Page 1 of 41 Submitted to Macromolecules 1 2 3 4 A Study of the Degree of Fluorination in Regioregular Poly(3-hexylthiophene) 5 6 7 8 9 10 J. Terence Blaskovits1#, Thomas Bura1#, Serge Beaupré1, Steven A. Lopez2, Carl Roy1, Julio de 11 12 Goes Soares 2, Adam Oh2, Jesse Quinn3, Yuning Li3, Alán Aspuru-Guzik2, Mario Leclerc1* 13 14 15 16 # J.T.B. and T.B. contributed equally 17 18 19 [*] Prof. Mario Leclerc 20 21 22 1) Canada Research Chair on Electroactive and Photoactive Polymers, Department of Chemistry, 23 24 Université Laval, Quebec City, Quebec, Canada 25 26 27 2) Department of Chemistry and Chemical Biology, Harvard University, Cambridge, 28 29 Massachusetts, U.S.A. 30 31 3) Department of Chemical Engineering, University of Waterloo, Waterloo, Ontario, Canada 32 33 34 35 36 E-mail : Mario.Leclerc@chm.ulaval.ca 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1 ACS Paragon Plus Environment Submitted to Macromolecules 1 2 3 4 For Table of Contents Use Only: 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment Page 2 of 41 2 Page 3 of 41 Submitted to Macromolecules 1 2 3 4 Abstract 5 6 7 We systematically varied the degree of fluorination along the backbone of a series of highly 8 9 regioregular 3-hexylthiophene-based polymers, P3HT-50F, P3HT-33F and P3HT-25F, in 10 11 which 50%, 33% and 25% of the thiophene units within the polymer chain contain fluorine 12 13 14 atoms in the available 4-position. These materials were homopolymerized using the Kumada 15 16 catalyst transfer polycondensation method from a set of mono-fluorinated bi-, ter- and 17 18 19 quarterthiophenes, to ensure high polymer regioregularity and evenly-spaced fluorine atoms 20 21 along the conjugated thiophene backbone. The monomers were obtained from a synthetic route 22 23 consisting of iterative Migita-Stille couplings of fluorinated and non-fluorinated 3- 24 25 26 hexylthiophenes. The effect of the fluorine atoms on both polymer structure and properties is 27 28 presented, with supporting quantum mechanical calculations that rationalize the intrinsic 29 30 31 conformation preferences of the three P3HT derivatives. P3HT-50F (‫ܯ‬ഥ௡ = 34 kg/mol, 98.5% rr), 32 33 P3HT-33F (‫ܯ‬ഥ௡ = 46 kg/mol, 98% rr) and P3HT-25F (‫ܯ‬ഥ௡ = 53 kg/mol, 95% rr) displayed 34 35 HOMO levels of -5.34, -5.26 and -5.24 eV, bandgaps of 1.98, 1.98 and 1.97 eV, and average 36 37 38 field-effect transistor hole mobilities of 4.5×10-3, 2.7×10-2, and 1.2×10-2 cm2 V-1s-1, respectively. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3 ACS Paragon Plus Environment Submitted to Macromolecules Page 4 of 41 1 2 3 4 Introduction 5 6 7 The impact of incorporating fluorine atoms into conjugated polymers has been the subject of 8 9 much investigation. The incorporation of fluorine onto conjugated systems favors π-π stacking,1, 10 11 2 and often results in a decrease of both the HOMO and LUMO frontier molecular orbital 12 13 14 energies.3, 4 The bandgap is thus constant, leading to a maintained absorption range of the 15 16 material, while increasing the open-circuit voltage (VOC) in bulk heterojunction solar cells.5 The 17 18 19 electronegativity of fluorine atoms can result in additional non-covalent interactions with 20 21 adjacent repeating units which possess sulfur or hydrogen atoms in close spatial proximity to the 22 23 fluorine moiety. These interactions promote coplanarity within a polymer chain, and strengthen 24 25 26 inter- and intra-chain interactions.6, 7 However, the impact of fluorination on polymer 27 28 morphology remains unpredictable, and both reduced solubility during polymerization and 29 30 aggregation during device fabrication should be considered carefully when working with 31 32 33 fluorinated homo- and copolymers.8 Fluorination at the 3-position of some thiophene-based 34 35 units of conjugated polymers increases both the difficulty and cost of monomer synthesis, 36 37 38 potentially limiting large-scale applications of these materials. 39 40 41 A number of studies have focused specifically on the preparation of fluorinated polythiophenes. 42 43 These are summarized in Figure 1. Following initial reports of the electropolymerization of 3- 44 45 fluorothiophene9 and the synthesis of a perfluorinated oligothiophene10 in the early 2000s, a 46 47 48 patent describing different synthetic routes to obtain poly(3-alkylthiophene) derivatives with 49 50 fluorinated backbones was presented.11 Drawing on these synthetic procedures, Roncali and 51 52 53 colleagues have since described the electropolymerization of an alkylated mono-fluorinated 54 55 terthiophene.12 Swager et al. also post-functionalized a brominated poly(3-hexylthiophene) at the 56 57 4-position using various electrophilic reagents.13 Using this method, the authors were able to 58 59 60 4 ACS Paragon Plus Environment Page 5 of 41 Submitted to Macromolecules 1 2 3 4 incorporate fluorine atoms into 67% of the thiophene units using N-fluorobenzenesulfonimide 5 6 (NFSI). 7 8 9 Two recently reported methods describe the synthesis of fluorinated thiophene monomers which 10 11 were then polymerized using polycondensation techniques. First, the Kumada catalyst transfer 12 13 14 polycondensation (KCTP) of 2,5-dibromo-3-fluoro-4-hexylthiophene resulted in a relatively 15 16 regioregular (93% regioregularity) fluorinated poly(3-hexylthiohene) (P3HT) by Heeney et al.14 17 18 19 However, the resulting polymer displayed very low solubility, and further characterization could 20 21 not be undertaken. The authors then prepared analogue polymers with octyl and ethylhexyl side 22 23 chains to improve solubility and better compare the effect of fluorination. All three fluorinated 24 25 26 poly(3-alkyl)thiophenes displayed improved charge carrier mobilities despite a lower degree of 27 28 crystallinity (as compared to P3HT), which may result from increased coplanarity, as indicated 29 30 by conformational analysis using density functional theory (DFT). Second, Coughlin and co- 31 32 33 workers increased the solubility of fluorinated P3HT via a direct (hetero)arylation (DHAP) 34 35 protocol.15 This was performed by varying the initial ratio of two monomers: 2-bromo-3- 36 37 hexylthiophene and 2-bromo-4-fluoro-3-hexylthiophene. However, with an increase in 38 39 40 fluorinated thiophene content from 25 to 100%, regioregularity dropped systematically from 41 42 89% to 78%. Hole mobility of approximately 2-3 × 10-5 cm2 V-1 s-1 (as determined by the space 43 44 45 charge limited current method) and power conversion efficiency (PCE) of OPV devices also 46 47 decreased with increased fluorination. The phenomenon of reduced PCE with reduced 48 49 regioregularity has been previously observed in non-fluorinated P3HT.16 In the DHAP reaction, 50 51 52 the presence of an electron-withdrawing group adjacent to the desired C-H bond increases 53 54 reactivity,17 which may have an adverse effect on the random nature of such a fluorinated 55 56 copolymer. In addition to this, the authors used DHAP conditions suitable for copolymers but 57 58 59 60 5 ACS Paragon Plus Environment Submitted to Macromolecules Page 6 of 41 1 2 3 4 necessarily optimal for the preparation of poly(3-alkyl)thiophenes.18 These factors rendered an 5 6 assessment of the impact of the random incorporation of fluorinated thiophene units into a P3HT 7 8 chain difficult. 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Figure 1. A summary of the synthesis of various fluorinated oligo- and polythiophenes from 56 57 58 1994 to the present: a) the electropolymerization of 3-fluorothiophene;9 b) the step-wise 59 60 6 ACS Paragon Plus Environment Page 7 of 41 Submitted to Macromolecules 1 2 3 4 synthesis of tetradecafluorosexithiophene;10 c) the electropolymerization of an alkylated mono- 5 6 fluorinated terthiophene;12 d) the post-functionalization of brominated P3HT with the 7 8 electrophilic fluorine source N-fluorobenzenesulfonimide (NFSI);13 e) the KCTP of 2,5-dibromo- 9 10 11 3-fluoro-4-alkylthiophenes;14 f) the random DHAP copolymerization of 2-bromo-4-fluoro-3- 12 13 hexylthophene with 2-bromo-3-hexylthiophene;15 g) this report. 14 15 16 17 18 19 For this reason, we synthesized a series of highly regioregular fluorinated P3HT-based 20 21 22 homopolymers via iterative thiophene couplings. We wished to assess the influence of 23 24 fluorination on the optical and electronic properties of P3HT, especially with regards to the 25 26 27 degree of fluorination required to modulate the HOMO and bandgap of P3HT. This experiment 28 29 also served to determine if the increasing, precisely ordered presence of fluorine atoms along the 30 31 polymer backbone leads to a change in charge-carrier mobility through electronic or steric 32 33 34 influences. Kumada catalyst transfer polycondensation (KCTP) was used to allow more direct 35 36 comparison with previously reported fully-fluorinated P3HT derivatives, which have been also 37 38 prepared using this method.14 39 40 41 42 43 44 45 Results and Discussion 46 47 48 1. Monomer Synthesis 49 50 51 The starting fluorinated unit 2-bromo-4-fluoro-3-hexylthiophene (2Br4F3HT) was synthesized 52 53 54 according to previously reported methods,10-12 albeit with a slight modification to the protocol of 55 56 the fluorination step (see Supporting Information). The bromine-fluorine halogen exchange 57 58 59 60 7 ACS Paragon Plus Environment Submitted to Macromolecules Page 8 of 41 1 2 3 4 was undertaken with both the organolithium reaction mixture and the solution of freshly 5 6 recrystallized N-fluorobenzenesulfonimide (NFSI) at – 100 ºC. This differs from previously 7 8 recorded protocols, in which the reaction mixture was cooled to – 78 ºC and the solution of NFSI 9 10 11 was maintained at room temperature. The NFSI used was also freshly recrystallized in diethyl 12 13 ether, according to a previously reported protocol.10 In this way, less dehalogenation was 14 15 observed at the 3-position of the thiophene, and the isolated yield for this step was consequently 16 17 18 improved from approximately 50 to 65%. 19 20 21 A series of iterative stannylation reactions and Migita-Stille couplings was then undertaken in 22 23 order to obtain compounds 4, 6 and 8 (as shown in Figure 2), in a similar fashion to a previous 24 25 26 report.19 In each case, the 2-position of the thiophene was protected with a trimethylsilyl moiety 27 28 in order to prevent stannylation at this position. An initial synthetic approach using n- 29 30 butyllithium with tetramethylethylenediamine (TMEDA) was attempted in order to selectively 31 32 33 insert the organotin moiety in the 5-position. As noted by Smith and Barratt,20 the use of amine 34 35 ligands or even sterically hindered lithium amide reagents such as LiTMP (lithium 2,2,6,6- 36 37 tetramethylpiperidide) do not entirely eliminate the possibility of a mixture of regioisomers 38 39 40 following electrophilic substitution. Even with high ratios of 5-substituted to 2-substituted 41 42 thiophene derivatives, the presence of irregularly stannylated thiophenes result in some undesired 43 44 45 coupling products, rendering purification procedures more difficult. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 8 ACS Paragon Plus Environment Page 9 of 41 Submitted to Macromolecules 1 2 3 C6H13 4 5S 61 7 8 9i 10 C6H13 C6H13 C6H13 11 ii iii 12 S Br S TMS Me3Sn S TMS 13 23 4 14 15 16 C6H13 C6H13 17 18 iv S S TMS iii Me3Sn S S TMS 19 20 C6H13 5 C6H13 6 21 22 23 24 iv C6H13 C6H13 iii C6H13 C6H13 25 26 S S S TMS S Me3Sn S S TMS 27 C6H13 7 C6H13 8 28 29 30 Figure 2. Iterative synthesis of stannylthiophene derivatives 4, 6 and 8. Reaction conditions: i) 31 32 33 NBS in 1:1 CHCl3:AcOH; ii) n-BuLi in THF, -78 ºC, then Me3SiCl; iii) n-BuLi in THF, -78 ºC, 34 35 then Me3SnCl; iv) Pd2dba3 3 mol%, P(o-tol)3 12 mol%, toluene, reflux. 36 37 38 39 40 41 42 Following the preparation of the stannylated mono-, bi- and terthiophenes, each of these was 43 44 coupled to 2Br4F3HT via Migita-Stille coupling, as presented in Figure 3. The desired products 45 46 9, 11 and 13 were obtained in 88%, 84% and 95% yield, respectively, and in each case the 47 48 49 principal side-product was isolated and identified by 1H NMR analysis as being the result of 50 51 stannylthiophene homocoupling. Once the fluorinated thiophene unit was inserted onto the 52 53 thiophene chain, the monomers were readily purified using reverse-phase (C-18) flash column 54 55 56 chromatography, which turned out to be quite advantageous, as some impurities could not be 57 58 59 60 9 ACS Paragon Plus Environment Submitted to Macromolecules Page 10 of 41 1 2 3 4 separated from the desired product using silica as the stationary phase. This ensured the purity of 5 6 the monomers and, by extension, the purity of the resulting polymers. Interestingly, the final step 7 8 to obtain the di-brominated monomers 10, 12 and 14 required three different bromination 9 10 11 reaction conditions (see Figure 3 for details). The bromination of 9 was performed as previously 12 13 reported in numerous studies, by using N-bromosuccinimide (NBS) in a 1:1 mixture of 14 15 chloroform and acetic acid, affording 10 in a 93% yield. However, these conditions led to 16 17 18 unexpected products when applied to 11 and 13. Following optimization of the bromination 19 20 reaction, we found that the bromination of 11 with NBS afforded the desired product in acid-free 21 22 chloroform. Acid-free chloroform was obtained by filtering the solvent through aluminum oxide 23 24 25 (basic). In this way, 12 was obtained in 52% yield. NBS was found to be an ineffective 26 27 brominating agent in the preparation of 14, and 1,3-dibromo-5,5-dimethylhydantoin was 28 29 30 employed to afford the desired product, albeit in an isolated yield of 51%. All synthetic 31 32 protocols are detailed in the Supporting Information. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 Figure 3. Synthesis of monomers from 2Br4F3HT coupled with stannyl thiophenes 4, 6 and 8; 55 56 57 polymerization of 10, 12 and 14 to afford P3HT-50F, P3HT-33F and P3HT-25F, respectively. 58 59 60 10 ACS Paragon Plus Environment Page 11 of 41 Submitted to Macromolecules 1 2 3 4 Reaction conditions: iv) Pd2dba3 3 mol%, P(o-tol)3 12 mol%, toluene, reflux; v) NBS, 1:1 5 6 CHCl3:AcOH; vi) NBS, CHCl3; vii) 1,3-dibromo-5,5-dimethylhydantoin, CHCl3; viii) 0.97 7 8 mol% i-PrMgCl.LiCl, THF, 0 ºC, then Ni(dppp)Cl2 1.5 mol%, 80 ºC. 9 10 11 12 13 14 15 2. Polymer Synthesis and Structural Characterization 16 17 18 Polymerization of 10, 12 and 14, which afforded respectively P3HT-50F, P3HT-33F and 19 20 P3HT-25F, was performed using similar conditions to those reported previously for P3HT and 21 22 23 fluorinated P3HT.14, 21-23 These consisted of selectively forming the Grignard on the fluorinated 24 25 thiophene using the Turbo Grignard reagent (i-PrMgCl.LiCl) in slightly sub-stoechiometric 26 27 quantities (0.97 equivalent) at 0 ºC in THF, followed by the addition of the precatalyst [1,3- 28 29 30 bis(diphenylphosphino)propane]dichloronickel(II) (Ni(dppp)Cl2) to the reaction mixture and 31 32 heating to 80 ºC. Results for the three polymers are summarized in Table 1, along with those of 33 34 P3HT prepared by both KCTP and DHAP.24 35 36 37 38 Table 1. Molecular weight, molar-mass dispersity, degree of polymerization, yield and 39 40 regioregularity of P3HT-50F, P3HT-33F and P3HT-25F 41 42 43 44 Polymer ࡹഥ ࢔ (kg/mol) a ÐM a DP b Yield (%) c RR (%) d 45 46 P3HT-50F 34 1.5 97 72 98.5 47 48 P3HT-33F 46 1.3 88 78 98 49 50 51 P3HT-25F 53 1.5 78 59 95 52 53 P3HT (KCTP) e 88 1.5 530 - 98 54 55 P3HT (DHAP) e 33 1.8 199 96 > 99 56 57 58 59 60 11 ACS Paragon Plus Environment Submitted to Macromolecules Page 12 of 41 1 2 3 4 a Number average molar mass (‫ܯ‬ഥ௡) and molar-mass dispersity (ÐM) were obtained from size 5 6 exclusion chromatography (SEC) measurements. b Degree of polymerization (DP) was 7 8 calculated from the number average molar mass using the molar weight of the repeating unit. c 9 10 11 Yield was obtained following Soxhlet extraction and polymer washing. d Regioregularity was 12 13 determined from integration of the resonances in 19F NMR spectra. e Previously-reported 14 15 samples of P3HT.24 16 17 18 19 20 21 22 Crude polymers were precipitated in methanol and purified using Soxhlet extraction (acetone, 23 24 hexane and ortho-dichlorobenzene [o-DCB]), affording P3HT-50F, P3HT-33F and P3HT-25F 25 26 27 with degrees of polymerization (DP) of 97, 88 and 78, and isolated yields of 72, 78 and 59%, 28 29 respectively. Despite not being extracted with chloroform or chlorobenzene, molar-mass 30 31 dispersity values remained relatively and comparatively narrow for each polymer. Some 32 33 34 insoluble material remained in the Soxhlet cartridge following the extraction of P3HT-50F with 35 36 o-DCB. The polymer washing step was used in order to reduce metallic impurities, following a 37 38 reported procedure for polymers prepared by the KCTP method.25 P3HT-50F was found to be 39 40 41 soluble in chlorinated solvents (o-DCB and 1,2,4-trichlorobenzene) above 100 ºC, as was P3HT- 42 43 33F around 80 ºC. P3HT-25F was found to be soluble in these solvents at room temperature. 44 45 46 MALDI-ToF analysis was undertaken in order to identify end-chains (full spectra are available 47 48 49 in the Supporting Information). In every case, the signature peaks separated by the molecular 50 51 weight of the repeating unit was well-defined. For all three polymers, the main peak could be 52 53 attributed to Br/Br end-chains and the secondary peak to the expected H/Br end-chains. 54 55 56 Interestingly, the loss of a C5H11 chain radical, which has been previously observed in P3HT26 57 58 59 60 12 ACS Paragon Plus Environment Page 13 of 41 Submitted to Macromolecules 1 2 3 4 and random fluorinated P3HT15 was not identified in any of the spectra. Nor was there any 5 6 evidence of iso-propyl endgroups, potentially originating from brominated end groups reacting 7 8 with unreacted iso-propylmagnesium chloride. Only in P3HT-25F were weak signals associated 9 10 11 with H/H end-chains found, possibly arising from magnesium-halogen exchange when the 12 13 reaction mixture temporarily gelified at 80 ºC (an increase in temperature to 120 ºC was 14 15 necessary to resolubilize the mixture).27 In KCTP, H/Br is the expected endcap pair, due to the 16 17 18 quenching of the active Grignard chain ends with methanol. In this case, we hypothesized that 19 20 the prevalence of Br/Br endcapping may arise from reductive homocoupling of two Grignard 21 22 23 chain ends (leading to the regeneration of the initial catalytic species Ni(dppp)Cl2), or from chain 24 25 termination, in which the catalytic species simply decoordinates from the polymer chain before 26 27 being inserted into a new C-Br bond. This phenomenon becomes more likely as the reaction time 28 29 30 increases.28 As described above, magnesium-halogen exchange could also be a cause of the 31 32 observed end-chains, but as there is very little (P3HT-25F) or no evidence (P3HT-50F and 33 34 P3HT-33F) of the accompanying H/H originating from the quenching of the bis-Grignard 35 36 37 polymer chain, this is unlikely. Due to the fact the molar masses observed in the MALDI-ToF 38 39 spectra are considerably lower than those obtained from SEC and NMR measurements (see 40 41 below), it can be presumed that only the shorter chains within the polymer sample are readily 42 43 44 ionised when this method of characterization is used. This may lead to the disproportionately 45 46 large fraction of Br/Br end capped chains observed, as these were terminated early on in the 47 48 polymerization reaction. It is therefore not possible to ascertain the nature of the end groups of 49 50 51 longer polymer chains. 52 53 54 Polymer regioregularity and end-chains were studied by 1H and 19F NMR analysis in 55 56 tetrachloroethane (TCE) at 110 ºC, due to the limited solubility of P3HT-50F and P3HT-33F at 57 58 59 60 13 ACS Paragon Plus Environment Submitted to Macromolecules Page 14 of 41 1 2 3 4 lower temperatures. The full 1H NMR spectra as well as details of the aromatic and methylene 5 6 regions are shown in Figure 4. The signal at 2.83 ppm corresponds to the methylene of the 7 8 fluorinated thiophene unit, and the signal at 2.91 ppm to that of the non-fluorinated thiophene, 9 10 11 although peak multiplicity was lost due to the heating necessary to fully solubilize P3HT-50F 12 13 and P3HT-33F. The integration of these signals against one another serves as a tool to validate 14 15 the polymer structure: the resonance at both 2.83 and 2.91 ppm have the same integration value 16 17 18 in P3HT-50F, the resonance at 2.91 ppm is twice as intense as the resonance at 2.83 ppm P3HT- 19 20 33F, and three time as intense in P3HT-25F. This effect can be readily correlated with the 21 22 equivalent signals of the corresponding monomers: the well-defined triplet at ca. 2.57 ppm 23 24 25 corresponds to the methylene of fluorinated thiophene, and the multiplet centered on ca. 2.7 ppm 26 27 to those of all the other, non-fluorinated thiophenes. Coughlin and colleagues also observed this 28 29 30 difference in methylene signals in their study of random fluorinated P3HT copolymers.15 31 32 33 In P3HT, the chemical shift of the central aromatic proton in the HT-HT triad is 6.98 ppm. In 34 35 P3HT-50F, the principal aromatic resonance (H(A)) is found at 7.14 ppm, indicating that the 50% 36 37 fluorinated equivalent of P3HT in the spectral conditions used (110 ºC) is downshifted by 38 39 40 approximately 0.16 ppm vis-à-vis P3HT. The combination of the electronic effect of the fluorine 41 42 atoms along the polymer backbone and the use of a different solvent (TCE) at a high temperature 43 44 45 modifies the absolute chemical shift of the protons, but should not change the relative shift 46 47 between protons. For this reason, it is possible to associate the secondary resonance visible in the 48 49 P3HT-50F spectrum at 6.96 ppm to a resonance which might appear at ca. 6.80 ppm in P3HT. 50 51 52 This corresponds most closely to the proton on a HT-HT brominated terminal thiophene (H(B), 53 54 which is situated at 6.82 ppm in non-fluorinated P3HT), confirming the results of MALDI-ToF 55 56 analysis. It appears in 3% of the intensity of the principal HT-HT signal, meaning that if, as the 57 58 59 60 14 ACS Paragon Plus Environment Page 15 of 41 Submitted to Macromolecules 1 2 3 4 MALDI-ToF results indicate, the vast majority of end-chains are Br/Br, this corresponds to a 5 6 degree of polymerization (DP) of 67. The fact that the DP as calculated using NMR spectra is 7 8 greatly inferior to the DP as determined by SEC measurements (97) is widely documented in the 9 10 11 characterization of conjugated polymers, including poly(thiophene)s.29 This is due to the 12 13 discrepancy of the hydrodynamic radius of such polymers as compared to the radii of the 14 15 poly(styrene) standards used for the calibration of SEC methods. 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 15 ACS Paragon Plus Environment Submitted to Macromolecules Page 16 of 41 1 2 3 4 5 6 7 8 9 10 11 H(D) 12 13 14 H(A) 15 H(C) 16 17 18 19 20 21 22 23 24 25 26 27 H(B) H(E) 28 29 30 31 32 33 34 35 36 Figure 4. Full 1H NMR spectra (above) and detail of the aromatic and methylene regions for 37 38 39 P3HT-50F (green), P3HT-33F (blue) and P3HT-25F (red) in TCE at 110 ºC. The resonances at 40 41 ca. 7.12 ppm correspond to the protons on the non-fluorinated thiophenes (H(A)). The resonances 42 43 44 at ca. 2.83 ppm correspond to the first CH2 (methylene) of the hexyl chains of the fluorinated 45 46 thiophene units (H(C)), and those at ca. 2.91 ppm to the methylene on non-fluorinated thiophenes 47 48 (H(D)). H(B) and H(E) correspond to the aromatic and methylene protons on the end-chain units, 49 50 51 respectively. Integration of P3HT-50F alone has been shown for clarity in the lower image. 52 53 54 55 56 57 58 59 60 16 ACS Paragon Plus Environment Page 17 of 41 Submitted to Macromolecules 1 2 3 4 In addition to this, HT-HH defects in P3HT appear as a signal with a shift of 7.02 ppm. This 5 6 would correspond to a resonance at 7.18 in P3HT-50F. There is no evidence of a resonance with 7 8 this chemical shift, consistent with the fact that head-to-head homocoupling defects do not occur 9 10 11 in P3HT prepared by the KCTP method.30 There is no other signal visible in the aromatic region 12 13 of the 1H NMR spectra, confirming both the polymer regioregularity and that initial TT coupling 14 15 is the only observable defect within each polymer chain. 16 17 18 19 In the methylene region, a single secondary resonance appears at 2.69 ppm, upshifted by 0.20 20 21 ppm from the abovementioned principal methylene resonance of the non-fluorinated thiophene 22 23 (H(D), 2.89 ppm). The main explanation for this is an HT dyad end group. This is supported by 24 25 26 the 1H NMR spectrum of a sample of regioregular P3HT in the same analysis conditions used for 27 28 P3HT-50F: although all resonance resolution is lost in these conditions, the secondary 29 30 methylene resonance corresponding to the end-chains is also found at ca. 2.69 ppm, upshifted 31 32 33 from the principal resonance by 0.21 ppm (see Supporting Information for the P3HT 34 35 spectrum). As the principal methylene resonances (H(C) and H(D)) occur in a 4:1 ratio to the 36 37 38 principal aromatic resonance of P3HT-50F (H(A)), so should the brominated chain-end 39 40 methylene resonance appear in a 4:1 ratio to the aromatic proton on the brominated thiophene 41 42 end group. This turns out to be the case, in that the secondary resonance at 2.69 ppm corresponds 43 44 45 to 4% of the intensity of the two principal methylene resonances. This same reasoning was 46 47 applied to the spectra of P3HT-33F and P3HT-25F, with similar results. The secondary 48 49 resonances in the aromatic and methylene regions integrated for 3% of the principal resonances 50 51 52 in both P3HT-33F and P3HT-25F. 53 54 55 In P3HT, the TT-HT triad produced by the initiation of KCTP chain growth can be observed in 56 57 the aromatic region of 1H NMR spectra. However, due to the presence of the fluorine atoms on 58 59 60 17 ACS Paragon Plus Environment Submitted to Macromolecules Page 18 of 41 1 2 3 4 the thiophene backbone in fluorinated P3HT derivatives, this is not possible. We therefore turned 5 6 to interpreting the 19F NMR spectra of the three polymers for insight into the TT defect. 7 8 Analyses reveal an upfield shift of the principal resonance (F(A)) going from δ = -120.97 ppm in 9 10 11 P3HT-50F to δ = -122.43 ppm in P3HT-33F, and another slight shift to δ = -122.50 ppm for 12 13 P3HT-25F. In all of these spectra, a secondary signal upfield from the principal resonance is 14 15 visible, and as the rate of fluorination decreases, this resonance draws nearer to the principal 16 17 18 resonance. In P3HT-50F, this resonance has 3% of the intensity of the principal resonance, and 19 20 in P3HT-33F 4%. In P3HT-25F the primary and secondary signals are differentiated from one 21 22 another with difficulty, but the secondary resonance shows an intensity of approximately 10% 23 24 25 relative to the principal resonance (see Supporting Information for full 1H and 19F NMR 26 27 spectra). 28 29 30 Due to the symmetrical nature of the inherent TT defect, the presence of fluorine atoms in each 31 32 33 of the repeating units amplifies the defect twofold. If there is only one TT defect within the 34 35 polymer chain, the 3% intensity signal in P3HT-50F would therefore represent a TT defect (F(B)) 36 37 in approximately 1.5% of the polymer chain, hence a 98.5% regioregularity. This same reasoning 38 39 40 leads to a regioregularity of 98% for P3HT-33F and 95% for P3HT-25F. 41 42 43 In the 19F NMR spectrum of P3HT-50F, a very weak signal can be observed at -120.27 ppm. 44 45 This may be produced by the small rate of TT initiation-induced homocoupling defects which, 46 47 48 rather than occurring somewhere within the polymer chain, appear at the beginning of the chain, 49 50 as was originally assumed to be the only position of the defect in KCTP-prepared polymers. One 51 52 53 possible piece of evidence that this may be the case is that this resonance does not appear in the 54 55 two other polymers. This may be due to the nature of the monomers used in that, as the rate of 56 57 fluorination decreases, the fluorine atom is increasingly distant from the chain end, i.e. one 58 59 60 18 ACS Paragon Plus Environment Page 19 of 41 Submitted to Macromolecules 1 2 3 4 thiophene separates it from the terminal C-Br bond in P3HT-50F, two in P3HT-33F and three in 5 6 P3HT-25F. Therefore, the occurrence of a TT homocoupling defect present at a chain end for 7 8 P3HT-33F and P3HT-25F would be too distant from the nearest fluorine atom for the electronic 9 10 11 environment of P3HT-25F to be significantly different than a TT defect within the chain. 12 13 14 The values of regioregularity obtained using this method offer more of an indication as to the 15 16 degree of polymerization for each chain, rather than the rate of defects. Since no other clear 17 18 19 signals indicating other defects are present, one may conclude that there is only one TT defect 20 21 per polymer chain. This TT defect is then ‘diluted’ in the number of regioregular HT couplings 22 23 corresponding to the degree of polymerization of the chain. Despite the imprecision of NMR 24 25 26 analysis in giving precise ratios for high molecular weight polymers, the values of 27 28 regioregularity are in relatively good agreement with those of the soluble 100% fluorinated 29 30 poly(3-alkylthiophene)s synthesized by Heeney et al. (in which the hexyl side chain was 31 32 33 replaced by octyl and ethylhexyl chains).14 34 35 36 In P3HT, DSC measurements serve as an indicator of the degree of regioregularity. A study in 37 38 which defects were deliberately incorporated into the polymer chain showed a strong correlation 39 40 41 between the regioregularity and the crystallization temperature and enthalpy.31 Heating and 42 43 cooling traces are shown in Figure 5, and Table 2 details the numerical values from these 44 45 experiments, along with those of P3HT. All three samples display relatively sharp crystallization 46 47 48 and melting peaks, indicating a high regioregularity for P3HT-50F, P3HT-33F and P3HT-25F. 49 50 No evidence of glass transitions was found. With decreasing fluorination, the melting 51 52 53 temperatures were 264, 254 and 250 ºC, crystallization temperatures were 239, 228 and 221 ºC, 54 55 and crystallization enthalpies were 32, 27 and 25 J/g. The lower melting temperature and 56 57 crystallization enthalpies for the polymers with a lower degree of fluorination may be due to 58 59 60 19 ACS Paragon Plus Environment Submitted to Macromolecules Page 20 of 41 1 2 3 4 reduced interactions between polymer chains. Melting and crystallization temperatures for 5 6 P3HT-50F were only slightly lower than those reported for the 100% fluorinated P3HT prepared 7 8 by the same KCTP method (Tfus,p = 267 ºC and 279 ºC; Tc,p = 245 ºC), and unlike the 100% 9 10 11 fluorinated poly(3-alkylthiophene)s previously reported, P3HT-50F possessed a single melting 12 13 peak, rather than two.14 Interestingly, the crystallization enthalpy of P3HT-50F was equivalent 14 15 to that of the all-fluorinated P3HT. All three polymers also display high thermal stability, with 16 17 18 nearly identical degradation temperatures. 19 20 21 22 23 24 Table 2. Physical properties of P3HT-50F, P3HT-33F, P3HT-25F, and P3HT 25 26 27 28 Polymer Tfus,p (ºC) a Tc,p (ºC) a ∆H(Tc) (J/g) a Td (ºC) b 29 30 P3HT-50F 264 239 32 436 31 32 P3HT-33F 254 228 27 438 33 34 35 P3HT-25F 250 221 25 440 36 37 P3HT (KCTP) c 234 198 15 464 38 39 P3HT (DHAP) c 237 208 16 464 40 41 42 43 44 45 a Melting temperature (Tfus,p), crystallization temperature (Tc,p) and enthalpy of crystallization 46 47 48 (∆H(Tc)) were determined by differential scanning calorimetry (DSC) under nitrogen at a 49 50 scanning speed of 10 ºC/min. b Degradation temperature (Td) was determined at 5% weight loss 51 52 via thermogravimetric analysis (TGA). c Previously reported samples of P3HT.24 53 54 55 56 57 58 59 60 20 ACS Paragon Plus Environment Page 21 of 41 Submitted to Macromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Figure 5. DSC traces for the second cycle of heating and cooling for P3HT-50F, P3HT-33F, 28 29 P3HT-25F and P3HT under nitrogen at a scanning speed of 10 ºC/min. Endothermic transitions 30 31 32 point downward. 33 34 35 36 37 38 39 40 41 3. Conformational analysis 42 43 44 45 We employed density functional theory (DFT) calculations in order to understand the intrinsic 46 47 conformational preferences of P3HT-50, P3HT-33 and P3HT-25. To reduce the computational 48 49 expense and to simplify the potential energy surface, we truncated all alkyl sidechains to methyl 50 51 52 groups with the common assumption that this modification would not substantially change their 53 54 electronic properties. Our calculations were performed on model oligomers (monomers, dimers, 55 56 and trimers) of the polymers studied experimentally (Figure 6). 57 58 59 60 21 ACS Paragon Plus Environment Submitted to Macromolecules Page 22 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Figure 6. The series of truncated fluorinated P3HT oligomers studied using DFT calculations. 21 22 23 We employed the B3LYP/6-311+G(d,p)-D3BJ)-IEFPCM(ε=4)//B97D/6-31G(d,p)-IEFPCM(ε=4) 24 25 level of theory to find the two lowest energy conformers based on this model chemistry for 26 27 28 P3HT-50F-monomer. The structures and relative energies are shown in Figure 7. 29 30 31 32 33 34 35 36 36˚ 28˚ 37 38 39 40 syn anti 41 ∆G = 0.6 kcal mol-1 ∆G = 0.0 kcal mol-1 42 43 44 45 Figure 7. The lowest energy conformers of P3HT-50F-monomer. Computed using B3LYP/6- 46 47 311+G(d,p)-D3BJ)-IEFPCM(ε=4)//B97D/6-31G(d,p)-IEFPCM(ε=4). 48 49 50 For P3HT-50F-monomer, the lowest energy conformer features the thiophenes pointing in 51 52 opposite directions (anti). We refer to the dihedral angle between the two thiophenes as SCCS; 53 54 55 the dihedral angle includes the sulfur atoms and connecting carbons between adjacent 56 57 thiophenes. This conformation has a relatively distorted backbone; the SCCS dihedral angle is 58 59 60 22 ACS Paragon Plus Environment Page 23 of 41 Submitted to Macromolecules 1 2 3 4 28°. The structure is not planar in the gas phase because of competing S-S closed-shell 5 6 repulsions and a steric clash between methyl hydrogens and the nearby sulfur lone pair orbitals 7 8 (2.90 Å). The syn conformer of P3HT-50F-monomer is higher in free energy by 0.6 kcal mol–1 9 10 11 and the backbone SCCS dihedral angle is 36°. The destabilizing S-S repulsion outweighs the 12 13 CH3-S steric clash, and we expect this class of fluorinated thiophenes to exhibit more 14 15 16 polymorphism and disorder than thiophenes unsubstituted at the 3-position. We extended our 17 18 conformational search with P3HT-50F-dimer; Figure 8 shows the lowest energy conformer. 19 20 21 22 23 24 25 26 27 22˚ LUMO= –2.26 eV 28 1˚ 29 14˚ 30 31 32 33 34 35 HOMO= –5.14eV 36 37 38 39 Figure 8. The lowest energy conformation of P3HT-50F-dimer. Computed using B3LYP/6- 40 41 311+G(d,p)-D3BJ)-IEFPCM(ε=4)//B97D/6-31G(d,p)-IEFPCM(ε=4). 42 43 44 The lowest energy conformer features an all-anti relationship for the thiophene sulfurs. The 45 46 47 central dihedral angle is virtually planar (1° rotated from planarity). De la Cruz et al. have 48 49 suggested that S-F interactions are slightly stabilizing,32 which appears to preserve planarity 50 51 52 about the central SCCS dihedral in P3HT-50-dimer. The outermost thiophenes are more 53 54 distorted from planarity (14° and 22°) because of CH3-S steric clashes identified in Figure 8. We 55 56 then computed P3HT-50-trimer, but only considered the all-syn and all-anti conformers. The 57 58 59 60 23 ACS Paragon Plus Environment Submitted to Macromolecules Page 24 of 41 1 2 3 4 all-syn conformer is higher in free energy by 1.2 kcal mol–1 and is helical in the gas phase to 5 6 avoid three pairs of S-S repulsions. Only the all-anti conformation of P3HT-50-trimer is 7 8 considered, and is shown in Figure 9. 9 10 11 12 13 14 15 24˚ 3˚ 12˚ 2˚ 16˚ 16 17 18 19 20 21 22 Figure 9. The all-anti conformation of P3HT-50-trimer, computed using B97D/6-31G(d,p)- 23 24 IEFPCM(ε=4). 25 26 27 28 29 30 Wherever S-F interactions are possible, the corresponding SCCS dihedral angles are 31 32 33 considerably more planar (2-3° vs 12-24°). Again, SCCS dihedral angles for the outermost 34 35 36 thiophenes are larger (16° and 24°) to avoid CH3-S steric repulsions. We report the computed 37 38 HOMO and LUMO energies for these systems in Table 3. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 24 ACS Paragon Plus Environment Page 25 of 41 Submitted to Macromolecules 1 2 3 4 Table 3. The computed frontier molecular orbitals for 1 in the all-anti conformations. 5 6 7 Truncated oligomer HOMO LUMO HOMO-LUMO 8 9 model (eV) (eV) gap (eV) 10 11 12 P3HT-50-monomer –5.69 –1.34 4.35 13 14 P3HT-50-dimer –5.14 –2.26 2.88 15 16 P3HT-50-trimer –4.99 –2.51 2.48 17 18 19 P3HT-33-dimer –4.91 –2.42 2.49 20 21 P3HT-25-dimer –4.82 –2.53 2.29 22 23 24 25 26 27 Table 3 shows that the HOMO-LUMO gap is extremely large for the monomer (4.35 eV) 28 29 because of the short conjugation length in this oversimplified system. The dimer and trimer 30 31 32 approach those measured experimentally (Table 4, see below). Our results are consistent with 33 34 many other reports of decreasing HOMO and LUMO energies with increased proportions of 35 36 fluorinated backbone. The HOMO-LUMO gap is larger for P3HT-50-dimer than P3HT-33- 37 38 39 dimer and P3HT-25-dimer because fluorination lowers the energy of the P3HT-50F HOMO 40 41 more than that of the LUMO. This behavior is due to the fluorine lone pair being part of the 42 43 HOMO, but not the LUMO. 44 45 46 47 For P3HT-33 and P3HT-25, we computed the truncated structures P3HT-33-dimer and P3HT- 48 49 25-dimer in the all-anti conformation and calculated the corresponding frontier molecular 50 51 orbitals (Figure 10). 52 53 54 55 56 57 58 59 60 25 ACS Paragon Plus Environment Submitted to Macromolecules Page 26 of 41 1 2 3 4 5 6 20° 16° 1° 9° 18° 7 15° 17° 8° 1° 14° 18° 18° 8 9 10 11 12 13 14 LUMO: -2.42 eV LUMO= –2.53 eV 15 16 17 18 19 20 21 HOMO: -4.91 eV HOMO = –4.82 eV 22 23 24 Figure 10. The structure of P3HT-33-dimer and P3HT-25-dimer with the corresponding 25 26 27 frontier molecular orbitals. Computed using B3LYP/6-311+G(d,p)-D3BJ)- 28 29 IEFPCM(ε=4)//B97D/6-31G(d,p)-IEFPCM(ε=4). 30 31 32 33 34 35 36 In P3HT-33-dimer, the SCCS dihedral angles range from 1-20° and in P3HT-25-dimer the 37 38 dihedral angles range from 1-18°, with the central SCCS dihedral angle virtually planar (1° out 39 40 41 of plane twist). Through-space S-F interactions reinforce the planarity and CH3-S steric clashes 42 43 increase the SCCS angles in these dimers. The extensive delocalization of the frontier molecular 44 45 orbitals of P3HT-25-dimer relative to P3HT-33-dimer results in a smaller HOMO-LUMO gap 46 47 48 (2.29 vs. 2.49 eV). 49 50 51 52 53 54 55 56 57 58 59 60 26 ACS Paragon Plus Environment Page 27 of 41 Submitted to Macromolecules 1 2 3 4 4. Polymer Optical and Electronic Properties 5 6 7 Solution UV-visible analyses were undertaken using a temperature gradient in order to assess the 8 9 aggregation of the polymers in solution (solution and solid-state UV-visible spectra are shown in 10 11 Figure 11, and numerical values from solution and film UV-visible spectroscopic analyses are 12 13 14 detailed in Table 4). The spectra of fluorinated polymers are highly temperature-dependent, as 15 16 previously reported.33 Results indicate that P3HT-25F is soluble in o-DCB at room temperature 17 18 19 and that P3HT-33F becomes fully solubilized at ca. 70 ºC. In P3HT-50F, the shape of the 20 21 absorbance spectrum indicates the presence of aggregates, which persist until 100 ºC. Above 22 23 these temperatures, all three polymers produce similar, featureless spectra with absorbance 24 25 26 maxima at 422, 430 and 432 nm for P3HT-50F, P3HT-33F and P3HT-25F, respectively. The 27 28 temperature-dependant evolution of the spectra was found to be reversible. 29 30 31 Following the observation of these aggregates at relatively high temperatures, films for solid- 32 33 34 state absorbance spectra were spin-coated from o-DCB solutions at 120 ºC in order to ensure the 35 36 full solubility of all three polymers prior to deposition. The three polymers display almost 37 38 identical absorbance spectra in the solid state, indicating similar inter- and intra-chain 39 40 41 organization upon cooling. The two principal absorbance maxima appear at 548/596, 550/595 42 43 and 553/600 nm for P3HT-50F, P3HT-33F and P3HT-25F, respectively, which can be 44 45 46 attributed to the A0-0 and A0-1 spectral transitions.34 The slight variation in absorbance is not due 47 48 to variation in molecular weights, as all three polymers have surpassed the effective conjugation 49 50 length of poly(thiophene)s at the observed chain length.35 With decreased fluorination, both 51 52 53 spectral transitions shift towards longer wavelengths and approach the standard values of P3HT 54 55 in the solid state (555/610 nm).36 56 57 58 59 60 27 ACS Paragon Plus Environment Submitted to Macromolecules Page 28 of 41 1 2 3 4 The ratio of the intensity of these two absorbance maxima (A0-0/A0-1) is comparable for the three 5 6 polymers. In P3HT, a high A0-0/A0-1 ratio is indicative of both a high regioregularity and high 7 8 degree of polymerization. A substantial increase of DP would be necessary to compensate for the 9 10 11 impact of a small reduction in regioregularity on the A0-0/A0-1 ratio, due to the effective 12 13 conjugation length in the crystallized polymer.34 Consequently, the similar A0-0/A0-1 ratios of the 14 15 three polymers (0.965, 0.920 and 0.956, respectively) fits with the slight reduction in 16 17 18 regioregularity from P3HT-50F to P3HT-25F (98%, 97% and 95%) at the same time as a 19 20 relatively large increase of molecular weight values. On the other hand, as observed from NMR 21 22 spectra, the only noticeable defect is the tail-to-tail homocoupling issuing from the initiation of 23 24 25 the KCTP reaction, as discussed earlier. Unlike a head-to-head defect, a tail-to-tail defect in the 26 27 polymer would likely not induce twisting of the conjugated backbone. It is possible that the 28 29 30 presence of fluorine substituents in adjacent thiophene units may in fact favor localized 31 32 coplanarity and rigidity. Therefore, the regioregularity may have a negligible effect on the 33 34 observed A0-0/A0-1 values. 35 36 37 The optical bandgap, as calculated using the solid-state UV-visible spectra, changes very little 38 39 40 between the three polymers, indicating that both the HOMO and LUMO levels deepen in parallel 41 42 with an increased rate of fluorination (numerical values of electronic measurements are given in 43 44 45 Table 4). However, the observed bandgaps are slightly larger than reported values for P3HT (1.9 46 47 eV).37 The bandgap is very similar to the previously reported randomly fluorinated P3HT 48 49 analogues, although the bandgap increases from 1.93 in the 25% fluorinated polymer to 1.99 eV 50 51 52 in the 100% fluorinated polymer.15 The similar bandgaps for P3HT-50F, P3HT-33F and P3HT- 53 54 25F may be indicative of the comparable regioregularity of the three polymers. 55 56 57 58 59 60 28 ACS Paragon Plus Environment Page 29 of 41 Submitted to Macromolecules 1 2 3 4 Table 4. Optical and electronic measurements for P3HT-50F, P3HT-33F and P3HT-25F 5 6 7 Polymer A0-0 Film a Solution Solution Eg opt HOMO LUMO 8 9 /A0-1 (RT in o- (140 ºC in (eV) c (eV) d (eV) e 10 11 12 DCB) b o-DCB) b 13 14 P3HT-50F 0.965 548, 596 548, 594 422 1.98 -5.34 -3.36 15 16 P3HT-33F 0.920 550, 595 460, 594 430 1.98 -5.26 -3.28 17 18 19 P3HT-25F 0.956 553, 600 462 432 1.97 -5.24 -3.27 20 21 22 23 24 a Solid-state UV-visible spectra were obtained from as-spun thin films prepared via spin coating 25 26 27 of a 2.0 mg/ml solution of the polymer in o-DCB at 120 ºC. b Solution UV-visible spectra were 28 29 obtained from a 0.01 mg/ml solution of the polymer heated until fully soluble in o-DCB and then 30 31 cooled to the desired temperature for analysis. c Optical bandgap for each polymer was 32 33 34 determined using the tangent of the onset of low-energy absorbance in the corresponding UV- 35 36 visible spectrum. d The onset of oxidation was determined by cyclic voltammetry of polymer 37 38 39 films cast on the electrode from an o-DCB solution; an external reference of ferrocene (Fc/Fc+) 40 41 was used. e The LUMO energy was estimated by subtracting the optical bandgap (Eg opt) from the 42 43 observed HOMO. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 29 ACS Paragon Plus Environment Submitted to Macromolecules Page 30 of 41 1 2 3 4 5 6 7 (a) (b) 8 9 10 11 12 13 14 15 16 17 18 19 20 21 (c) (d) 22 23 24 25 26 27 28 29 30 31 32 33 34 Figure 11. UV-visible spectra of polymers P3HT-50F, P3HT-33F and P3HT-25F in solution 35 36 (a, b and c) and in thin films (d). Solid-state UV-visible spectra were obtained from as-spun thin 37 38 39 films prepared via spin coating of a 2.0 mg/ml solution of the polymer in o-DCB at 120 ºC. 40 41 Solution UV-visible spectra were obtained from a 0.01 mg/ml solution of the polymer heated 42 43 44 until fully soluble in o-DCB and then cooled to 30ºC, at which point the temperature was 45 46 increased in 10º increments. 47 48 49 50 51 52 Cyclic voltammetry was used to compare the onset of oxidation of the three polymers with that 53 54 55 of P3HT. Cyclic voltammograms for P3HT-50F, P3HT-33F and P3HT-25F, as well as for 56 57 P3HT are shown in Figure 12, along with that of the external ferrocene reference (numerical 58 59 60 30 ACS Paragon Plus Environment Page 31 of 41 Submitted to Macromolecules 1 2 3 4 values are found in Table 4). In all three voltammograms, three oxidative doping peaks are 5 6 clearly visible. Due to this similarity with the voltammograms of P3HT, it is possible that these 7 8 peaks also correspond to the formation of a polaron and bipolaron. The reductive peaks are also 9 10 11 somewhat visible in all three polymers. P3HT-50F shows the most narrow oxidation peaks, 12 13 perhaps due to the narrow spacing of the fluorinated thiophenes which may lead to the most 14 15 isoenergetic polymer chain conjugation.38 The three oxidation peaks in P3HT-25F appear at a 16 17 18 higher voltage than those of the two other fluorinated polymers and, in keeping with this, the 19 20 onset of oxidation of P3HT-50F, P3HT-33F and P3HT-25F is -5.34, -5.26 and -5.24 eV, 21 22 respectively, as compared to –5.18 eV for the P3HT reference polymer analyzed in the same 23 24 25 conditions. It may be concluded that below 33%, fluorination has very little effect on the 26 27 electrochemical properties of P3HT, as the oxidation onset between P3HT-33F and P3HT-25F 28 29 30 changes very little. A parallel can be drawn here with the study of DHAP-prepared random 31 32 fluorinated copolymers, the authors of which identified that the HOMO changes very little 33 34 between the 50%, 75% and 100% fluorinated materials, possibly due to ‘donor’ and ‘acceptor’ 35 36 37 effects.15 The electronic effect of the fluorine substituent on the HOMO does therefore seem to 38 39 have the greatest impact somewhere between 33% and 50% substitution, and beyond this point 40 41 increased fluorination only reduces solubility. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 31 ACS Paragon Plus Environment Submitted to Macromolecules Page 32 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 12. Cyclic voltammetry measurements for P3HT-50F, P3HT-33F and P3HT-25F, along 31 32 33 with a P3HT sample and the external ferrocene (Fc/Fc+) reference. Polymer films were cast onto 34 35 the electrode from an o-DCB solution. 36 37 38 39 40 41 For comparison purposes, field effect transistors were prepared in the bottom-gate bottom- 42 43 44 contact configuration (BGBC) to assess charge carrier mobilities in the three polymers. Due to 45 46 the limited solubility of P3HT-50F, all polymers were deposited via spin-coating of an o-DCB 47 48 49 solution. Full details of transistor preparation are available in the Supporting Information. 50 51 Average and maximum charge carrier mobilities, on/off current ratios and threshold voltages for 52 53 each polymer at these two temperatures are detailed in Table 5, and are accompanied by 54 55 56 reference values for P3HT. The hole mobility of P3HT-50F (0.0079 cm2 V-1 s-1) and P3HT-25F 57 58 59 60 32 ACS Paragon Plus Environment Page 33 of 41 Submitted to Macromolecules 1 2 3 4 (0.0013 cm2 V-1 s-1) were found to be optimal at 100 oC and the highest average was obtained 5 6 with P3HT-33F at 150 oC (0.0289 cm2 V-1 s-1), although mobility values were comparable at 7 8 these two annealing temperatures. The fact that P3HT-33F led to the best results was possibly 9 10 11 due to an optimized trade-off between the increasing planarity of the polymer backbone and 12 13 decreasing crystallinity when the degree of fluorination is increased, as observed in previous 14 15 fluorinated polythiophenes.14, 15 In similarly-prepared devices, highly regioregular non- 16 17 18 fluorinated P3HT prepared by both KCTP and DHAP displays much higher mobilities (0.08 and 19 20 0.14 cm2 V-1 s-1, respectively).24 It is, however, interesting to note that on/off current ratios for all 21 22 three polymers surpass not only those of non-fluorinated P3HT which was prepared in the same 23 24 25 conditions (103), but also those of previously reported all-fluorinated P3HT (104-105).14, 24 The 26 27 most likely explanation may be related to the deeper HOMO levels (~ -5.3 eV as opposed to ~ - 28 29 30 5.2 for P3HT). Indeed, the three fluorinated polymers are less susceptible to doping by ambient 31 32 oxygen than P3HT, preserving their insulating properties and, by extension, limiting their 33 34 electrical conductivity and minimizing their off current. No electron transport behavior was 35 36 37 observed for any of the three polymers. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 33 ACS Paragon Plus Environment Submitted to Macromolecules Page 34 of 41 1 2 3 4 5 6 7 Table 5. Average and maximum mobility in field-effect transistors. 8 9 10 Polymer T (oC) a µ (× 10-2 cm2 V-1 s-1) b ION/OFF c VTH d 11 12 100 0.45 ± 0.20 (0.79) ~105-106 6.9 13 14 15 P3HT-50F 150 0.43 ± 0.20 (0.75) ~106 7.1 16 17 100 2.66 ± 0.13 (2.82) ~106 10.5 18 P3HT-33F 19 20 150 2.72 ± 0.16 (2.89) ~106-107 7.9 21 22 100 1.18 ± 0.07 (1.30) ~106-107 7.7 23 P3HT-25F 24 150 1.05 ± 0.07 (1.17) ~105-106 1.1 25 26 27 28 29 30 a Annealing temperature. b Charge carrier mobility values are written out in the form Average ± 31 32 33 standard deviation (maximum). c On/off current ratio. d Average threshold voltage. 34 35 36 37 38 39 Conclusion 40 41 42 Three derivatives of poly(3-hexylthiophene) which differ in the ratio of fluorinated thiophene 43 44 45 repeating units were reported. P3HT-50F, P3HT-33F and P3HT-25F were synthesized via the 46 47 Kumada catalyst transfer polymerization of monomers which were prepared in turn using 48 49 iterative Migita-Stille couplings from 2-bromo-4-fluoro-3-hexylthiophene (2Br4F3HT). This 50 51 52 method allowed for a regular separation of fluorine atoms along the polythiophenes backbone. 53 54 Detailed 1H and 19F NMR spectroscopic studies was undertaken, and from these it could be 55 56 determined that all three polymers displayed comparable and high regioregularity. Well-defined 57 58 59 60 34 ACS Paragon Plus Environment Page 35 of 41 Submitted to Macromolecules 1 2 3 4 UV-visible absorbance spectra and sharp peaks in differential scanning calorimetry 5 6 measurements both confirmed the high regioregularity of the three polymers. Computational 7 8 analyses indicated that favourable fluorine-sulfur interactions contribute to the planarity of the 9 10 11 anti-conformation between thiophene units of truncated models. The dihedral angle between two 12 13 thiophenes is minimized when one of the thiophenes possesses a fluorine substituent in the 4- 14 15 position. It was determined from this that the favourable sulfur-fluorine interaction overcomes 16 17 18 the otherwise unfavorable sulfur-methyl steric clash. Optical bandgaps of 1.98 eV for both 19 20 P3HT-50F and P3HT-33F, as opposed to 1.97 eV for P3HT-25F, and HOMO energy levels of - 21 22 5.34, -5.26 and -5.24 eV, as determined from electrochemical measurements for P3HT-50F, 23 24 25 P3HT-33F and P3HT-25F, respectively, indicated that below the 33% fluorination point, 26 27 fluorination had little effect on the electrochemical properties of P3HT. P3HT-33F also 28 29 30 possessed the best results in field effect transistors, reaching hole mobilities of 2.89 × 10-2 cm2 31 32 V-1 s-1. Despite the lowering of the HOMO levels and the improved on/off current ratios, the 33 34 overall OFET performances of highly regioregular P3HT are still better than those of their 35 36 37 regioregular, evenly fluorinated counterparts. 38 39 40 41 42 43 Acknowledgements 44 45 46 The authors acknowledge the Natural Sciences and Engineering Research Council of Canada 47 48 49 (NSERC) and the Canadian Institute for Advanced Research (CIFAR) for their support. J. T. B. 50 51 thanks the NSERC for a Canada Graduate Scholarship, S. A. L. thanks the U.S. Department of 52 53 Energy Energy Efficiency and Renewable Energy postdoctoral fellowship (Solar Energy 54 55 56 program) for funding and J. de G. S. thanks the Brazilian Coordination for the Improvement of 57 58 59 60 35 ACS Paragon Plus Environment Submitted to Macromolecules Page 36 of 41 1 2 3 4 Higher Education Personnel (CAPES) (Program number: Julio de Goes 5 6 Soares/88888.934050/2014-00). A.A.-G. acknowledges support from the Department of Energy 7 8 Basic, Energy Sciences Program in Theory and Modeling under contract DE-SC0015959. J. T. 9 10 11 B. thanks Rodica Neagu Plesu for help with certain polymer characterization methods. 12 13 14 15 16 17 Supporting Information 18 19 20 Additional figures, NMR spectra, details of synthesis, computational data, MALDI-ToF data, 21 22 23 fabrication methods for the preparation of OTFTs and device output and transfer curves. 24 25 26 27 28 29 References 30 31 32 1. Okamoto, T.; Nakahara, K.; Saeki, A.; Seki, S.; Oh, J. H.; Akkerman, H. B.; Bao, Z.; 33 34 35 Matsuo, Y., Aryl−Perfluoroaryl Substituted Tetracene: Induction of Face-to-Face π−π Stacking 36 37 and Enhancement of Charge Carrier Properties. Chem. Mater. 2011, 23 (7), 1646-1649. 38 39 2. Stuart, A. C.; Tumbleston, J. R.; Zhou, H.; Li, W.; Liu, S.; Ade, H.; You, W., Fluorine 40 41 42 Substituents Reduce Charge Recombination and Drive Structure and Morphology Development 43 44 in Polymer Solar Cells. J. Am. Chem. Soc. 2013, 135 (5), 1806-1815. 45 46 47 3. Chen, H.-Y.; Hou, J.; Zhang, S.; Liang, Y.; Yang, G.; Yang, Y.; Yu, L.; Wu, Y.; Li, G., 48 49 Polymer solar cells with enhanced open-circuit voltage and efficiency. Nature Photon. 2009, 3 50 51 (11), 649-653. 52 53 54 55 56 57 58 59 60 36 ACS Paragon Plus Environment Page 37 of 41 Submitted to Macromolecules 1 2 3 4 4. Liang, Y.; Feng, D.; Wu, Y.; Tsai, S.-T.; Li, G.; Ray, C.; Yu, L., Highly Efficient Solar 5 6 Cell Polymers Developed via Fine-Tuning of Structural and Electronic Properties. J. Am. Chem. 7 8 Soc. 2009, 131 (22), 7792-7799. 9 10 11 5. Koster, L. J. A.; Mihailetchi, V. D.; Blom, P. W. M., Ultimate efficiency of 12 13 polymer/fullerene bulk heterojunction solar cells. Appl. Phys. Lett. 2006, 88 (9), 093511. 14 15 6. Yum, S.; An, T. K.; Wang, X.; Lee, W.; Uddin, M. A.; Kim, Y. J.; Nguyen, T. L.; Xu, S.; 16 17 18 Hwang, S.; Park, C. E.; Woo, H. Y., Benzotriazole-Containing Planar Conjugated Polymers with 19 20 Noncovalent Conformational Locks for Thermally Stable and Efficient Polymer Field-Effect 21 22 Transistors. Chem. Mater. 2014, 26 (6), 2147-2154. 23 24 25 7. Nguyen, T. L.; Choi, H.; Ko, S. J.; Uddin, M. A.; Walker, B.; Yum, S.; Jeong, J. E.; Yun, 26 27 M. H.; Shin, T. J.; Hwang, S.; Kim, J. Y.; Woo, H. Y., Semi-crystalline photovoltaic polymers 28 29 30 with efficiency exceeding 9% in a [similar]300 nm thick conventional single-cell device. Energ. 31 32 Environ. Sci. 2014, 7 (9), 3040-3051. 33 34 8. Leclerc, N.; Chávez, P.; Ibraikulov, O.; Heiser, T.; Lévêque, P., Impact of Backbone 35 36 37 Fluorination on π-Conjugated Polymers in Organic Photovoltaic Devices: A Review. Polymers 38 39 2016, 8 (1), 11. 40 41 9. El Kassmi, A.; Fache, F.; Lemaire, M., Poly(3-fluorothiophene). J. Electroanal. Chem. 42 43 44 1994, 373 (1), 241-244. 45 46 10. Sakamoto, Y.; Komatsu, S.; Suzuki, T., Tetradecafluorosexithiophene:  The First 47 48 Perfluorinated Oligothiophene. J. Am. Chem. Soc. 2001, 123 (19), 4643-4644. 49 50 51 11. Heeney, M.; Farrand, L.; Giles, M.; Thompson, M.; Tierney, S.; Shkunov, M.; Sparrowe, 52 53 D.; McCullouch, I. U.S. Patent 6676857 B2, 2004. 54 55 56 57 58 59 60 37 ACS Paragon Plus Environment Submitted to Macromolecules Page 38 of 41 1 2 3 4 12. Gohier, F.; Frère, P.; Roncali, J., 3-Fluoro-4-hexylthiophene as a Building Block for 5 6 Tuning the Electronic Properties of Conjugated Polythiophenes. J. Org. Chem. 2013, 78 (4), 7 8 1497-1503. 9 10 11 13. Koo, B.; Sletten, E. M.; Swager, T. M., Functionalized Poly(3-hexylthiophene)s via 12 13 Lithium–Bromine Exchange. Macromolecules 2015, 48 (1), 229-235. 14 15 14. Fei, Z.; Boufflet, P.; Wood, S.; Wade, J.; Moriarty, J.; Gann, E.; Ratcliff, E. L.; McNeill, 16 17 18 C. R.; Sirringhaus, H.; Kim, J.-S.; Heeney, M., Influence of Backbone Fluorination in 19 20 Regioregular Poly(3-alkyl-4-fluoro)thiophenes. J. Am. Chem. Soc. 2015, 137 (21), 6866-6879. 21 22 15. Homyak, P. D.; Liu, Y.; Harris, J. D.; Liu, F.; Carter, K. R.; Russell, T. P.; Coughlin, E. 23 24 25 B., Systematic Fluorination of P3HT: Synthesis of P(3HT-co-3H4FT)s by Direct Arylation 26 27 Polymerization, Characterization, and Device Performance in OPVs. Macromolecules 2016, 49 28 29 30 (8), 3028-3037. 31 32 16. Hoth, C. N.; Choulis, S. A.; Schilinsky, P.; Brabec, C. J., On the effect of poly(3- 33 34 hexylthiophene) regioregularity on inkjet printed organic solar cells. J. Mater. Chem. 2009, 19 35 36 37 (30), 5398-5404. 38 39 17. Gao, Y.; Zhang, X.; Tian, H.; Zhang, J.; Yan, D.; Geng, Y.; Wang, F., High Mobility 40 41 Ambipolar Diketopyrrolopyrrole-Based Conjugated Polymer Synthesized Via Direct Arylation 42 43 44 Polycondensation. Adv. Mater. 2015, 27 (42), 6753-6759. 45 46 18. Bura, T.; Blaskovits, J. T.; Leclerc, M., Direct (Hetero)arylation Polymerization: Trends 47 48 and Perspectives. J. Am. Chem. Soc. 2016, 138 (32), 10056-10071. 49 50 51 19. Koch, F. P. V.; Smith, P.; Heeney, M., “Fibonacci’s Route” to Regioregular Oligo(3- 52 53 hexylthiophene)s. J. Am. Chem. Soc. 2013, 135 (37), 13695-13698. 54 55 56 57 58 59 60 38 ACS Paragon Plus Environment Page 39 of 41 Submitted to Macromolecules 1 2 3 4 20. Smith, K.; Barratt, M. L., Highly Selective 5-Substitution of 3-Methylthiophene via 5 6 Directed Lithiation. J. Org. Chem. 2007, 72 (3), 1031-1034. 7 8 21. Loewe, R. S.; Khersonsky, S. M.; McCullough, R. D., A Simple Method to Prepare 9 10 11 Head-to-Tail Coupled, Regioregular Poly(3-alkylthiophenes) Using Grignard Metathesis. Adv. 12 13 Mater. 1999, 11 (3), 250-253. 14 15 22. Loewe, R. S.; Ewbank, P. C.; Liu, J.; Zhai, L.; McCullough, R. D., Regioregular, Head- 16 17 18 to-Tail Coupled Poly(3-alkylthiophenes) Made Easy by the GRIM Method:  Investigation of the 19 20 Reaction and the Origin of Regioselectivity. Macromolecules 2001, 34 (13), 4324-4333. 21 22 23. Miyakoshi, R.; Yokoyama, A.; Yokozawa, T., Catalyst-Transfer Polycondensation. 23 24 25 Mechanism of Ni-Catalyzed Chain-Growth Polymerization Leading to Well-Defined Poly(3- 26 27 hexylthiophene). J. Am. Chem. Soc. 2005, 127 (49), 17542-17547. 28 29 30 24. Pouliot, J.-R.; Wakioka, M.; Ozawa, F.; Li, Y.; Leclerc, M., Structural Analysis of 31 32 Poly(3-hexylthiophene) Prepared via Direct Heteroarylation Polymerization. Macromol. Chem. 33 34 Phys. 2016, 217 (13), 1493-1500. 35 36 37 25. Xiao, S. S.; Qiu, C.; Qiu, C. X. U.S. Patent 6894145 B2, 2005. 38 39 26. Winter, J. D.; Deshayes, G.; Boon, F.; Coulembier, O.; Dubois, P.; Gerbaux, P., MALDI- 40 41 ToF analysis of polythiophene: use of trans-2-[3-(4-t-butyl-phenyl)-2-methyl- 2- 42 43 44 propenylidene]malononitrile—DCTB—as matrix. J. Mass Spectrom. 2011, 46 (3), 237-246. 45 46 27. Liu, J.; Loewe, R. S.; McCullough, R. D., Employing MALDI-MS on 47 48 Poly(alkylthiophenes):  Analysis of Molecular Weights, Molecular Weight Distributions, End- 49 50 51 Group Structures, and End-Group Modifications. Macromolecules 1999, 32 (18), 5777-5785. 52 53 28. Komber, H.; Senkovskyy, V.; Tkachov, R.; Johnson, K.; Kiriy, A.; Huck, W. T. S.; 54 55 56 Sommer, M., Ring Walking versus Trapping of Nickel(0) during Kumada Catalyst Transfer 57 58 59 60 39 ACS Paragon Plus Environment Submitted to Macromolecules Page 40 of 41 1 2 3 4 Polycondensation Using Externally Initiated Electron-Accepting Thiophene–Benzothiadiazole– 5 6 Thiophene Precursors. Macromolecules 2011, 44 (23), 9164-9172. 7 8 29. Holdcroft, S., Determination of molecular weights and Mark–Houwink constants for 9 10 11 soluble electronically conducting polymers. J. Polym. Sci., Part B: Polym. Phys. 1991, 29 (13), 12 13 1585-1588. 14 15 30. Kiriy, A.; Senkovskyy, V.; Sommer, M., Kumada Catalyst-Transfer Polycondensation: 16 17 18 Mechanism, Opportunities, and Challenges. Macromol. Rapid Commun. 2011, 32 (19), 1503- 19 20 1517. 21 22 31. Kim, J.-S.; Kim, J.-H.; Lee, W.; Yu, H.; Kim, H. J.; Song, I.; Shin, M.; Oh, J. H.; Jeong, 23 24 25 U.; Kim, T.-S.; Kim, B. J., Tuning Mechanical and Optoelectrical Properties of Poly(3- 26 27 hexylthiophene) through Systematic Regioregularity Control. Macromolecules 2015, 48 (13), 28 29 30 4339-4346. 31 32 32. Jackson, N. E.; Savoie, B. M.; Kohlstedt, K. L.; Olvera de la Cruz, M.; Schatz, G. C.; 33 34 Chen, L. X.; Ratner, M. A., Controlling Conformations of Conjugated Polymers and Small 35 36 37 Molecules: The Role of Nonbonding Interactions. J. Am. Chem. Soc. 2013, 135 (28), 10475- 38 39 10483. 40 41 33. Li, Z.; Lin, H.; Jiang, K.; Carpenter, J.; Li, Y.; Liu, Y.; Hu, H.; Zhao, J.; Ma, W.; Ade, 42 43 44 H.; Yan, H., Dramatic performance enhancement for large bandgap thick-film polymer solar 45 46 cells introduced by a difluorinated donor unit. Nano Energy 2015, 15, 607-615. 47 48 34. Kohn, P.; Huettner, S.; Komber, H.; Senkovskyy, V.; Tkachov, R.; Kiriy, A.; Friend, R. 49 50 51 H.; Steiner, U.; Huck, W. T. S.; Sommer, J.-U.; Sommer, M., On the Role of Single Regiodefects 52 53 and Polydispersity in Regioregular Poly(3-hexylthiophene): Defect Distribution, Synthesis of 54 55 56 57 58 59 60 40 ACS Paragon Plus Environment Page 41 of 41 Submitted to Macromolecules 1 2 3 4 Defect-Free Chains, and a Simple Model for the Determination of Crystallinity. J. Am. Chem. 5 6 Soc. 2012, 134 (10), 4790-4805. 7 8 35. Zhang, L.; Colella, N. S.; Liu, F.; Trahan, S.; Baral, J. K.; Winter, H. H.; Mannsfeld, S. 9 10 11 C. B.; Briseno, A. L., Synthesis, Electronic Structure, Molecular Packing/Morphology Evolution, 12 13 and Carrier Mobilities of Pure Oligo-/Poly(alkylthiophenes). J. Am. Chem. Soc. 2013, 135 (2), 14 15 844-854. 16 17 18 36. Perepichka, I. F.; Perepichka, D. F., Handbook of Thiophene-based Materials: 19 20 Applications in Organic Electronics and Photonics. John Wiley and Sons, Ltd. : 2009; Vol. 1, p 21 22 417. 23 24 25 37. McCullough, R. D., The Chemistry of Conducting Polythiophenes. Adv. Mater. 1998, 10 26 27 (2), 93-116. 28 29 30 38. Trznadel, M.; Pron, A.; Zagorska, M.; Chrzaszcz, R.; Pielichowski, J., Effect of 31 32 Molecular Weight on Spectroscopic and Spectroelectrochemical Properties of Regioregular 33 34 Poly(3-hexylthiophene). Macromolecules 1998, 31 (15), 5051-5058. 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 41 ACS Paragon Plus Environment