Optical, Electrochemical, Thermal, and Structural Properties of Synthesized Fluorene / Dibenzosilole-Benzothiadiazole Dicarboxylic Imide Alternating Organic Copolymers for Photovoltaic Applications

: In this work, four donor–acceptor copolymers, PFDTBTDI-DMO, PFDTBTDI-8, PDBSDTBTDI-DMO, and PDBSDTBTDI-8, based on alternating 2,7-ﬂuorene or 2,7-dibenzosilole ﬂanked by thienyl units, as electron-donor moieties and benzothiadiazole dicarboxylic imide (BTDI) as electron-accepting units, have been designed and synthesized for photovoltaic applications. All polymers were synthesized in good yields via Suzuki polymerization. The impact of attaching two di ﬀ erent alkyl chains (3,7-dimethyloctyl vs. n -octyl) to the BTDI units upon the solubilities, molecular weights, optical and electrochemical properties, and thermal and structural properties of the resulting polymers was investigated. PFDTBTDI-8 has the highest number average molecular weight ( M n = 24,900 g · mol − 1 ) among all polymers prepared. Dibenzosilole-based polymers have slightly lower optical band gaps relative to their ﬂuorene-based analogues. All polymers displayed deep-lying HOMO levels. Their HOMO energy levels are una ﬀ ected by the nature of either the alkyl substituents or the donor moieties. Similarly, the LUMO levels are almost identical for all polymers. All polymers exhibit excellent thermal stability with T d exceeding 350 ◦ C. X-ray powder di ﬀ raction (XRD) studies have shown that all polymers have an amorphous nature in the solid state.


Introduction
Solution-processable polymer solar cells (PSCs) have received substantial consideration as a renewable energy source due to their benefits such as being flexible devices, light weight, low costs, and easy fabrication [1][2][3][4]. The most successful method to build an active layer of PSCs is based on the The photovoltaic performance of this polymer with PC 61 BM as the acceptor delivered a PCE of 1.6%. Soon after, Cao et al. independently reported a higher PCE of 5.4% for the same polymer using 4-fold higher M n (79,000 g·mol −1 ) [35]. The improved performance of PDBSDTBT compared to PFDTBT analogue was due to higher hole mobility measured by FET and broader absorption spectrum. In addition, the C-Si bond in DBS unit is longer than C-C bond in fluorene unit. Consequently, DBS units create less steric hindrance compared to their fluorene counterparts and therefore a better π-π stacking between polymer chains is expected [36].

Materials
All the starting materials and reagents obtained from Sigma-Aldrich and Alfa Aesar and used without further purification. Most of the reactions were carried out under argon atmosphere. Anhydrous solvents used for the reactions obtained from Grubbs solvent purification system within the Sheffield University/Chemistry Department. All the monomers used for preparing the polymers in this article were synthesized according to the procedure part.

Measurements
All 1 H NMR and 13 C NMR nuclear magnetic resonance (NMR) spectra for the monomers measured either with a Bruker Avance AV 3HD 400 (400 MHz) spectrometer in deuterated chloroform (CDCl 3 ), deuterated acetone (CD 3 COCD 3 ) or deuterated dimethyl sulfoxide (CD 3 SOCD 3 ) as the solvents at room temperature. The 1 H NMR spectra for the polymers were measured with Bruker AV 3HD 500 (500 MHz) in deuterated 1,1,2,2-tetrachloroethane (C 2 D 2 Cl 4 ) as the solvent at 100 • C. The chemical shifts were measured in parts per million (ppm). The coupling constants (J) are calculated in Hertz (Hz). The 1 H NMR and 13 C NMR spectra were analyzed using Bruker TopSpin 3.2 software. Elemental analysis (CHN) was performed by either the Perkin Elmer 2400 CHNS/O Series II Elemental Analyzer or Vario MICRO Cube CHN/S Elemental Analyzer for CHN analysis. Anion analysis (Br, I, and S) was performed by the Schöniger oxygen flask combustion method. Mass spectra for the monomers were recorded on Agilent 7200 accurate mass Q-TOF GC-MS spectrometer. Helium is used as a carrier gas in rate of (1.2 mL·min −1 ), the injection volume is (1.0 µL) and the concentration of measured sample is (5 mg·mL −1 ) in CHCl 3 solvent. The temperature program is between 60 to 320 • C at 10 • C·min −1 . Mass spectra for the monomers were obtained by the electron ionization method (EI). Gel permeation chromatography (GPC) measurements accomplished by Viscotek GPC Max, a waters 410 instrument with a differential refractive index detector, two polymer labs PLgel 5µ Mixed C (7.5 × 300 mm) columns and a guard (7.5 × 50 mm). Molecular weights for the polymers were determined by preparing polymer solutions (2.5 mg·mL −1 ) using HPLC grade CHCl3. The columns were thermostated at 40 • C using CHCl 3 . UV-vis absorption spectra were measured by SPECORD S600 UV/visible Spectrophotometer at room temperature. The absorbance of the polymers was measured in CHCl 3 solution using quartz cuvettes (light path length = 10 mm) and blank quartz cuvettes including CHCl 3 was used as a reference. The polymers were coated on quartz substrates from CHCl 3 solutions (1 mg·mL −1 ) and blank quartz substrate was used as a reference. Thermogravimetric analysis (TGA) measurements were recorded by Perkin Elmer (Pyris 1) thermogravimetric Analyzer. Platinum pans were used as sample holder and the weight of the measured samples was about (3 mg). Cyclic voltammograms were measured using a Model 263A Potentiostat/Galvanostat-Princeton Applied Research. A standard Coatings 2020, 10, 1147 5 of 22 three electrode system was used based on a Pt disk working electrode, a silver wire reference electrode (Ag/Ag + ) inserted in (0.01 M) AgNO 3 solution in acetonitrile and put it in the electrolyte solution and a Pt wire counter electrode was purged with argon atmosphere during all measurements at room temperature. Tetrabutylammonium perchlorate in acetonitrile (Bu 4 NClO 4 /CH 3 CN) (0.1 M) was used as the electrolyte. Polymer thin films were drop cast onto the Pt disk from polymer solutions in CH 3 Cl (1 mg·mL −1 ) and dried under nitrogen prior to measurement. Ferrocene (Fc/Fc + ) was used as a reference redox system. Powder X-ray diffraction (XRD) for the polymers was measured by Bruker D8 ADVANCE X-ray powder diffractometer. Infrared absorption spectra recorded on ATR Perkin Elmer Rx/FT-IR system and Nicolet Model 205 FT-IR spectrometer.

Monomers and Polymers Synthesis
2.3.1. Synthesis of 2,5-dibromothiophene (1) Thiophene (25.00 g, 297.12 mmol) in DMF (250 mL) added to a flask and cooled to −15 • C. To this solution, NBS (110.00 g, 618.04 mmol) in DMF (300 mL) was added dropwise in the dark and the reaction stirred overnight at RT. The reaction contents were put into ice and DCM and subsequently extracted with DCM and the organic phase washed with deionized H 2 O to a neutral pH. The organic layer was collected and dried over MgSO 4 and the solvent concentrated to afford the product which purified by vacuum distillation and gave 1 as a yellow oil (59.30 g, 245 mmol, 82% yield) [37]. 1 (2) Concentrated H 2 SO 4 (150 mL) and fuming H 2 SO 4 (150 mL, 20% free SO 3 ) combined in a flask. This flask was cooled to 0 • C and 1 (26.00 g, 107.46 mmol) was added dropwise. Concentrated nitric acid (125 mL) was added dropwise and the reaction contents kept under 20 • C. During addition of nitric acid yellow precipitate formed quickly. The mixture was stirred for 3 h at 20-30 • C. Then, the mixture was poured into ice and upon melting of the ice a yellow precipitate filtrated and washed thoroughly with deionized H 2 O. The product recrystallized from methanol to afford 2 as yellow crystals (32.50 g, 98 mmol, 91% yield) [38]. 13 (3) In a flask, 2 (9.90 g, 29.82 mmol), 2-(tributylstannyl)thiophene (27.82 g, 74.54 mmol) and PdCl 2 (PPh 3 ) 2 (0.45 g, 0.64 mmol) added. The system was degassed under argon and anhydrous toluene (100 mL) added and heated at 115 • C for 24 h. The flask was cooled to RT and the volatiles removed to obtain the product which purified by column chromatography with gradient (petroleum ether, 0-50% DCM) to obtain an orange solid and the product further purified by recrystallization from methanol to afford 3 as orange crystals (9.10 g, 27 mmol, 90% yield) [39]. 1  then stirred vigorously and filtered through celite. The product was extracted with toluene and the organic phases washed with NaCl and subsequently dried over MgSO 4 . The solvent was concentrated to obtain the 4 as a brown solid (2.40 g, 9 mmol, 97% yield) [40]. 1 (5) 4 (1.67 g, 5.99 mmol) dissolved in dry pyridine (30 mL) in a flask and degassed under argon. To this mixture, N-thionylaniline (1.60 g, 11.49 mmol) was added drop wise and chlorotrimethylsilane (4.50 g, 41.42 mmol) then added drop wise, resulting in a dark blue color. The reaction contents were stirred for 3 h at RT and then put into DCM. The solution was washed with HCl and with deionized water and extracted with DCM. The organic phase was dried over anhydrous MgSO 4 and subsequently filtered. The solvent was evaporated to afford the product which purified via chromatography with DCM to afford 5 as blue crystals (1.72 g, 6 mmol, 93% yield) [41]. 1  2.3.6. Synthesis of 4,7-di(thien-2-yl)-2,1,3-benzothiadiazole-5,6-dimethyl Ester (6) 5 (1.86 g, 6.06 mmol) and dimethyl acetylenedicarboxylate (1.73 g, 12.17 mmol) was combined in a flask. The system was evacuated and refilled with argon for three cycles before anhydrous xylene (40 mL) added. The reaction contents were refluxed for 24 h. The flask was cooled to RT and the solvent removed to afford the product which was purified by column chromatography with gradient (petroleum ether, 0-50% DCM) to afford 6 as yellow crystals (2.37 g, 6 mmol, 94% yield) [23]. 1  2.3.7. Synthesis of 4,7-di(thien-2-yl)-2,1,3-benzothiadiazole-5,6-dicarboxylic Acid (7) Sodium hydroxide (4.00 g, 100.00 mmol) was dissolved in deionized water (30 mL) and added to a flask. To this solution, ethanol (200 mL) and 6 (2.27 g, 5.45 mmol) was added and the reaction contents refluxed for 24 h. The flask was cooled to RT and deionized H 2 O added. This mixture was cooled to 0 • C and neutralized by HCl to precipitate the product. The precipitate was filtered and subsequently washed with deionized H 2 O. The precipitate was dried under high vacuum to afford 7 as yellow solid (1.80 g, 5 mmol, 85% yield) [20]. 1 13  2.3.8. Synthesis of 4,7-di(thien-2-yl)-2,1,3-benzothiadiazole-5,6-dicarboxylic Anhydride (8) 7 (1.15 g, 2.96 mmol) and anhydrous acetic anhydride (10.00 g, 97.95 mmol) combined in a flask. The system was evacuated and refilled with argon for three cycles before anhydrous xylene (30 mL) added. The mixture was heated at 130 • C for 6 h. The mixture was cooled to RT, and the solvent evaporated to obtain 8 as red solid (1.06 g, 3 mmol, 97% yield) [42]. 1 13 (9) Triphenylphosphine (21.10 g, 80.44 mmol) was added to a mixture of 3,7-dimethyloctyl alcohol (12.61 g, 79.69 mmol) and dichloromethane (250 mL) and stirred in a flask. To this mixture, NBS (14.26 g, 80.14 mmol) was added portionwise and stirred at RT for 90 min. The mixture was washed with NaHCO 3 solution, dried over MgSO 4 , filtered, and the solvent evaporated. The substance was stirred in petroleum ether for 1 h at RT, filtered and the filtrate evaporated. The product was purified by chromatography with petroleum ether to yield 9 as colorless oil (23.00 g, 59 mmol, 73% yield) [43]. 1 (10) 9 (4.07 g, 18.40 mmol) and anhydrous DMF (20 mL) added into a flask. To this mixture, potassium phthalimide (3.75 g, 20.27 mmol) was added and the reaction contents heated to 90 • C for 17 h. The mixture was cooled to RT and put in deionized H 2 O and the product subsequently extracted with DCM. The organic extracts were combined, washed with KOH and deionized water. The organic phase was dried over MgSO 4 and the solvent was evaporated to obtain the product which was purified via chromatography with dichloromethane to yield 10 as colorless oil (5.29 g, 18 mmol, 91% yield) [44]. 1  2.3.11. Synthesis of 3,7-dimethyl-1-octanamine (11) 10 (6.03 g, 20.98 mmol), hydrazine hydrate (4.0 mL, 65.0 mmol, 51%) and methanol (100 mL) were combined in a flask. The reaction contents were refluxed until the starting material disappeared. Upon completion, excess HCl was added and the mixture refluxed for 1 h and then cooled to RT. The precipitate was filtered and washed with water. The methanol was concentrated and the residue diluted with dichloromethane. The organic layer was washed with KOH and the product extracted with dichloromethane. The organic phase was washed with NaCl, dried over MgSO 4 and the solvent was concentrated to yield 11 as a brown oil (2.85 g, 18 mmol, 86% yield) [45]. 1  2.3.12. Synthesis of 4,7-di(thien-2-yl)-2,1,3-benzothiadiazole-5,6-N-(3,7-dimethyloctyl)dicarboxylic Imide (12) 8 (1.00 g, 2.69 mmol), acetic acid (50 mL, 100%) and 11 (0.88 g, 5.59 mmol) was combined in a flask. The system was evacuated and refilled with argon for three cycles and heated at 110 • C overnight. The mixture was cooled to RT, acetic anhydride (20 mL) added and heated at 110 • C for 6 h. The mixture was cooled to RT and the solvent concentrated to yield the product which was purified by 2.3.13. Synthesis of 4,7-di(thien-2-yl)-2,1,3-benzothiadiazole-5,6-N-octyl-dicarboxylic imide (13) 13 was prepared followed by the same procedure for synthesis of 12 except N-octylamine (1.20 g, 9.28 mmol) was used. 13 was obtained as an orange solid (1.20 g, 2.5 mmol, 93% yield) [42]. 1  12 (1.00 g, 1.96 mmol) and THF (100 mL) was combined in a flask. To this mixture, NBS (1.74 g, 9.77 mmol) was added and stirred at RT overnight in the dark. The solvent was evaporated to obtain the product as red solid, subsequently washed with cold CH 3 OH, filtered and dried. The product was purified via chromatography with DCM to yield M1 as red solid (1.28 g, 2 mmol, 98% yield) [20]. 1  2.3.15. Synthesis of 4,7-di(5-bromo-thien-2-yl)-2,1,3-benzothiadiazole-5,6-N-octyl-dicarboxylic Imide (M2) M2 was prepared followed by the same procedure for synthesis of M1, where compound 13 (1.00 g, 2.07 mmol), used with THF (100 mL) and NBS (1.84 g, 10.33 mmol). M2 was obtained as red solid (1.27 g, 2 mmol, 96% yield) [20]. 1  2,5-Dibromonitrobenzene (50.00 g, 179.30 mmol) and Cu powder (25.00 g, 393.39 mmol) was combined in a flask. The mixture was evacuated and refilled with argon for three cycles before anhydrous DMF (230 mL) added and heated at 125 • C for 3 h. The reaction contents were cooled to RT and dissolved in toluene and filtered. NaHCO 3 solution was added to the filtrate and the mixture extracted with toluene and the organic phases combined and washed with deionized H 2 O numerous times until it became neutral. The organic phase was dried over anhydrous MgSO 4 , filtered and the solvent evaporated to afford the product which was recrystallized from isopropanol to yield 14 as yellow crystals (32.61 g, 162 mmol, 90% yield) [46]. 1 (15) 14 (6.00 g, 14.92 mmol), ethanol (74 mL) and HCl (43 mL, 32%) was added to a flask. To this mixture, Sn (7.00 g, 58.96 mmol) was added over 10 min and refluxed for 90 min. The reaction contents were cooled to RT and another portion of Sn (7.00 g, 58.96 mmol) added and refluxed for 90 min. The flask was cooled to RT and the mixture filtered and deionized water added to the filtrate. NaOH solution was added dropwise until pH became approximately 9. The mixture was extracted with Et 2 O and organic phase washed with NaCl, dried over MgSO 4 and then filtered. The solvent was concentrated to obtain the 15 as brown crystals (3.58 g, 10.5 mmol, 70% yield) [47]. 1 (16) 15 (5.00 g, 14.61 mmol), HCl (50 mL, 32 %), H 2 O (200 mL) and acetonitrile (200 mL) were combined in a flask and cooled to 0 • C. To this mixture, NaNO 2 (4.59 g, 66.53 mmol) was dissolved in deionized H 2 O (25 mL), added dropwise, and stirred for 1 hour between -5 and -10 • C. KI (22.28 g, 134.21 mmol), dissolved in deionized H 2 O (50 mL), and cooled to 0 • C and added dropwise while the reaction temperature maintained at -10 to -15 • C. After addition was completed, the temperature was raised to RT and then heated to 80 • C for 20 h and cooled to RT. The mixture was extracted with DCM, the collected organic layers washed with Na 2 S 2 O 3 solution, deionized water and NaCl solution. The collected organic layers were dried over MgSO 4 , filtered, and the solvent was concentrated to afford the product which was purified via chromatography with petroleum ether. It was further purified by recrystallization from n-hexane to yield 16 as white crystals (4.23 g, 7.5 mmol, 51% yield) [48]. 1 13 (17) 16 (4.20 g, 7.44 mmol) was added to a flask. Anhydrous THF (84 mL) was added and the mixture cooled to −78 • C and then the system degassed under argon. To this mixture, n-BuLi (12.00 mL, 30.0 mmol) was added dropwise over 2 h. The reaction contents were stirred for 1 h, and dichlorodioctylsilane (4.86 g, 14.95 mmol) was added dropwise over 5 min and the temperature raised to RT and mixture stirred overnight. Deionized water was added, and the product extracted with Et 2 O and organic layers collected and washed with brine. The organic layer was dried over MgSO 4 , filtered and the solvent concentrated to yield a product which purified using chromatography with petroleum ether to yield 17 as a colorless oil (3.80 g, 7 mmol, 90% yield) [49]. 1 13 9.29 mmol) and PdCl 2 (dppf) (0.12 g, 0.14 mmol, 5.54 mol%) were combined in a flask and then the system degassed under argon. To this mixture, anhydrous DMF (30 mL) was added and heated at 100 • C for 48 h. The flask was cooled to RT and the product extracted with Et 2 O and organic phases washed with deionized H 2 O. The organic layers were separated and dried over MgSO 4 , filtered and the solvent evaporated to obtain a product which was recrystallized from methanol which passed through the basic alumina to remove the acidic protons to obtain M4 as brown crystals (0.92 g, 1.4 mmol, 53% yield) [50]. 1  Anhydrous THF (10 mL) was followed by sodium hydrogen carbonate solution (2.5 mL, 5% wt., degassed), added, and the system degassed again. To this mixture, Pd(OAc) 2 (3.7 mg, 0.0168 mmol) and P(o-tol) 3 (10.2 mg, 0.0336 mmol) was added, degassed, and heated at 90 • C for 30 h. The flask was cooled to RT, the polymer dissolved in CHCl 3 (200 mL) and an NH 4 OH solution (50 mL, 35% in H 2 O) added and the mixture stirred overnight. The organic phase was separated and washed with deionized H 2 O. The organic phase was concentrated to around (50 mL) and put into methanol (300 mL) and stirred overnight. The mixture was filtered and the polymer cleaned using Soxhlet extraction with methanol (300 mL), acetone (300 mL), hexane (300 mL) and then toluene (300 mL). The toluene fraction was concentrated to around (50 mL) and then put into methanol (300 mL). The mixture was stirred overnight, and the pure polymer was recovered by filtration to afford PFDTBTDI-DMO as purple powders (170 mg, 0.18 mmol, 85% yield) [21].
xii) xiii) xii) xiii)  Reagents and conditions: (i) anhydrous THF, NaHCO3, Pd(OAc)2, P(o-tol)3, 90 °C, 21-30 h. Molecular weights of the polymers were measured by GPC using chloroform at 40 °C relative to polystyrene standards, as shown in Table 1. Substituting 3,7-dimethyloctyl chains in PFDTBTDI-DMO for octyl chains in PFDTBTDI-8 on the BTDI building blocks results in a polymer with lower Mn values for the toluene fractions of the polymers. However, PFDTBTDI-8 afforded another fraction in chloroform of a higher Mn value that was not soluble in toluene while PFDTBTDI-DMO did not provide chloroform fraction. The results indicate that a higher solubility of the polymer with 3,7-dimethyloctyl chains as a result of the branching of its substituents. The toluene fractions of the dibenzosilole-based polymers (PDBSDTBTDI-DMO and PDBSDTBTDI-8) have similar Mn values. However, PDBSDTBTDI-8 has a lower Mn value to that of PDBSDTBTDI-DMO for the chloroform fractions. This could be attributed to the effect of the branched chains in PDBSDTBTDI-DMO which provides it with a greater solubility and allows it to provide higher molecular weight fractions. Polymers with n-octyl chains provided lower yields compared to those with branched chains. Moreover, the yield of fluorene-based polymers is higher relative to dibenzosilole-based polymers. This could be due to more aggregation in dibenzosilole-based polymers with more intermolecular interactions relative to fluorene-based polymers. Reagents and conditions: (i) anhydrous THF, NaHCO 3 , Pd(OAc) 2 , P(o-tol) 3 , 90 • C, 21-30 h. Molecular weights of the polymers were measured by GPC using chloroform at 40 • C relative to polystyrene standards, as shown in Table 1. Substituting 3,7-dimethyloctyl chains in PFDTBTDI-DMO for octyl chains in PFDTBTDI-8 on the BTDI building blocks results in a polymer with lower M n values for the toluene fractions of the polymers. However, PFDTBTDI-8 afforded another fraction in chloroform of a higher M n value that was not soluble in toluene while PFDTBTDI-DMO did not provide chloroform fraction. The results indicate that a higher solubility of the polymer with 3,7-dimethyloctyl chains as a result of the branching of its substituents. The toluene fractions of the dibenzosilole-based polymers (PDBSDTBTDI-DMO and PDBSDTBTDI-8) have similar M n values. However, PDBSDTBTDI-8 has a lower M n value to that of PDBSDTBTDI-DMO for the chloroform fractions. This could be attributed to the effect of the branched chains in PDBSDTBTDI-DMO which provides it with a greater solubility and allows it to provide higher molecular weight fractions. Polymers with n-octyl chains provided lower yields compared to those with branched chains. Moreover, the yield of fluorene-based polymers is higher relative to dibenzosilole-based polymers. This could be due to more aggregation in dibenzosilole-based polymers with more intermolecular interactions relative to fluorene-based polymers.

Optical Properties
The normalized UV-vis absorption spectra of all polymers in chloroform solutions and in thin films are shown in Figure 1. The optical properties of these polymers are summarized in Table 2. All polymers show two absorption bands at short and long wavelengths. The peak at shorter wavelengths could be related to π-π* transition. However, the other band at lower energy is related to the intramolecular charge transfer (ICT) between donor (D) and acceptor (A) units. The band gaps (E g ) of the polymers are assessed from the absorption onsets in thin films. In solutions, all polymers display comparable absorption maxima. In thin films, the absorption spectra of the polymers show red-shifted absorption maxima by 22-34 nm relative to their absorption in solutions. This could be explained by stronger intermolecular π-π interaction and a more planar structure in the solid state. Compared with fluorene-based polymers, dibenzosilole-based polymers have broader absorption bands and therefore lower E g values. A change of alkyl chains on BTDI units from 3,7-dimethyloctyl chains to n-octyl chains has a negligible impact on the E g of the resulting polymers. PFDTBTDI-8 is red-shifted relative to its PFDTBTDI-DMO analogue; this may arise from the fact that the former polymer has a higher molecular weight than the latter polymer.

Optical Properties
The normalized UV-vis absorption spectra of all polymers in chloroform solutions and in thin films are shown in Figure 1. The optical properties of these polymers are summarized in Table 2. All polymers show two absorption bands at short and long wavelengths. The peak at shorter wavelengths could be related to π-π* transition. However, the other band at lower energy is related to the intramolecular charge transfer (ICT) between donor (D) and acceptor (A) units. The band gaps (Eg) of the polymers are assessed from the absorption onsets in thin films. In solutions, all polymers display comparable absorption maxima. In thin films, the absorption spectra of the polymers show red-shifted absorption maxima by 22-34 nm relative to their absorption in solutions. This could be explained by stronger intermolecular π-π interaction and a more planar structure in the solid state. Compared with fluorene-based polymers, dibenzosilole-based polymers have broader absorption bands and therefore lower Eg values. A change of alkyl chains on BTDI units from 3,7-dimethyloctyl chains to n-octyl chains has a negligible impact on the Eg of the resulting polymers. PFDTBTDI-8 is red-shifted relative to its PFDTBTDI-DMO analogue; this may arise from the fact that the former polymer has a higher molecular weight than the latter polymer.    PFDTBTDI-DMO and PFDTBTDI-8 have lower E g relative to P(BTI-F) which has two extra thiophene spacers between fluorene and DTBTD I units. The latter polymer is blue-shifted around 10-25 nm relative to the former polymers. PFDTBTDI-DMO and PFDTBTDI-8 have lower E g values around 0.1 eV compared with that of PFDTBT due to the stronger electron-accepting strength of the BTDI building blocks than BT unit [18]. Similarly, PDBSDTBTDI-DMO and PDBSDTBTDI-8 have lower E g relative to PDBSDTBT analogue [35]. The absorption coefficients (ε) of the dibenzosilole-based polymers are higher than fluorene-based polymers as illustrated in Table 2. This indicates that all things been equal, the dibenzosilole polymers should lead to more efficient OPV devices.

Electrochemical Properties
Cyclic voltammetry (CV) is considered to be an effective technique to study the electroactivity of active species and it is also suitable technique for characterizing the oxidation and reduction potentials of the different phases present in the conducting polymers [53]. In addition, CV can be employed to analyze the charge storage mechanism and the charge transfer between the electrodes in the electrochemical double layer capacitor (EDLC) [54][55][56]. Therefore, in this work CV was used to study the electrochemical properties of the polymers. The LUMO and HOMO levels of all polymers calculated from the onsets of reduction and oxidation potentials, respectively ( Figure 2 and Table 3 PFDTBTDI-DMO and PFDTBTDI-8 have lower Eg relative to P(BTI-F) which has two extra thiophene spacers between fluorene and DTBTD I units. The latter polymer is blue-shifted around 10-25 nm relative to the former polymers. PFDTBTDI-DMO and PFDTBTDI-8 have lower Eg values around 0.1 eV compared with that of PFDTBT due to the stronger electron-accepting strength of the BTDI building blocks than BT unit [18]. Similarly, PDBSDTBTDI-DMO and PDBSDTBTDI-8 have lower Eg relative to PDBSDTBT analogue [35]. The absorption coefficients (ε) of the dibenzosilole-based polymers are higher than fluorene-based polymers as illustrated in Table 2. This indicates that all things been equal, the dibenzosilole polymers should lead to more efficient OPV devices.

Electrochemical Properties
Cyclic voltammetry (CV) is considered to be an effective technique to study the electroactivity of active species and it is also suitable technique for characterizing the oxidation and reduction potentials of the different phases present in the conducting polymers [53]. In addition, CV can be employed to analyze the charge storage mechanism and the charge transfer between the electrodes in the electrochemical double layer capacitor (EDLC) [54][55][56]. Therefore, in this work CV was used to study the electrochemical properties of the polymers. The LUMO and HOMO levels of all polymers calculated from the onsets of reduction and oxidation potentials, respectively ( Figure 2 and Table 3). The onsets were determined from cyclic voltammograms on drop cast polymer films on Pt electrode as working electrode in Bu4NClO4/CH3CN (0.1 M) vs. Ag/Ag + reference electrode. All polymers show the same HOMO energy levels. The results indicate that switching from fluorene to DBS moiety does not alter the HOMO levels of the resulting polymers. The HOMO level is dominated by the nature of the donor unit and that both fluorene and DBS units are weak electron donors of comparable strength. All polymers display low-lying HOMO energy levels which are beneficial for the chemical stability of polymers in oxygen and should lead to higher Voc values of the fabricated OPV devices including these polymers as donor materials. All polymers have nearly identical LUMO energy levels, as all polymers have the same BTDI acceptor units which control the LUMO levels in these materials. Furthermore, anchoring different alkyl chains on BTDI units has little impact on the LUMO levels of the resulting polymers.    All polymers have comparable HOMO levels relative to P(BTI-F), which has a HOMO level of −5.60 eV [24]. This indicates that the incorporation of two extra thiophene spacers between fluorene and DTBTDI units has negligible effect on the HOMO levels of the resulting polymers.
However, all polymers have deeper HOMO levels compared to those of PDI-BDTT , PDI-BDTO, BBTI-1, and BBTI-2. This could be attributed to fluorene and DBS units being weaker electron donors than BDT and BTT units [20,23]. The LUMO levels of the polymers are higher than those of PFDTBT and PDBSDTBT (−3.8 eV), which are related to the stronger electron-accepting ability of BTDI moiety than BT unit [18,35].

Thermal Properties
The thermal properties of the polymers were studied by TGA (Figure 3 and Table 3). All polymers show high thermal stability with T d up to 350 • C. The thermal stability of the polymers with n-octyl chains on BTDI moiety is significantly lower than those with 3,7-dimethyloctyl chains. The results show that the thermal properties of the polymers are mostly affected by the size of the alkyl chains anchored to the BTDI units as well as by the type of the donor building blocks.
Coatings 2020, 10, x FOR PEER REVIEW 17 of 22 All polymers have comparable HOMO levels relative to P(BTI-F), which has a HOMO level of −5.60 eV [24]. This indicates that the incorporation of two extra thiophene spacers between fluorene and DTBTDI units has negligible effect on the HOMO levels of the resulting polymers. However, all polymers have deeper HOMO levels compared to those of PDI-BDTT, PDI-BDTO, BBTI-1, and BBTI-2. This could be attributed to fluorene and DBS units being weaker electron donors than BDT and BTT units [20,23]. The LUMO levels of the polymers are higher than those of PFDTBT and PDBSDTBT (−3.8 eV), which are related to the stronger electron-accepting ability of BTDI moiety than BT unit [18,35].

Thermal Properties
The thermal properties of the polymers were studied by TGA (Figure 3 and Table 3). All polymers show high thermal stability with Td up to 350 °C. The thermal stability of the polymers with n-octyl chains on BTDI moiety is significantly lower than those with 3,7-dimethyloctyl chains. The results show that the thermal properties of the polymers are mostly affected by the size of the alkyl chains anchored to the BTDI units as well as by the type of the donor building blocks.

Powder X-ray Diffraction (XRD)
Since a single polymer can come in various forms including crystalline, amorphous and semi-crystalline phases, which effect on its mechanical and electrical properties, it is necessary to accurately determine the structural properties using XRD technique [57][58][59]. The structural properties of the prepared polymers were investigated by powder XRD in solid state as shown in Figure 4. The XRD of the PFDTBTDI-DMO, PFDTBTDI-8, PDBSDTBTDI-DMO, and PDBSDTBTDI-8 show diffraction peaks at 20.0°, 20.5°, 19.7° and 20.3° corresponding to the π-π stacking distance of 4.43 Å, 4.32 Å, 4.5 Å and 4.36 Å, respectively. The results presented in the current study confirm that all the prepared polymers have an amorphous nature, which are similar to the most conductive polymers from the literature [53,60,61]. It is also worth noting that the peaks for the polymers containing n-octyl chains are more pronounced which indicates more aggregation than those polymers including 3,7-dimethyloctyl chains as also indicated by their lower solubility.

Powder X-ray Diffraction (XRD)
Since a single polymer can come in various forms including crystalline, amorphous and semi-crystalline phases, which effect on its mechanical and electrical properties, it is necessary to accurately determine the structural properties using XRD technique [57][58][59]. The structural properties of the prepared polymers were investigated by powder XRD in solid state as shown in Figure 4. The XRD of the PFDTBTDI-DMO, PFDTBTDI-8, PDBSDTBTDI-DMO, and PDBSDTBTDI-8 show diffraction peaks at 20.0 • , 20.5 • , 19.7 • and 20.3 • corresponding to the π-π stacking distance of 4.43 Å, 4.32 Å, 4.5 Å and 4.36 Å, respectively. The results presented in the current study confirm that all the prepared polymers have an amorphous nature, which are similar to the most conductive polymers from the literature [53,60,61]. It is also worth noting that the peaks for the polymers containing n-octyl chains are more pronounced which indicates more aggregation than those polymers including 3,7-dimethyloctyl chains as also indicated by their lower solubility.

Conclusions
In summary, four fluorene and dibenzosilole-based copolymers were prepared by copolymerizing 2,7-fluorene and 2,7-dibenzosilole (DBS) with both M1 and M2 and yielded PFDTBTDI-DMO, PFDTBTDI-8, PDBSDTBTDI-DMO, and PDBSDTBTDI-8, respectively. All polymers exhibit good solubility in common organic solvents. Changing the alkyl chains on BTDI moieties has a substantial influence on the solubility of the polymers. The use of 3,7-dimethyloctyl side groups on BTDI units in the fluorene-based polymers afforded PFDTBTDI-DMO in high yield; however, the polymer was extracted in the toluene fraction due to its high solubility. The use of n-octyl side groups on BTDI units yielded PFDTBTDI-8, which has a lower solubility. In addition to the toluene fraction, another fraction from chloroform which a higher molecular weight was obtained. However, in the case of dibenzosilole-based copolymers, linear octyl side chains have a negative impact on the molecular weight and the solubility of the resulting polymer. In solutions, all polymers show similar absorption maxima. In thin films, the absorption spectra of the polymers display bathochromic shift absorption maxima relative to their absorption in solutions. The optical band gaps of the fluorene-based copolymers are slightly higher than those of dibenzosilole-based copolymers. The band gaps of the fluorene-based polymers are slightly changed by substituting 3,7-dimethyloctyl chains with n-octyl chains on BTDI units, while the band gaps of dibenzosilole-based polymers are the same. Upon varying fluorene to DBS unit, the HOMO levels of the resulting polymers do not change. This is due to the fact that the HOMO energy levels are controlled by the nature of the donor units, both fluorine and DBS units are weak electron donors of similar strength. All polymers show deep-lying HOMO energy levels of −5.59 eV, which are advantageous for the chemical stability, and this would lead to higher Voc values using these polymers as electron-donating materials in the BHJ devices. All polymers have almost the same LUMO levels since they have the same BTDI acceptor units which dominate the LUMO levels in these materials. Moreover, attaching different alkyl chains on BTDI units has little impact on the LUMO energy levels of the resulting polymers. All polymers show excellent thermal stability with Td exceeding 350 °C. The polymers based on branched 3,7-dimethyloctyl chains have higher thermal stability than those polymers based on n-octyl chains. The thermal stability of the polymers is dependent upon the type of the alkyl substituents attached to the acceptor moieties. The X-ray powder diffraction studies of the polymers show diffraction peaks around 20.0° corresponding to the π-π stacking distance of about 4.0 Å which gives evidence for the amorphous nature of the polymer. All polymers have the amorphous nature.

Conclusions
In summary, four fluorene and dibenzosilole-based copolymers were prepared by copolymerizing 2,7-fluorene and 2,7-dibenzosilole (DBS) with both M1 and M2 and yielded PFDTBTDI-DMO, PFDTBTDI-8, PDBSDTBTDI-DMO, and PDBSDTBTDI-8, respectively. All polymers exhibit good solubility in common organic solvents. Changing the alkyl chains on BTDI moieties has a substantial influence on the solubility of the polymers. The use of 3,7-dimethyloctyl side groups on BTDI units in the fluorene-based polymers afforded PFDTBTDI-DMO in high yield; however, the polymer was extracted in the toluene fraction due to its high solubility. The use of n-octyl side groups on BTDI units yielded PFDTBTDI-8, which has a lower solubility. In addition to the toluene fraction, another fraction from chloroform which a higher molecular weight was obtained. However, in the case of dibenzosilole-based copolymers, linear octyl side chains have a negative impact on the molecular weight and the solubility of the resulting polymer. In solutions, all polymers show similar absorption maxima. In thin films, the absorption spectra of the polymers display bathochromic shift absorption maxima relative to their absorption in solutions. The optical band gaps of the fluorene-based copolymers are slightly higher than those of dibenzosilole-based copolymers. The band gaps of the fluorene-based polymers are slightly changed by substituting 3,7-dimethyloctyl chains with n-octyl chains on BTDI units, while the band gaps of dibenzosilole-based polymers are the same. Upon varying fluorene to DBS unit, the HOMO levels of the resulting polymers do not change. This is due to the fact that the HOMO energy levels are controlled by the nature of the donor units, both fluorine and DBS units are weak electron donors of similar strength. All polymers show deep-lying HOMO energy levels of −5.59 eV, which are advantageous for the chemical stability, and this would lead to higher V oc values using these polymers as electron-donating materials in the BHJ devices. All polymers have almost the same LUMO levels since they have the same BTDI acceptor units which dominate the LUMO levels in these materials. Moreover, attaching different alkyl chains on BTDI units has little impact on the LUMO energy levels of the resulting polymers. All polymers show excellent thermal stability with T d exceeding 350 • C. The polymers based on branched 3,7-dimethyloctyl chains have higher thermal stability than those polymers based on n-octyl chains. The thermal stability of the polymers is dependent upon the type of the alkyl substituents attached to the acceptor moieties. The X-ray powder diffraction studies of the polymers show diffraction peaks around 20.0 • corresponding to the π-π stacking distance of about 4.0 Å which gives evidence for the amorphous nature of the polymer. All polymers have the amorphous nature.