Transition-metal-free controlled polymerization for poly(p-aryleneethynylene)s

A catalytic amount of fluoride anions promoted the polymerization of 1-pentafluorophenyl-4-[(trimethylsilyl)ethynyl]benzene, providing a high-molecular-weight polymer in a chain-growth-like manner.


Introduction
Among the most extensively studied families of molecular optoelectronic materials, such as organic eld-effect transistors (FETs) and organic electroluminescent (EL) devices, poly(paryleneethynylene)s (PAEs) are one of the most important types of materials. 1 The polymers are commonly prepared by polycondensations using Sonogashira-or Stille-type cross couplings 2 or by an alkyne metathesis. 3 In all cases, however, the molecular weight (MW) and polydispersity index (PDI) of the polymers are not controlled because the polycondensations proceed by a step-growth mechanism. In contrast to these conventional step-growth polycondensations, chain-growth polymerizations have been demonstrated recently, but they require transition-metal catalysts. 4 There remains a need to achieve additional synthetic methods for well-dened polymers with control of the MW, the polydispersity, modication of the end group, and also block copolymerization, which could offer new architectures and materials. 5 Here, we describe a transition-metal-free controlled polymerization for the attainment of PAEs. The polymerization proceeds in a chain-growth-like manner to afford the polymers with controlled MWs and low PDIs. This could be an alternative synthetic method for well-controlled PAEs.

Results and discussion
We designed 1-pentauorophenyl-4-[(trimethylsilyl)ethynyl] benzene 1 as a monomer and examined its polymerization with a catalytic amount of uoride anions (Scheme 1). This is because uoride anions were found to catalyze silylacetylene activation for subsequent reaction with a number of electrophiles, 6,7 and also because regioselective S N Ar reactions of per-uoroaryl groups with nucleophiles 8,9 are well studied. Very recently, we demonstrated the transition-metal-free polymerization of 2-peruoroaryl-5-trimethylsilylthiophenes promoted by uoride anions to afford polymers with controlled MWs and low PDIs. 10 In contrast, the reported polycondensation of hexauorobenzene and 1,4-bis[(trimethylsilyl)ethynyl]benzene was not controlled. 11 Recently, Bielawski et al. reported a controlled Pd catalyzed transfer polycondensation for poly-(p-phenyleneethynylene). 12 A catalytic amount of uoride anions smoothly promoted the polymerization of 1. Table 1 summarizes the polymerization results. For example, the reaction of 1 with 5 mol% tetrabutylammonium uoride (TBAF) 13 as a uoride anion source in tetrahydrofuran (THF) at room temperature for 2 h led to poly(p-tetrauorophenylene-phenylene-ethynylene) 2 with a numberaveraged MW (M n ) of 4400 and a PDI of 1.31 in an 81% isolated yield (entry 2). The MWs measured by size-exclusion chromatography (SEC)-multi-angle light scattering (MALS) were almost the same as those measured by SEC. When the polymerization of 1 was examined with respect to varying the mol% of TBAF, the MW of the polymer linearly increased with an increasing monomer to TBAF ratio (entries 1-4), while the PDIs were Scheme 1 Polymerization of 1.
relatively low (#2). Addition of 1 to a solution of TBAF also afforded the polymer (entry 5). However, tetramethylammonium uoride was not a good initiator because some remained undissolved in THF (entry 6). Potassium uoride or cesium uoride in the presence of 18-crown-6 or cryptand[2.2.2] did not give a good result (entries 7-9). Tetrabutylammonium diuorotriphenylsilicate was not effective for the polymerization of 1 (entry 10).
The obtained polymer has a structure with high regioregularity, as demonstrated by 1 H, 13 C, and 19 F NMR analyses ( Fig. S1 in the ESI †). The 1 H NMR spectrum of 2 shows major signals at around 0.8, 1.4, and 4.0 ppm owing to the side chain on the phenylene units, along with a small signal at 3.7 ppm arising from the ethynyl group at the polymer end. In the 19 F NMR spectrum of the polymer, two strong signals and three weak signals were found, which are assigned to the 1,4-tetra-uorophenylene units in the main chain and the penta-uorophenyl group at the polymer end, respectively. Thus, the NMR spectra are consistent with a high regioregularity for the polymer main chain, indicating that the polymerization process itself must be highly regioselective. In addition, because the polymer ends are designated as the pentauorophenyl and the ethynyl groups, the integral ratio of the peaks from the side chain on the main chain and the end group in the 1 H and 19 F NMR spectra provide a M n of ca. 5000, which is in reasonable agreement with the MW estimated by SEC (Table 1, entry 2). Furthermore, matrix-assisted laser desorption ionization timeof-ight (MALDI-TOF) mass spectra also indicate that the polymer has pentauorophenyl and ethynyl groups at its ends ( Fig. S2 in the ESI †).
The controlled MWs and relatively low PDIs found in Table 1 indicate that the polymerization proceeds in a chain-growthlike manner under the specied conditions. To provide further evidence for this, we monitored the polymerization of 1 with 2.5 mol% TBAF as a function of monomer conversion. Fig. 1 shows the M n and PDI as functions of monomer conversion. The polymerization of 1 was fast, with the conversion of 1 being up to 50% aer a few minutes (Table S1 and Fig. S3 in the ESI †). As shown in Fig. 1, the linear relationship between M n and monomer conversion, and the relatively low PDI, conrm the chain-growth-like process.
Scheme 2 shows a possible polymerization mechanism. In the initial reaction step, a uoride anion attacks the trimethylsilyl group of 1 to form a pentacoordinate silicate. The silicate is quite reactive and could regioselectively attack the 4position of the pentauorophenyl group of 1 to reproduce a uoride anion. The uoride anion would then transfer intramolecularly to the trimethylsilylethynyl group at the polymer end, where there may be an anion-p interaction that produces an associated pair. 14 Then, the uoride anion catalyzes the polymerization to give polymer 2. In the polymerization, the reactivity of the trimethylsilyl group and/or the silicate at the polymer end is changed by replacing the 4-position of the pentauorophenyl group of 1. Watson et al. discussed the change in reactivity in the polymerization of 1,4-bis [(trimethylsilyl)ethynyl]benzene and hexauorobenzene by  substitution with uoride anions. 11 Further description of the polymerization mechanism requires further studies. According to the polymerization mechanism, all of the propagating polymer chains contain reactive pentacoordinate silicate at a chain terminus. Indeed, a pentacoordinated alkynylsilicate, which is prepared in situ by the reaction of 1-(4-trimethylsilyl)phenyl-2-(triethoxysilyl)acetylene 3 and potassium t-butoxide, 15 efficiently initiated the polymerization of 1 in the presence of cryptand[2.2.2] to give polymer 4 with a M n of 12 300 and a PDI of 2.08 in 89% yield (Scheme 3). 16 The 1 H NMR spectrum of 4 shows peaks at around 0.3 and 3.4 ppm arising from the trimethylsilyl and ethynyl groups at the polymer ends, respectively, where the observed integral ratio is in good agreement with the calculated value (calcd for -SiMe 3 /-C^CH 9.0, found 8.8) within the experimental errors (Fig. S4 in the ESI †). 17 Next, we demonstrated the anionic "living" character of this polymerization end by applying it to the synthesis of a well-dened block copolymer (Scheme 4). Aer polymerization of 1 with a catalytic amount of TBAF, part of the mixture was studied to analyze the MW of the rst block, and then a solution of 5 was added to the reaction mixture to give block copolymer 6. 18 As shown in Fig. 2, the SEC curve of the polymer obtained at the end of the reaction was shied toward a higher MW compared to that of the rst polymerization. The rst/second block ratio (n/m) of 6 was found to be 0.98 by 1 H NMR analysis (Fig. S6 in the ESI †).
We also extended our method to the polymerization of a variety of aromatic monomers, 7-9, which possess trimethylsilylethynyl and pentauorophenyl groups (Scheme 5). The reaction of 2-pentauorophenyl-5-[(trimethylsilyl)ethynyl]thiophene 7 with 5 mol% TBAF gave polymer 10 with a M n of 11 200 and a PDI of 2.09 in a moderate yield. Fluorene-incorporated 8 was polymerized by addition of 5 mol% TBAF to give polymer 11 with a M n of 10 600 and a PDI of 1.99 in a good yield. A phenyleneethynylene monomer 9 was also polymerized to afford polymer 12 with a M n of 11 300 and a PDI of 1.85. The structures of these polymers are highly ordered, as demonstrated by NMR analyses (Fig. S7-S9 in the ESI †), indicating that the polymerizations also proceed with high regioselectivities.
Next, we turned our attention to exploring the surfaceterminated polymerization of 1. Au surfaces of nanometer-sized particles are an ideal substrate for this study because of their well-established chemistry and their features which could be exploited for molecular device applications. 19 Surface-terminated polymerization was accomplished by the addition of a surface-enhanced Raman scattering (SERS)-active Au nanoparticle (f 15 nm) array 20 to a polymerization mixture of 1 (Scheme 6). Aer standing for 1 h, the plate was washed with THF and then analyzed by Raman spectroscopy. SEC analysis of Scheme 2 Possible polymerization mechanism of 1.
Scheme 3 Polymerization of 1 with a pentacoordinated alkynylsilicate prepared by the reaction of 3 and potassium t-butoxide.  the polymerization mixture shows that the resulting polymer, 13, has a M n of 4700 and a PDI of 1.79. The n C^C stretching region in the SERS spectra is particularly informative. The Raman spectrum of 2 shows a signal at around $2200 cm À1 owing to the ethynyl groups in the polymer main chain (Fig. 3). On the other hand, in the Raman spectrum of 13, a broad and downshied signal is observed at 2050 cm À1 . The red-shi of $150 cm À1 indicates a strong interaction with the Au substrate, which is in agreement with previous reports. 21 Thus, the ethynyl group of the polymer end is attached to the gold substrate via covalent Au-C^C bonds. Although this is a preliminarily demonstration, it represents a possible step for molecular device applications. Further details will be reported.

Conclusions
We have developed a transition-metal-free controlled synthesis of PAEs promoted by uoride anions. The polymerization proceeds in an anionic chain-growth-like manner to afford PAEs with controlled MWs and relatively low PDIs. The polymerization end is active and affords a block copolymer. We also demonstrated the synthesis of a surface-terminated PAE on a Au nanoparticle array. We expect that this concept can be extended to prepare other well-controlled conjugated polymers, providing new opportunities in optoelectronics and other applications. Further study along this line is currently in progress.