Synthesis of Terpyridine End-Modified Polystyrenes through ATRP for Facile Construction of Metallo-Supramolecular P3HT-b-PS Diblock Copolymers

Complementary complexation between 2,2′:6′,2″-terpyridine (tpy) and 6,6″-dianthracenyl-substituted tpy in the presence of Zn(II) ions provided an efficient strategy for construction of metallo-supramolecular diblock copolymers. To synthesize well-defined tpy-modified polystyrenes (PSs), an Fe(II) bis(tpy) complex bearing α-bromoester as a metallo-initiator was applied to atom transfer radical polymerization (ATRP) to avoid poisoning the Cu(I) catalyst. Subsequently, a series of tpy-functionalized PSs was obtained after the decomplexation of  junction by tetrakis(triethylammonium) ethylenediaminetetraacetate (TEA-EDTA) under mild conditions. The metallo-supramolecular poly(3-hexylthiophene) (P3HT)-block-PS diblock copolymers were prepared by simply mixing the corresponding terminally tpy-modified homopolymers with Zn(II) ions, and further characterized by 1H NMR and diffusion ordered spectroscopy (DOSY) experiments. The approach using metallo-initiators for ATRP offers an opportunity to construct tpy-functionalized polymers with controllable molecular weights and low polydispersities. Through the spontaneous heteroleptic complexation, a variety of metallo-supramolecular diblock copolymers with tunable block ratios can be easily constructed.


General Procedure for Complexation Reactions
To a CHCl 3 solution (5 mL) of PS n (n = 19, 33, 85, 106, 161, and 235) and P3HT 54 in an equimolar ratio calculated from the corresponding number average molecular weights (M n,NMR ), 1 Eq of Zn(OTf) 2 in MeOH (5 mL) was added. After the reaction mixture was stirred at room temperature for 5 min, the solvent was evaporated under reduced pressure to give the corresponding diblock copolymers.

Results and Discussion
Inspired by the pioneering research of metallo-initiators for ATRP, we designed and synthesized an Fe(II) bis(tpy) complex bearing α-bromoester as a bifunctional metallo-initiator (4) (Scheme 1). The α-bromoester modified tpy-based initiator (3) was synthesized from the precursor 4'-(4-hydroxyphenyl)-tpy (1) in moderate yield. First, 1 was alkylated by 2-chloroethanol to give 4 -(4-(2-hydroxyethoxy)phenyl)-tpy (2). The following esterification was achieved by reaction of 2 with α-bromoisobutyryl bromide in the presence of triethylamine at room temperature to afford compound 3. Subsequently, the complexation was conducted by adding a MeOH solution of FeCl 2 (0.5 Eq) into a CHCl 3 solution of ligand 3 (1 Eq) at room temperature, followed by counter-anion exchange with NH 4 PF 6 (10 Eq), to yield 4 as a dark purple powder in quantitative yield. The suitable crystals for single-crystal X-ray crystallography were obtained by slow diffusion of diethyl ether into an MeCN solution of 4. The structure of complex 4 was unequivocally established by its crystal structure ( Figure S8), NMR spectroscopy ( Figures S5 and S6) and ESI-MS ( Figure S7). hydroxyethoxy)phenyl)-tpy (2). The following esterification was achieved by reaction of 2 with αbromoisobutyryl bromide in the presence of triethylamine at room temperature to afford compound 3. Subsequently, the complexation was conducted by adding a MeOH solution of FeCl2 (0.5 Eq) into a CHCl3 solution of ligand 3 (1 Eq) at room temperature, followed by counter-anion exchange with NH4PF6 (10 Eq), to yield 4 as a dark purple powder in quantitative yield. The suitable crystals for single-crystal X-ray crystallography were obtained by slow diffusion of diethyl ether into an MeCN solution of 4. The structure of complex 4 was unequivocally established by its crystal structure ( Figure  S8), NMR spectroscopy ( Figures S5 and S6) and ESI-MS ( Figure S7).
Our attempts to conduct Cu(I)-mediated ATRP of styrene using the initiator 3 were unsuccessful possibly due to the strong chelating ability of tpy ligands acting as a catalyst poison. To investigate the controllability of ATRP of styrene initiated by the bifunctional metallo-initiator 4, a kinetic study was performed via a typical ATRP protocol as follows. Initiator 4 (45.2 mg, 34.9 μmol) and CuBr (15.0 mg, 104.8 μmol) were added to a degassed Schlenk flask equipped with a stir bar. Subsequently, MeCN (1.3 mL) and styrene (3.6 mL, 31.4 mmol) were added into the flask, which was degassed by three freeze-pump-thaw cycles, followed by the addition of PMDETA (21.9 μL, 104.8 μmol), and then stirred at 110 °C. The polymerization solution was periodically sampled via a pre-degassed syringe to monitor the conversion of monomer by 1 H NMR and calculate the theoretical molecular weight (Mn,theo). The sampled solution was further treated with TEA-EDTA in dimethylformamide (DMF) for 1 day at room temperature to decomplex the <tpy-Fe(II)-tpy> junction to afford the tpy end-modified PSs. The corresponding Mn,GPC and PDI (Mw/Mn) were determined by GPC, and Mn,NMR was calculated from the 1 H NMR peak integral ratios of polymerized styrene and terminal tpy. Our attempts to conduct Cu(I)-mediated ATRP of styrene using the initiator 3 were unsuccessful possibly due to the strong chelating ability of tpy ligands acting as a catalyst poison. To investigate the controllability of ATRP of styrene initiated by the bifunctional metallo-initiator 4, a kinetic study was performed via a typical ATRP protocol as follows. Initiator 4 (45.2 mg, 34.9 µmol) and CuBr (15.0 mg, 104.8 µmol) were added to a degassed Schlenk flask equipped with a stir bar. Subsequently, MeCN (1.3 mL) and styrene (3.6 mL, 31.4 mmol) were added into the flask, which was degassed by three freeze-pump-thaw cycles, followed by the addition of PMDETA (21.9 µL, 104.8 µmol), and then stirred at 110 • C. The polymerization solution was periodically sampled via a pre-degassed syringe to monitor the conversion of monomer by 1 H NMR and calculate the theoretical molecular weight (M n,theo ). The sampled solution was further treated with TEA-EDTA in dimethylformamide (DMF) for 1 day at room temperature to decomplex the <tpy-Fe(II)-tpy> junction to afford the tpy end-modified PSs. The corresponding M n,GPC and PDI (M w /M n ) were determined by GPC, and M n,NMR was calculated from the 1 H NMR peak integral ratios of polymerized styrene and terminal tpy.
The kinetic study on ATRP of styrene initiated by the metallo-initiator 4 was summarized in Figure 1. The semilogarithmic kinetic plot of ln([M] 0 /[M]) versus reaction time indicated the first-order radical polymerization process and the radical concentration was kept constant during the polymerization (Figure 1a). In addition, the experimental molecular weights (M n,NMR ) were in good agreement with the theoretical ones (M n,theo ) and linearly increased with respect to the monomer conversion (Figure 1b), implying the absence of significant chain transfer reactions. Moreover, the polydispersities were decreased with increasing conversion (Figure 1b), and a clear shift to higher molecular weights with a mono-distribution was evidenced by GPC traces (Figure 1c). These observations suggested that the well-controlled ATRP of styrene could be initiated by 4.
Polymers 2020, 12, x 4 of 10 conversion (Figure 1b), implying the absence of significant chain transfer reactions. Moreover, the polydispersities were decreased with increasing conversion (Figure 1b), and a clear shift to higher molecular weights with a mono-distribution was evidenced by GPC traces (Figure 1c). These observations suggested that the well-controlled ATRP of styrene could be initiated by 4. Based on the kinetic result of ATRP of styrene, a series of tpy-functionalized PSn (n = 19, 33, 85, 106, 161, and 235) with varying chain lengths was prepared via ATRP under optimized conditions ( Table 1). The formation of well-defined PSn could be evident from the narrow molecular weight distributions and the consistency between Mn,NMR and Mn,GPC. It is noteworthy that the terminal bromide at PS chain-ends susceptible to elimination easily led to formation of a double bond during MALDI-TOF-MS measurements [50]. Nevertheless, the high fidelity in the tpy chain-end functionality was verified by the corresponding MALDI-TOF-MS peaks (Figures S10, S12, S14, S16, and S18). On the other hand, the well-defined P3HT54 (DP = 54, Mn,GPC = 8,800 Da, Mw/Mn = 1.23) endfunctionalized with a 4-(4'-(6,6"-dianthracenylterpyridyl))phenyl group was obtained through the Suzuki-Miyaura coupling reaction of mono-brominated P3HT (Br-P3HT) with 6,6"-dianthracenyl-4'-(4-boronophenyl)tpy (Figure 2a) [19]. Notably, Br-P3HT prepared by Grignard metathesis (GRIM) polymerization method possessed two isomeric chain-end structures, i.e., head-to-head and head-totail orientations, which could not be differentiated by the 1 H NMR spectrum of Br-P3HT but clearly seen in that of P3HT54 (Figure 2b) [51,52]. Therefore, the 3-hexylthiophene coupled with two 6,6 "dianthracenyl-substituted tpys (L1) was synthesized as a model compound to ensure the proper 1 H NMR assignments. The two sets of tpy signals of L1 corresponded to two types of chain-end connections, and the chain-end head-to-tail content of P3HT54 was estimated to be 22% ( Figure S26). The single molecular weight distribution in the MALDI-TOF-MS spectra ( Figure S27) strongly supported the high chain-end functionality for P3HT54.  Based on the kinetic result of ATRP of styrene, a series of tpy-functionalized PS n (n = 19, 33, 85, 106, 161, and 235) with varying chain lengths was prepared via ATRP under optimized conditions ( Table 1). The formation of well-defined PS n could be evident from the narrow molecular weight distributions and the consistency between M n,NMR and M n,GPC . It is noteworthy that the terminal bromide at PS chain-ends susceptible to elimination easily led to formation of a double bond during MALDI-TOF-MS measurements [50]. Nevertheless, the high fidelity in the tpy chain-end functionality was verified by the corresponding MALDI-TOF-MS peaks (Figures S10, S12, S14, S16, and S18). On the other hand, the well-defined P3HT 54 (DP = 54, M n,GPC = 8800 Da, M w /M n = 1.23) end-functionalized with a 4-(4 -(6,6"-dianthracenylterpyridyl))phenyl group was obtained through the Suzuki-Miyaura coupling reaction of mono-brominated P3HT (Br-P3HT) with 6,6"-dianthracenyl-4 -(4-boronophenyl)tpy (Figure 2a) [19]. Notably, Br-P3HT prepared by Grignard metathesis (GRIM) polymerization method possessed two isomeric chain-end structures, i.e., head-to-head and head-to-tail orientations, which could not be differentiated by the 1 H NMR spectrum of Br-P3HT but clearly seen in that of P3HT 54 (Figure 2b) [51,52]. Therefore, the 3-hexylthiophene coupled with two 6,6"-dianthracenyl-substituted tpys (L1) was synthesized as a model compound to ensure the proper 1 H NMR assignments. The two sets of tpy signals of L1 corresponded to two types of chain-end connections, and the chain-end head-to-tail content of P3HT 54 was estimated to be 22% ( Figure S26). The single molecular weight distribution in the MALDI-TOF-MS spectra ( Figure S27) strongly supported the high chain-end functionality for P3HT 54 .  We have demonstrated that the complementary complexation between 6,6"-substituted and unsubstituted tpy ligands with Zn(II) under ambient conditions could be applied to construction of the metallo-supramolecular diblock copolymers from two distinct tpy-modified homopolymers [19,48]. In the ligand design, the bulky 9-anthracenyl substituents effectively decelerated the formation rate of homoleptic complexes. Moreover, the X-ray single-crystal structure of [L2-Zn-L3] (Figure 3a) exhibited the π-π interactions between unsubstituted L3 and two anthracenyl substituents to facilitate the formation of heteroleptic complexes. Consequently, a series of metallosupramolecular diblock copolymers of [P3HT54-Zn-PSn] (n = 19, 33, 85, 106, 161, and 235) could be readily constructed from homopolymers PSn and P3HT54 in the presence of Zn(II) ions (Scheme 2). Due to the labile coordination bonds, the intact copolymers could not be detected by MALDI-TOF-MS [53]. Hence, the resultant diblock copolymers were characterized by 1   We have demonstrated that the complementary complexation between 6,6"-substituted and unsubstituted tpy ligands with Zn(II) under ambient conditions could be applied to construction of the metallo-supramolecular diblock copolymers from two distinct tpy-modified homopolymers [19,48]. In the ligand design, the bulky 9-anthracenyl substituents effectively decelerated the formation rate of homoleptic complexes. Moreover, the X-ray single-crystal structure of [L2-Zn-L3] (Figure 3a) exhibited the π-π interactions between unsubstituted L3 and two anthracenyl substituents to facilitate the formation of heteroleptic complexes. Consequently, a series of metallo-supramolecular diblock copolymers of [P3HT 54 -Zn-PS n ] (n = 19, 33, 85, 106, 161, and 235) could be readily constructed from homopolymers PS n and P3HT 54 in the presence of Zn(II) ions (Scheme 2). Due to the labile coordination bonds, the intact copolymers could not be detected by MALDI-TOF-MS [53]. Hence, the resultant diblock copolymers were characterized by 1  showed much smaller diffusion coefficients than expected, presumably because of the severe intermolecular aggregation, where the shorter PS length attenuated the interference in the assembly of P3HT segments [55].