[1,2-b]thiophene-6-ylidene)]dipropanedinitriles [37]. The double thioannulation of ethynylanthraquinone substrates with sodium sulfide is a key stage in the assembly of a πconjugated system consisting of two AT fragments connected by functionalized arene linkers. Note that organolithium reagents and expensive thiophene intermediates were not used. For comparison, literature methods for the synthesis of AT-5,10-diones through condensation reactions requires a multi-step preparation of thiophene intermediates [44–48]. In addition, they are not suitable for the synthesis of AT-5,10-diones containing aryl or hetaryl substituents in the thiophene ring. Alkyne substrates 1 containing chlorine atoms were prepared by Sonogashira reaction of 2-chloro-3-iodoanthraquinone with diethynylarenes in aqueous pyridine with Na2CO3 as the base at reflux temperatures. As expected, the transition metal-free double thioannulation of alkyne substrates 1 with sodium sulfide in pyridine at 115 °C for 3–10 h gave corresponding thienoquinones 2 in 50–71% yield. Aliphatic substituents have been used to make polycyclic molecules soluble in organic solvents. Synthesized diquinone 2 are promising substrates for the introduction of dicyanomethylene groups through condensation with malononitrile. However, this reaction has some limitations in the case of thieno-fused anthraquinones. De la Cruz and co-workers reported that the Knoevenagel condensation of AT-5,10dione with malononitrile catalyzed by TiCl4/pyridine (Lehnert reagent) in dichloromethane at room temperature for 24 h led to a mixture of 2.2.3.3. 2,2′-[2,2′-(2-Octyloxy-5-tert-pentylbenzene-1,3-diyl)bis(anthra [2,3-b]thiophene-5,10-diylidene)]tetrapropanedinitrile (3c). Yield 190 mg (76%), mp 350–352 °C (ethyl acetate). 1H NMR (CDCl3, 400 MHz) δ: 0.70 (t, J =7.0 Hz, 3H, Me), 0.81 (t, J =7.3 Hz, 3H, Me), 1.02 (m, 8H, 4CH2), 1.14 (m, 2H, CH2), 1.41 (s, 6H, 2Me), 1.68 (m, 2H, CH2), 1.75 (m, 2H, CH2), 3.59 (t, J =6.6 Hz, 2H, OCH2), 7.69 (s, 2H, HAr), 7.75 (m, 4H, HAr), 7.92 (s, 2H, Hthiophene), 8.27 (m, 4H, HAr), 8.65 (s, 2H, HAr), 8.73 (s, 2H, HAr). IR (v/cm−1): 2924, 2854 (Alk), 2222 (C≡N). Found (%): C, 76.15; H, 4.60; N, 11.11; S, 6.73. Calc. for C63H44N8OS2 (%): C, 76.18; H, 4.47; N, 11.28; S, 6.46. 2.3. Optical and atomic-force microscopy measurements UV/Vis absorption spectra were recorded using Agilent 8453 spectrophotometer (Agilent Technologies). In these experiments chlorobenzene was used as solvent and typical concentrations were 10−5 M. Thin films were prepared from filtered through PTFE hydrophobic filter (0.45 μm) chlorobenzene solutions with concentration of 10 mg/ml (w/ w ratio was 1:1 for PCDTBT/3 blends) by spin-coating at 2000 rpm for 30 s. After that PCDTBT/3 composites were annealed at 120 °C for 10 min. Atomic-force microscope (AFM; NT-MDT, Russia) in semicontact mode was used to estimate topography and thickness of obtained films. 3 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. Scheme 1. Synthetic route to TCAQ-based molecules 3. tetracyano- (40%) and isomeric dicyano derivatives (19%) [49]. The transformation of ADT-5,11-dione under the same reaction conditions for 17 h gave TCNQ derivative in 8% yield [50]. Conversely, the exhaustive condensation of compounds2 with malononitrile led to octacyano products 3 in 74–81% yield. The compounds were characterized by elemental analysis, IR, 1H NMR,13C NMR and mass spectrometry. Table 1 Electrochemical properties of compounds 2 and 3. 3.2. Electrochemical properties Electrochemical properties of compounds 2 and 3 were studied in dichloromethane by cyclic voltammetry (CV). The saturated concentration of the studied compounds in dichloromethane was used for the CV measurements. The cyclic voltammograms for all studied compounds are presented in Fig. 1 and corresponding electrochemical data are summarized in Table 1. All compounds demonstrated reversible reduction processes. Irreversible oxidation processes for 2a,c and quasireversible ones for 2b, 3a,b were observed. TCAQ-based molecules 3 Compound HOMOa, eV LUMOa, eV Ega, eV Eox 1/2, vs Fc/Fc+ Ered 1/2, vs Fc/Fc+ 2a 2b 2c 3a 3b 3c −5.55 −5.76 −6.01 −5.62 −5.77 – −3.42 −3.43 −3.42 −3.94 −3.97 −3.95 2.13 2.33 2.59 1.67 1.80 – – 1.05 – 0.92 1.08 – −1.45 −1.43 −1.45 −0.84 −0.83 −0.84 a Energies estimated from onset oxidation and reduction potentials in CV. demonstrated significantly less negative (by ˜0.6 eV) reduction potentials as compared to AT-5,10-dione derivatives 2, indicating electron withdrawing effect of CN-groups. For both classes of compounds 2 and 3 the LUMO energy levels estimated from CV data demonstrated high conformity, indicating a weak effect of the central electron-donating fragment, whereas HOMO energy levels are highly dependent on the configuration and electronic structure of the central linker (Table 1). The E1/2 reduction potentials for 3 derivatives are about -0.84 eV (vs Fc/Fc+) which is comparable to that for TCAQ molecule (about −0.67 eV) [51]. Strong electron accepting character of 3 derivatives and low-lying LUMO levels are considered to be important prerequisites for applications of studied compounds as an electron acceptors. 3.3. Optical properties and quantum chemical calculations The electronic absorption spectra of chlorobenzene solutions and thin films of synthesized molecules are displayed in Fig. 2. Important spectroscopic information is also summarized in Table 2. To better understand electronic structure of compounds 2 and 3, we performed the DFT calculations of the HOMO-LUMO energy gap and time-dependent DFT calculations of their spectroscopic properties. Fig. 2 demonstrate that calculated positions of the electronic transitions coincide fairly well with the experimental UV–vis absorption spectra of compounds 2 and 3. Note, that long-wavelength transitions in the UV–vis spectra of 2a,b and 3a are almost exclusively due to the single Fig. 1. Cyclic voltammograms of studied compounds in CH2Cl2 solution. 4 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. Fig. 4. Frontier molecular orbitals, their energies and HOMO – LUMO energy gap (Eg) for compounds 2a-c and 3a-c calculated at the B3LYP/def2-TZVP level of theory for geometries optimized at the same level. Fig. 2. UV/Vis absorption spectra of: (a) solutions of 2 in chlorobenzene; (b) solutions of 3 in chlorobenzene; (c) thin films of 3a (from chloroform) and of 3b,c (from chlorobenzene); computed positions and oscillator strengths (f) of the electronic transitions for 2 (a) and 3 (b) are depicted as solid bars of corresponding color. HOMO → LUMO excitation (96–98%). For 3b,c, contribution of HOMO → LUMO excitation is 82 and 89%, respectively, and only for 2c three electronic excitations (HOMO → LUMO, HOMO-1 → LUMO and HOMO → LUMO + 1) contribute almost equally to the long-wavelength transition. As demonstrated in Fig. 3, the HOMOs of compounds under study are mainly localized at the central part of the molecules. In turn, the LUMOs are localized at one (2c, 3c) or both (2a,b and 3a,b) AT fragments. Therefore, the long-wavelength bands in the UV–vis spectra of all compounds (except 2c) are the charge-transfer (CT) bands and, as expected [52,53], TD-DFT calculations underestimate significantly the energy of CT-transitions. Better agreement of the calculated spectra with experiment for 2a,b is due to a more delocalized LUMO for these species comparing to that of 3 (Fig. 3). In case of compounds 2 (Fig. 2a), a bathochromic shift of absorption spectra is observed in the row c→b→a. When carbonyl groups were substituted by dicyanomethylene moieties (Fig. 2b), this trend continues and the shift becomes larger, this is associated with a decrease in energy of the LUMO. In thin films (Fig. 2c) of compounds 3b,c, absorption band broadening is observed, as compared with solutions, due to increased intermolecular interactions and formation of aggregates. It was not possible to register the absorption spectrum of neat 3a film Table 2 Optical properties of solutions and films of synthesized compounds. Compound ε, M−1 cm−1 (λmax)a h, nmb α, cm−1 (λmax)c Eg, eVd Eg, eVe 2a 2b 2c 3a 3b 3c 43400 47600 76200 20200 57200 13470 n.a. n.a. n.a. 2–3f 25–28 16–18 n.a. n.a. n.a. 106000 ± 18000f 102000 (391) 79000 (360) 2.48 2.58 2.74 2.07 2.17 2.28 2.86 2.95 3.34 2.36 2.45 2.69 a b c d e f (319) (388) (312) (348) (398) (359) Wavelengths correspond to the maximum of absorption in solution. Film thicknesses. Wavelengths correspond to the maximum absorption in films. The HOMO-LUMO gap determined from the absorption edge of solutions. The HOMO-LUMO gap calculated at the B3LYP/def2-TZVP level. Film was prepared from chloroform. Fig. 3. Frontier molecular orbitals (│isovalue│ = 0.03 a.u.) calculated for 2a-c and 3a-c at the B3LYP/def2-TZVP level of theory. 5 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. Fig. 5. AFM (2 × 2 μm2) images of (a) 3a, (b) 3b, (c) 3c and (d) PCDTBT/3b annealed at 120 °C. intensity of peak at 400 nm increased because of absorption of 3b,c in this region. Fig. 4 displays the calculated energies of the frontier molecular orbitals for all compounds under study. In full agreement with CV data, calculations demonstrate that the nature of the electron-donating linker has no effect on the energy of LUMO orbitals. Calculations also predict tremendous effect of the oxygen substitution by the dicyanomethylene group to the LUMO energy, although the calculated LUMO energy difference (˜1.1 eV) is much higher than that estimated from CV (˜0.5 eV). In agreement with CV estimations, the HOMO energy levels are highly dependent on both the substitution and nature of the linker: the substitution leads to about 0.4 eV energy reduction, the linker c also leads to ˜0.4 eV decrease in the HOMO energies of both 2c and 3c comparing to their cogeners (Fig. 3). Thus, the calculated HOMO – LUMO energy gap (Eg) depends strongly on the substitution going down by 0.5 eV for 3a,b and 0.65 eV for 3c. Note that calculated energy gaps are 0.3 – 0.4 eV higher than those estimated from experimental UV–vis spectra and by 0.62 – 0.73 eV than those estimated from CV data. Fig. 6. Dark and Light (under irradiation) EPR signals of PCDTBT/3b composite, T = 80 K. cast-from chlorobenzene due to poor solubility. In order to solve this problem, we made saturated solution of 3a in chloroform, followed by heating to boiling temperature for 10 min and after that ultrasonicated by ultrasonication probe for few min. Using this way it was possible to spin-coat film on substrate (3000 rpm for 30 s). The thickness of obtained film was 2–3 nm. (Fig. 2c) In PCDTBT/3a composite there are two absorption bands corresponding to PCDTBT and the presence of 3a had no effect on the ratio of their intensities (0.9:1) [54], which also indicates low solubility of 3a. For PCDTBT/3b,c composites the 3.4. Thin film morphology The morphology of thin films of 3 was studied by atomic force microscopy in semicontact mode. The layers were deposited on glass substrates by spin-coating from chlorobenzene solution. Root mean square (RMS) of 3a, 3b, 3c and PCDTBT/3b is 0.48, 0.20, 0.16 and 0.37 nm, respectively. It can be seen that central π-linker affects the morphology of the neat films (Fig. 5). The formation of crystalline structures is observed on the 3a and 3c surfaces. However, for 6 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. Table 3 Summary of blend performance. Device Jsc, mA/cm2 Voc, V FF, % PCE, % PCDTBT/PCBM PCDTBT/3b 12.70 ± 0.30 2.95 ± 0.26 0.857 ± 0.027 0.764 ± 0.017 37.8 ± 1.4 28.3 ± 0.4 4.12 ± 0.09 0.64 ± 0.06 compound 3b with fluorene π-linker such structures are not observed, what may have a positive effect on the morphological stability of the film. In PCDTBT/3b composite the size of acceptor and donor phases are not too large (˜60-70 nm), what would facilitate the exciton dissociation at the donor-acceptor interface. We did not investigate composites based on PCDTBT and 3a,c due to formation of large grains on the top of the film, which can be observed without using any optical equipment. 3.5. Light-induced EPR In order to obtain insight into light-induced charge generation in PCDTBT/3b composite, light-induced EPR experiments were performed at 80 K. The intensity of light EPR signal is more than ten-fold higher than that of dark EPR signal (see Fig. 6), which is typical for polymer/ fullerene OPV blends with efficient charge separation [55]. This points on low concentration of defect sites producing dark EPR signal, and on efficient generation of free charges in PCDTBT/3b composite. Single strong light-induced EPR line is observed for PCDTBT/3b composite, probably, because of the superposition of EPR lines of anion 3b− and cation PCDTBT+ with the similar values of g-factor and linewidth. Similar effect was observed earlier for P3HT/2,2′-[2,2′-arenediylbis(11oxoanthra[1,2-b]thiophene-6-ylidene)]dipropanedinitrile composites [37]. Light-induced EPR decay kinetics after switching light off in PCDTBT/3b and PCDTBT/PCBM composites are also similar (See Fig. 7). Both curves can be simulated in frame of the model which suggests effective bimolecular recombination rate with reaction order higher that two (reaction order is 9.7 for composite with PCBM and 8.8 for composite with 3b) [55]. Such behaviour is usually explained by energetic disorder of BHJ composite, which means that positive and negative charges should overcome local energetic barriers before they can meet each other at donor/acceptor interface and recombine. Overall, similarity of the effective reaction order and timescale of the recombination for PCDTBT/3b and PCDTBT/PCBM composites point on the same trap-limited bimolecular recombination mechanism [56]. Fig. 7. LEPR decay kinetics reflecting charge recombination for PCDTBT/PCBM (top) and for PCDTBT/3b (bottom) and their simulation in effective bimolecular model. T = 80 K. 3.6. Photovoltaic properties To demonstrate a potential of synthesized molecules for fullerene replacement in OPV, photovoltaic devices with ITO/PEDOT:PSS (35–38 nm)/PCDTBT:3b(25–28 nm)/FM structure were fabricated. PEDOT:PSS was used as hole transporting layer and FM was used as cathode. PCDTBT/PCBM was used as reference device. Typical currentvoltage curves are shown in Fig. 8 and OPV characteristics are summarized in Table 3. For best PCDTBT/PCBM device the short-circuit current Jsc, open circuit voltage Voc, fill factor FF and power conversion efficiency PCE are 13.1 mA/cm2, 0.831 V, 39.1% and 4.25%, respectively. However, when fullerene derivative was replaced by 3b, the values of Jsc, Voc, FF and PCE became lower, namely, 3.14 mA/cm2, 0.785 V, 28.3% and 0.71%, respectively. The lower Voc can be explained by the smaller difference between the HOMO of donor and LUMO acceptor (3.97 eV for 3b and 3.7 eV for PCBM), which creates a lower driving force inside the cell. But in the same time there is sharp decrease in Jsc and FF. The extinction coefficient of 3b is larger than that of PCBM, the morphology is quite favorable for successful charge separation and transport through the active layer. Also, judging on light-induced EPR data one Fig. 8. Current density-voltage characteristics of the organic solar cell devices based on PCDTBT/PCBM (top) and PCDTBT/3b (bottom). 7 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. can suggest that efficient charge separation takes place in the PCDTBT/ 3b composite. We assume that the worse cell performance may be due to the low electron mobility in acceptor phase with respect to PCBM, which makes the charge transport unbalanced PCDTBT/3b in composite and because of this both Jsc and FF may fall. It is worth noting that for best of our knowledge, the performance of cells fabricated in this work are comparable with the best performances of OPV devices based on non-fullerene compounds, fabricated by vacuum-free method [57,58]. 994–1004, https://doi.org/10.1002/ejoc.200700970. [12] N. Martin, R. García, J. Calbo, R. Viruela, M.A. Herranz, E. Ortí, The role of planarity vs. Non-planarity in the electronic communication of TCAQ-based push-pull chromophores, Chempluschem 83 (2018) 300–307, https://doi.org/10.1002/cplu. 201700553. [13] J. Liu, Q. Meng, X. Zhang, X. Lu, P. He, L. Jiang, H. Dong, W. Hu, Aggregationinduced emission enhancement based on 11,11,12,12,-tetracyano-9,10-anthraquinodimethane, Chem. Commun. 49 (2013) 1199–1201, https://doi.org/10.1039/ c2cc38817k. [14] E.M. Pérez, A.L. Capodilupo, G. Fernéndez, L. Sénchez, P.M. Viruela, R. Viruela, E. Ortí, M. Bietti, N. Martín, Weighting non-covalent forces in the molecular recognition of C60. Relevance of concave-convex complementarity, Chem. Commun. (2008) 4567–4569, https://doi.org/10.1039/B810177A. [15] A.E. Murschell, W.H. Kan, V. Thangadurai, T.C. Sutherland, Anthraquinone derivatives as electron-acceptors with liquid crystalline properties, Phys. Chem. Chem. Phys. 14 (2012) 4626–4634, https://doi.org/10.1039/c2cp23224c. [16] G. Zerza, M.C. Scharber, C.J. Brabec, N.S. Sariciftci, R. Gómez, J.L. Segura, N. Martín, V.I. Srdanov, Photoinduced charge transfer between tetracyano-anthraquino-Dimethane derivatives and conjugated polymers for photovoltaics, J. Phys. Chem. A 104 (2000) 8315–8322, https://doi.org/10.1021/jp000729u. [17] Q. Ye, C. Chi, Recent highlights and perspectives on acene based molecules and materials, Chem. Mater. 26 (2014) 4046–4056, https://doi.org/10.1021/ cm501536p. [18] A.N. Lakshminarayana, A. Ong, C. Chi, Modification of acenes for n-channel OFET materials, J. Mater. Chem. C 6 (2018) 3551–3563, https://doi.org/10.1039/ C8TC00146D. [19] M.L. Tang, Z. Bao, Halogenated materials as organic semiconductors, Chem. Mater. 23 (2011) 446–455, https://doi.org/10.1021/cm102182x. [20] J. Dhar, U. Salzner, S. Patil, Trends in molecular design strategies for ambient stable n-channel organic field effect transistors, J. Mater. Chem. C 5 (2017) 7404–7430, https://doi.org/10.1039/C6TC05467F. [21] U.H.F. Bunz, J.U. Engelhart, B.D. Lindner, M. Schaffroth, Large N-Heteroacenes: new tricks for very old dogs? Angew. Chem. Int. Ed. 52 (2013) 3810–3821, https:// doi.org/10.1002/anie.201209479. [22] U.H.F. Bunz, The larger linear N-Heteroacenes, Acc. Chem. Res. 48 (2015) 1676–1686, https://doi.org/10.1021/acs.accounts.5b00118. [23] J.T. Markiewicz, F. Wudl, Perylene, oligorylenes, and aza-analogs, ACS Appl. Mater. Interfaces 51 (2015) 28063–28085, https://doi.org/10.1021/acsami.5b02243. [24] S. Katsuta, K. Tanaka, Y. Maruya, S. Mori, S. Masuo, T. Okujima, H. Uno, K. Nakayama, H. Yamada, Synthesis of pentacene-, tetracene- and anthracene bisimides using double-cyclization reaction mediated by bismuth(III) triflate, Chem. Commun. 47 (2011) 10112–10114, https://doi.org/10.1039/C1CC13980K. [25] H. Usta, C. Kim, Z. Wang, S. Lu, H. Huang, A. Facchetti, T.J. Marks, Anthracenedicarboximide-based semiconductors for air-stable, n-channel organic thin-film transistors: materials design, synthesis, and structural characterization, J. Mater. Chem. 22 (2012) 4459–4472, https://doi.org/10.1039/C1JM14713G. [26] P. de la Cruz, N. Martin, F. Miguel, C. Seoane, A. Albert, F.H. Cano, A. Leverenz, M. Hanack, Syntheses, electrochemical properties and crystal structure of tetracyano-p-quinodimethane (TCNQ) derivatives with π-extended systems containing a sulfur atom, Synth. Met. 48 (1992) 59–64, https://doi.org/10.1016/0379-6779(92) 90050-S. [27] Y. Ren, S. Lee, J. Bertke, D.L. Grayd, J.S. Moorea, Synthesis and structures of 11,11,12,12-tetracyano-2,6-diiodo-9,10-anthraquinodimethane and its 2:1 cocrystals with anthracene, pyrene and tetrathiafulvalene, Acta Cryst. C72 (2016) 923–931, https://doi.org/10.1107/S2053229616016387. [28] X. Shi, C. Chi, Different strategies for the stabilization of acenes and acene analogues, Chem. Rec. 16 (2016) 1690–1700, https://doi.org/10.1002/tcr.201600031. [29] A. Maliakal, K. Raghavachari, H. Katz, E. Chandross, T. Siegrist, Photochemical stability of pentacene and a substituted pentacene in solution and in thin films, Chem. Mater. 24 (2004) 4980–4986, https://doi.org/10.1021/cm049060k. [30] J. Mei, Y. Diao, A.L. Appleton, L. Fang, Z. Bao, Integrated materials design of organic semiconductors for field-effect transistors, J. Am. Chem. Soc. 135 (2013) 6724–6746, https://doi.org/10.1021/ja400881n. [31] W. Zhao, S. Li, H. Yao, S. Zhang, Y. Zhang, B. Yang, J. Hou, Molecular optimization enables over 13% efficiency in organic solar cells, J. Am. Chem. Soc. 139 (2017) 7148–7151, https://doi.org/10.1021/jacs.7b02677. [32] F. Shena, J. Xua, X. Li, C. Zhan, Nonfullerene small-molecule acceptors with perpendicular side-chains for fullerene-free solar cells, J. Mater. Chem. A 6 (2018) 15433–15455, https://doi.org/10.1039/C8TA04718A. [33] J. Yuan, Y. Zhang, L. Zhou, G. Zhang, H.-L. Yip, T.-K. Lau, X. Lu, C. Zhu, H. Peng, P.A. Johnson, M. Leclerc, Y. Cao, J. Ulanski, Y. Li, Y. Zou, Single-junction organic solar cell with over 15% efficiency using fused-ring acceptor with electron-deficient core, Joule (2019), https://doi.org/10.1016/j.joule.2019.01.004. [34] N.S. Dokunikhin, B.N. Kolokolov, Synthesis of dihaloiodoanthraquinones containing different halides, Zhurnal Vsesojuznogo Himicheskogo. Obshchstva im D.I. Mendeleeva 8 (1963) 710-711 (in Russian), Chem. Abstr. 60 (1964) 9216. [35] X. Liu, M. Zhu, S. Chen, M. Yuan, Y. Guo, Y. Song, H. Liu, Y. Li, Organic-inorganic nanohybrids via directly grafting gold nanoparticles onto conjugated copolymers through the diels-alder reaction, Langmuir 24 (2008) 11967–11974, https://doi. org/10.1021/la8020639. [36] M.S. Maji, T. Pfeifer, A. Studer, Transition-metal-Free synthesis of conjugated polymers from bis-grignard reagents by using TEMPO as oxidant, Chem. Eur. J. 16 (2010) 5872–5875, https://doi.org/10.1002/chem.201000236. [37] D.S. Baranov, M.N. Uvarov, M.S. Kazantsev, E.M. Glebov, D.A. Nevostruev, E.A. Mostovich, O.V. Antonova, L.V. Kulik, A concise and efficient route to electronaccepting 2,2′-[2,2′-arenediylbis(11-oxoanthra[1,2-b]thiophene-6-ylidene)] 4. Conclusions Novel molecules based on thiophene-fused TCAQ moieties linked through functionalized π-conjugated arene bridges are synthesized. Introduction of dicyanomethylene groups reduces energy of LUMO making these compounds strong electron acceptors, and broadens absorbance spectra in visible region. Compound with fluorene π-linker has increased solubility in common organic solvents, higher extinction coefficient in film and very smooth surface, compared to compounds with benzene linker. OPV devices based on composite of PCDTBT and compound with fluorene π-linker 3b exhibit PCE as high as 0.71%, which are among the highest values reported for vacuum-free fabricated non-fullerene OPV. Strong light-induced EPR signal in PCDTBT/ 3b composite with decay kinetics similar to that in PCDTBT/PCBM composite confirms efficient light-induced charge separation. These preliminary results demonstrate that TCNQ modified anthra[2,3-b] thiophenes are promising building blocks for non-fullerene acceptors in OPV devices. Acknowledgments The research was performed under the financial support of the Russian Science Foundation (grant no. 17-73-10144). References [1] V.A. Starodub, T.N. Starodub, Radical anion salts and charge transfer complexes based on tetracyanoquinodimethane and other strong π-electron acceptors, Russ. Chem. Rev. 83 (2014) 391–438, https://doi.org/10.1070/ RC2014v083n05ABEH004299. [2] N. Castagnetti, A. Girlando, M. Masino, C. Rizzoli, M.R. Ajayakumar, M. MasTorrent, C. Rovira, Extensive study of the electron donor 1,1,4,4-tetrathiabutadiene (TTB) and of its charge transfer crystal with TCNQ, Synth. Met. 235 (2018) 29–33, https://doi.org/10.1016/j.synthmet.2017.11.007. [3] S. Reddy Kotla, S. Kakumanu, D. Williams, K. Kharel, Ö. Günaydın-Şen, J.T. Mague, P. Chandrasekaran, Synthesis and characterization of an acenaphthene–fused, πextended tetrathiafulvalene derivative, Synth. Met. 242 (2018) 49–54, https://doi. org/10.1016/j.synthmet.2018.05.003. [4] N. Martín, J.L. Segura, C. Seoane, Design and synthesis of TCNQ and DCNQI type electron acceptor molecules as precursors for ‘organic metals’, J. Mater. Chem. 7 (1997) 1661–1676, https://doi.org/10.1039/A702314F. [5] R. Gómez, C. Seoane, J.L. Segura, The first two decades of a versatile electron acceptor building block: 11,11,12,12-tetracyano-9,10-anthraquinodimethane (TCAQ), Chem. Soc. Rev. 36 (2007) 1305–1322, https://doi.org/10.1039/ B605735G. [6] Y.-A. Cao, Y.-B. Bai, Q.-J. Men, C.-H. Chen, J.-H. Yang, X.-D. Chai, W.-S. Yang, Z.W. Wu, T.-J. Li, Study on electroluminescent behavior of Ga/Alq3/TCAQ/ITO double layer device, Synth. Met. 85 (1997) 1267–1268, https://doi.org/10.1016/ S0379-6779(97)80234-2. [7] J.L. Segura, R. Gómez, R. Blanco, E. Reinold, P. Bäuerle, Synthesis and electronic properties of anthraquinone-, tetracyanoanthraquinodimethane-, and perylenetetracarboxylic diimide-functionalized poly(3,4-ethylenedioxythiophenes), Chem. Mater. 18 (2006) 2834–2847, https://doi.org/10.1021/cm0602085. [8] R. Gómez, J.L. Segura, N. Martín, Synthesis of an optically active electron-acceptor tetracyanoanthraquinodimethane (TCAQ) main-chain polyester, Tetrahedron Lett. 47 (2006) 6445–6448, https://doi.org/10.1016/j.tetlet.2006.06.112. [9] J. Wu, Z. Jin, Q. Zhang, Synthesis and characterization of conjugated polymer based on tetracyano-anthraquinodimethane, Polym. Bull. 72 (2015) 2553–2560, https:// doi.org/10.1007/s00289-015-1420-6. [10] F. Bures, O. Pytela, F. Diederich, Solvent effects on electronic absorption spectra of donor-substituted 11,11,12,12-tetracyano-9,10-anthraquinodimethanes (TCAQs), J. Phys. Org. Chem. 22 (2009) 155–162, https://doi.org/10.1002/poc.1443. [11] F. Bureš, W.B. Schweizer, C. Boudon, J.-P. Gisselbrecht, M. Gross, F. Diederich, New push-pull chromophores featuring TCAQ (11,11,12,12-Tetracyano- 9,10-anthraquinodimethane) and other dicyanovinyl acceptors, Eur. J. Org. Chem. 6 (2008) 8 Synthetic Metals 255 (2019) 116097 D.S. Baranov, et al. [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] 1252–1259, https://doi.org/10.1007/s10593-007-0191-x. [49] P. De la Cruz, N. Martin, F. Miguel, C. Seoane, A. Albert, F.H. Cano, A. Gonzalez, J.M. Pingarron, Novel π-Extended thiophene-fused Electron acceptors for organic metals, J. Org. Chem. 57 (1992) 6192–6198, https://doi.org/10.1021/ jo00049a027. [50] B. Djukic, D.F. Perepichka, Non-classical heteroacenes: synthesis and properties of anthra[2,3-c:6,7-c´]dithiophene derivatives, Chem. Commun. 47 (2011) 12619–12621, https://doi.org/10.1039/C1CC15623C. [51] M. Gruber, K. Padberg, J. Min, A.R. Waterloo, F. Hampel, H. Maid, T. Ameri, C.J. Brabec, R.R. Tykwinski, Acenequinocumulenes: lateral and vertical π-Extended analogues of tetracyanoquinodimethane (TCNQ), Chem. Eur. J. 23 (2017) 17829–17835, https://doi.org/10.1002/chem.201704314. [52] Y. Zhao, D.G. Truhlar, Density functionals with broad applicability in chemistry, Acc. Chem. Res. 41 (2008) 157–167, https://doi.org/10.1021/ar700111a. [53] F. Di Meo, P. Trouillas, C. Adamo, J.C. Sancho-Garcia, Application of recent doublehybrid density functionals to low-lying singlet-singlet excitation energies of large organic compounds, J. Chem. Phys. 139 (2013) 164104, , https://doi.org/10.1063/ 1.4825359. [54] S. Wakim, S. Beaupre, N. Blouin, B.-R. Aich, S. Rodman, R. Gaudiana, Y. Tao, M. Lecterc, Highly efficient organic solar cells based on a poly(2,7-carbazole) derivative, J. Mater. Chem. 19 (2009) 5351–5358, https://doi.org/10.1039/ b901302d. [55] E.A. Lukina, M.N. Uvarov, L.V. Kulik, Charge recombination in P3HT/PC60BM composite studied by light-induced EPR, J. Phys. Chem. C 118 (2014) 18307–18314, https://doi.org/10.1021/jp502299c. [56] K. Seki, K. Marumoto, M. Tachiya, Bulk recombination in organic bulk heterojunction solar cells under continuous and pulsed light irradiation, Appl. Phys. Express. 6 (2013) 051603, , https://doi.org/10.7567/APEX.6.051603. [57] P. Cheng, H.T. Bai, N.K. Zawacka, T.R. Andersen, W.Q. Liu, E. Bundgaard, M. Jorgensen, H. Chen, F.C. Krebs, X.W. Zhan, Roll-coated fabrication of fullerenefree organic solar cells with improved stability, Adv. Sci. 2 (2015) 1500096, , https://doi.org/10.1002/advs.201500096. [58] K. Liu, T.T. Larsen-Olsen, Y.Z. Lin, M. Beliatis, E. Bundgaard, M. Jorgensen, F.C. Krebs, X.W. Zhan, Roll-coating fabrication of flexible organic solar cells: comparison of fullerene and fullerene-free systems, J. Mater. Chem. A 4 (2016) 1044–1051, https://doi.org/10.1039/c5ta07357j. dipropanedinitriles, Eur. J. Org. Chem. (2018) 2259–2266, https://doi.org/10. 1002/ejoc.201800275. D. Barreiro-Arguelles, G. Ramos-Ortiz, J.L. Maldonado, E. Perez-Gutierrez, D. Romero-Borja, A. Alvarez-Fernandez, PTB7: PC71BM-Based solar cells fabricated with the eutectic alloy field’s metal as an alternative cathode and the influence of an Electron extraction layer, IEEE J. Photovolt 7 (2017) 191–198, https://doi.org/10. 1109/JPHOTOV.2016.2617087. A.D. Becke, Density-functional thermochemistry. III. The role of exact exchange, J. Chem. Phys. 98 (1993) 5648–5652, https://doi.org/10.1063/1.464913. C. Lee, W. Yang, R.G. Parr, Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density, Phys. Rev. B 37 (1988) 785–789, https://doi.org/10.1103/PhysRevB.37.785. F. Weigend, R. Ahlrichs, Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: design and assessment of accuracy, Phys. Chem. Chem. Phys. 7 (2005) 3297–3305, https://doi.org/10.1039/ B508541A. F. Neese, Software update: the ORCA program system, version 4.0, WIREs Comput. Mol. Sci. 8 (2018) e1327, https://doi.org/10.1002/wcms.1327. A. Dreuw, M. Head-Gordon, Single-reference ab initio methods for the calculation of excited states of large molecules, Chem. Rev. 105 (2005) 4009–4037, https://doi. org/10.1021/cr0505627. Y. Kita, S. Mohri, T. Tsugoshi, H. Maeda, Y. Tamura, Reaction of heteroaromatic analogs of homophthalic anhydride : synthesis of hetero analogs of peri-hydroxy polycyclic aromatic compounds, isocoumarins, isoquinolinones, and related compounds, Pharm. Bull. 33 (1985) 4723–4731, https://doi.org/10.1248/cpb.33.4723. L.M. Chaloner, A.P.A. Crew, P.M. O’Neill, R.C. Storr, M. Yelland, Heterocyclic fused 2,5-Dihydrothiophene S,S-Dioxides as precursors to heterocyclic oQuinodimethanes, Tetrahedron 48 (1992) 8101–8116, https://doi.org/10.1016/ S0040-4020(01)80480-4. K.J. van den Berg, A.M. van Leusen, Formation and [4+2] cycloaddition reactions of 2,3-dimethylene-2,3-dihydrothiophene, Red. Trav. Chim. Pays-Bas 112 (1993) 007–014, https://doi.org/10.1002/recl.19931120103. M.L. Tang, T. Okamoto, Z. Bao, High-performance organic semiconductors: asymmetric linear acenes containing sulphur, J. Am. Chem. Soc. 128 (2006) 16002–16003, https://doi.org/10.1021/ja066824j. Y.B. Sinkevich, A.E. Shchekotikhin, Y.N. Luzikov, V.N. Buyanov, L.V. Kovalenko, Synthesis of thiopheno-quinizarine derivatives, Chem. Het. Comp. 43 (2007) 9