Copolymers incorporated with β-substituted acrylate synthesized by organo-catalyzed group-transfer polymerization

Various copolymers incorporated with β-substituted acrylates, such as alkyl crotonates (e.g., methyl crotonate (MC), ethyl crotonate (EC), isopropyl crotonate (iPC), and n-butyl crotonate (nBC)) and methyl cinnamate (MCin), were synthesized by group-transfer polymerization (GTP) using a silicon-based Lewis acid catalyst. In addition to β-substituted acrylates, α-substituted acrylates (e.g., methyl methacrylate (MMA) and n-butyl methacrylate (nBMA)) were examined as comonomers. Proton nuclear magnetic resonance (1H NMR) spectroscopy and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) characterizations of the obtained copolymers revealed that each monomer component was incorporated sufficiently. The thermal stabilities of the resulting copolymers were investigated by dynamic mechanical analysis (DMA), indicating that the glass-transition temperature (Tg) of the copolymers can be widely varied over a relatively high-temperature range by selecting the optimal comonomer. More specifically, the Tg values of poly(MC-random-EC) (MC/EC molar ratio = 50/50), poly(MC-random-nBC) (MC/nBC molar ratio = 50/50), poly(MC-random-MCin) (MC/MCin molar ratio = 54/46), and poly(nBC-random-MCin) (nBC/MCin molar ratio = 56/44) were 173, 130, 216, and 167 °C, respectively. The copolymerization of β-substituted acrylates (i.e., crotonates and a cinnamate) using an organo-catalyzed group-transfer polymerization system with N-(tert-butyldimethylsilyl)bis(trifluoromethanesulfonyl)imide (Tf2NSitBuMe2) and 1-methoxy-1-(tert-butyldimethylsiloxy)-2-methyl-1-propene (tBuMe2SKA) was investigated. The tert-butyldimethylsilyl moiety served as an effective means for suppressing the termination reactions of the GTP system. One-stage and two-stage copolymerization strategies employing different monomer feed methods were evaluated and gave random and block copolymers, respectively. The glass-transition temperatures of the prepared copolymers varied significantly depending on the comonomer used during the synthesis.


Introduction
Group-transfer polymerization (GTP) was developed as a controlled living-polymerization method of (meth)acrylates by Webster et al. in the 1980s [1]. The first-generation GTP system used fluoride ions and silyl ketene acetals (SKAs) as the catalyst and the initiator, respectively. Since then, numerous improvements have been made to the GTP system to control the living-polymerization fashion and expand the range of available monomer substrates. Hertler et al. reported that Lewis acids, such as zinc halides, organo aluminum, and HgI 2 , can be applied to the controlled GTP of acrylates [2][3][4][5][6]. In addition, Kakuchi et al. developed a highly efficient organo-catalyzed GTP system using a silicon-based Lewis acid catalyst [7,8].
Because copolymerization can greatly enhance the extent of currently attainable polymeric properties, investigations into the copolymerizability and the applicable comonomer range of β-substituted acrylates could expand the potential polymer applications [31][32][33][34]. However, our GTP system did not guarantee copolymerization because the living chain end was frequently deactivated via cyclization termination reactions [29,30,35,36]. Recently, we demonstrated that changing the silyl groups of Tf 2 NSiMe 3 and MTS to a bulkier group, such as a tert-butyldimethylsilyl moiety, is an effective means for suppressing termination reactions in the GTP system [37]. Thus, we herein report our investigation into the copolymerization of β-substituted acrylates (i.e., crotonates and cinnamate) using an organo-catalyzed GTP system with N-(tert-butyldimethylsilyl)bis(trifluoromethanesulfonyl)imide (Tf 2 NSi t BuMe 2 ) and 1-methoxy-1-(tert-butyldimethylsiloxy)-2-methyl-1-propene ( tBuMe2 SKA). Two types of copolymerization strategies are evaluated based on different monomer feed methods. More specifically, a one-stage copolymerization is conducted to obtain a random copolymer, where two types of monomers are added simultaneously to the polymerization system. In addition, two-stage polymerization is conducted to obtain a block copolymer, where one monomer is homopolymerized in the first stage, and a second feed of comonomer is subsequently added to the system. The series of copolymers prepared via this one-stage copolymerization approach incorporate two types of β-substituted acrylates selected from MC: ethyl crotonate (EC), isopropyl crotonate (iPC), n-butyl crotonate (nBC), and methyl cinnamate (MCin). Moreover, block copolymers of β-substituted acrylates with common methacrylates, such as methyl methacrylate (MMA) and n-butyl methacrylate (nBMA), are prepared by the two-stage copolymerization approach using the livingpolymerization characteristics of the GTP system (Scheme 1). To the best of our knowledge, examples of addition copolymerization strategies incorporating cinnamates into the polymer backbone are limited [38][39][40]. We also investigated the thermal stabilities of a series of obtained copolymers by dynamic mechanical analysis (DMA) and thermogravimetric analysis (TGA).

Measurements
All molecular weight data for the polymer samples were obtained by gel permeation chromatography (GPC) at 40°C using a Shimadzu GPC system equipped with an RID-10A refractive index detector, a Shodex GPC K-802 column, and a Shodex GPC K-806 M column. CHCl 3 was used as the eluent at a flow rate of 0.8 mL min −1 . Polystyrene standards with a low polydispersity were used to obtain a calibration curve. 1 H NMR and 13 C NMR measurements were performed on a Varian NMR system. The samples were dissolved in CDCl 3, and 500 MHz 1 H NMR spectra were recorded at room temperature. The 125 MHz 13 C NMR spectra were recorded at room temperature in a CDCl 3 solution. The data were calibrated against tetramethylsilane (TMS, δ 0.00) and analyzed using Mnova NMR software.
TGA was performed on an SII Nano Technology TG/DTA7200 thermal analyzer from room temperature to 500°C in a nitrogen atmosphere at a heating rate of 10°C min −1 . T d5 and T d50 were defined as the temperature at which the weight loss reached 5% and 50%, respectively, of the total weight.
The T g of each polymer sample was measured by DMA (Perkin Elmer DMA800) under a nitrogen atmosphere by cooling the sample to −50°C at a rate of 10°C min −1 and then recording the DMA scan from −50 to +250°C at a heating rate of 5°C min −1 and a frequency of 1 Hz.

Group-transfer polymerization
All manipulations were performed in a dry and oxygen-free argon atmosphere using standard high-vacuum Schlenk techniques or a glovebox. A typical procedure for the onestage polymerization process (entry 3, Table 1) was as follows: MC (501 mg, 5 mmol), nBC (711 mg, 5 mmol), tBuMe2 SKA (10.8 mg, 0.050 mmol), and CH 2 Cl 2 (6.0 mL) were charged to a Schlenk tube in a glovebox. A CH 2 Cl 2 solution (2.7 mL) of Tf 2 NH (7.0 mg, 0.025 mmol) was added to initiate the polymerization reaction at 0°C. After 48 h, methanol (10 mL) was added to the reaction mixture, after which the solvents were removed by evaporation under reduced pressure. The residue was then dissolved in CHCl 3 , and the solution was poured into a large volume of hexane. The precipitated polymer was collected as the hexane-insoluble component by filtration, washed several times with hexane, and dried under vacuum for 24 h. For the two-stage polymerization process (entry 8, Table 1), the procedure was as follows: MC (501 mg, 5 mmol), tBuMe2 SKA (10.8 mg, 0.050 mmol), and CH 2 Cl 2 (4.0 mL) were charged to a Schlenk tube in a glovebox. A CH 2 Cl 2 solution (2.7 mL) of Tf 2 NH (7.0 mg, 0.025 mmol) was added to initiate the polymerization reaction at 0°C. After 24 h, a CH 2 Cl 2 solution (2.0 mL) of nBC (711 mg, 5 mmol) was added dropwise, and the polymerization reaction was allowed to continue for an additional 24 h. The second monomer feed was introduced without sampling from the first step because this manipulation may deactivate the living polymers by introducing trace amounts of water. Finally, the reaction was worked up using the same process that was employed for the one-stage polymerization.

Results and discussion
Copolymerization using two types of β-substituted acrylates Initially, the preparation of a series of copolymers incorporating two types of β-substituted acrylates (M 1 and M 2 ) selected from the MC, EC, iPC, nBC, and MCin groups was attempted by GTP using Tf 2 NSi t BuMe 2 (generated via the in situ reaction between tBuMe2 SKA and Tf 2 NH (Scheme S1, Supporting Information)) as the catalyst and tBuMe2 SKA as the initiator in CH 2 Cl 2 at 0°C at a molar feed ratio of M 1 : M 2 = 1:1. Table 1 summarizes the conversions of M 1 and M 2 (x 1 and x 2 , respectively), the polymer yields, the number-average molecular weights (M n ), and the dispersities (Đ) of the obtained polymers. Figure 1 shows the GPC curves for the synthesized polymers.
When the polymerization reaction was conducted using a simultaneous feed of M 1 and M 2 (one-stage polymerization) (entries 1-7, Table 1), the consumption of both monomers was observed. In the cases of entries 1-3, the use of MC as   initiator was 400, polymers possessing a high molecular weight (M n > 76 kg mol −1 ) and a relatively narrow dispersity (Ð < 1.2) were obtained. In particular, the polymer obtained by the reaction outlined in entry 3 (i.e., from MC and nBC) exhibited a high molecular weight of >107 kg mol −1 . In terms of entries 4-7, the use of alkyl crotonates (MC, EC, iPC, or nBC) as M 1 and MCin as M 2 , where the molar feed ratio of the sum of M 1 and M 2 versus the initiator was 100, gave polymers possessing a relatively high molecular weight (M n > 24 kg mol −1 ) and a relatively narrow dispersity (Ð < 1.3). The exception to this rule was entry 6, where the reactivity of iPC was low, which was likely due to the steric hindrance of the ester group.
In the case of the two-stage polymerization process (entries 8 and 9, Table 1), where M 1 was homopolymerized for 24 h in the first stage, followed by a second feed of M 2 and polymerization for an additional 24 h, the consumption of M 2 was barely observed. The molecular weights of the obtained polymers (M n = 53 100 g mol −1 in entry 8 and 63 900 g mol −1 in entry 9) were clearly smaller than that obtained for the reaction of entry 3 (M n = 107 600 g mol −1 ), where the polymerization process was conducted under conditions identical to those for entries 8 and 9, with the exception of the monomer feeding method. This result suggests that in the two-stage polymerization of β-substituted acrylates, the amount of living polymer derived from M 1 was low during the additional M 2 feed. This deactivation of a living polymer is thought to be caused by a side reaction, such as an intramolecular termination reaction (e.g., terminal cyclization of the propagating chain-end groups), as proposed for the GTP system [35,36].

Structural analyses of the obtained polymers
To confirm the structures of the obtained polymers, 1 H NMR analyses were conducted. Figure 2a shows the 1 H NMR spectrum of the polymer synthesized using MC as M 1 and EC as M 2 (entry 1, Table 1), in addition to those of poly (MC) and poly(EC), which were synthesized via a similar GTP route for comparison (Table S3, Supporting Information). Furthermore, Fig. 2b shows the 1 H NMR spectrum of the polymer synthesized using EC as M 1 and MCin as M 2 (entry 5, Table 1), in addition to those of the prepared poly (EC) and poly(MCin) (Table S3, Supporting Information). As shown in Fig. 2a, the 1 H NMR spectrum of entry 1 agrees well with the overlap of the spectra of poly(MC) and poly(EC). Similarly, as shown in Fig. 2b, the 1 H NMR spectrum of entry 5 is comparable to the superposition of  Table 1) and b EC as M 1 and MCin as M 2 (entry 5, Table 1), measured at 500 MHz in CDCl 3 at room temperature. The blue and green lines represent the 1 H NMR spectra of the homopolymers synthesized from M 1 and M 2 , respectively the spectra of poly(EC) and poly(MCin). These results suggest that the consumed monomers (M 1 and M 2 ) were successfully introduced into the polymer backbone.
The molecular structures of the obtained polymers were investigated in greater detail by MALDI-TOF-MS analysis. Figure 3a shows the mass spectrum of the polymer synthesized using MC as M 1 and iPC as M 2 (entry 2, Table 1), while Fig. 3b shows the mass spectrum of the polymer synthesized using EC as M 1 and MCin as M 2 (entry 5 in Table 1).
From the detailed analysis, it was confirmed that the observed peaks correspond to the sum of the masses of the copolymers containing M 1 and M 2 generated by GTP using tBuMe2 SKA and Na + (derived from CF 3 COONa added as an ionizing agent in the MALDI-TOF-MS measurements). The polymer shown in Fig. 3a was therefore determined to have the structural formula (CH 3

Time-conversion curve
Subsequently, we attained the time-conversion curve data to obtain information regarding the monomer unit arrangement in the copolymer backbone. Figure 4a shows the time-conversion curve for the copolymerization reaction using MC as M 1 and EC as M 2 (entry 1, Table 1), while Fig. 4b shows the time-conversion curve for the  Table 1) and b EC as M 1 and MCin as M 2 (entry 5, Table 1), measured using dithranol as a matrix and CF 3 COONa as an ionizing agent  Table 1).
As shown in Fig. 4a, the conversion ratio of M 1 (MC) versus M 2 (EC) was within the range of 0.5-1 from the initial term to the end term of the copolymerization process, suggesting that the polymerization rate of EC was only slightly faster than that of MC. This result indicates that in the case of the copolymer obtained for the reaction outlined in entry 1 (Table 1), the MC and EC monomer units were incorporated throughout the polymer backbone.
In contrast, as shown in Fig. 4(b), the conversion of M 1 (EC) was approximately tenfold greater than that of M 2 (MCin) in the initial stage of copolymerization. This result indicates that in the case of entry 5 (Table 1), EC monomer units were primarily incorporated at the head part of the polymer backbone, while MCin monomer units were mainly incorporated at the tail end. Thus, copolymerization using two types of β-substituted acrylates with greatly different polymerizabilities leads to the generation of a gradient-like random copolymer.
Copolymer of α-substituted acrylate and β-substituted acrylate Subsequently, the preparation of a series of copolymers from an α-substituted acrylate (M 1 , MMA or nBMA) and a β-substituted acrylate (M 2 , MC, nBC, or MCin) was attempted using the GTP protocol with Tf 2 NSi t BuMe 2 as the catalyst and tBuMe2 SKA as the initiator in CH 2 Cl 2 at 0°C , using a molar feed ratio of M 1 :M 2 = 1:1. Table 2 summarizes the conversions of M 1 and M 2 (x 1 and x 2 , respectively), the polymer yields, and the M n and Đ values of the obtained polymers, while Fig. 5 shows their GPC curves.
In the case of the two-stage polymerization reactions, where the α-substituted acrylate was homopolymerized in the first stage, followed by a second feed of the βsubstituted acrylate and an additional polymerization phase (entries 10-15, Table 2), sufficient consumption of both monomers was observed. When MMA or nBMA was employed as M 1 and an alkyl crotonate (MC or nBC) was used as M 2 , where the molar feed ratio of the sum of M 1 and M 2 versus the initiator was 400 (entries 10-14, Table 2), polymers possessing a relatively high molecular weight (M n > 56 kg mol −1 ) were obtained. In contrast, (entry 15, Table 2), when nBMA was used as M 1 and MCin was employed as M 2 , where the molar feed ratio of the sum of M 1 and M 2 versus the initiator was 100, a polymer possessing a relatively low molecular weight of 16,800 g mol −1 was obtained.
On the other hand, in the case of entry 16 (Table 2), where the first feed monomer was a β-substituted acrylate (MC) and the second feed monomer was an α-substituted acrylate (nBMA), the conversion of nBMA was particularly low (x 2 < 10%). In addition, in the case of entry 17, where a Performed on an aluminum block at 0°C under an argon atmosphere. The polymerization conditions were as follows: M 1 feed and M 2 feed = 5.0 mmol, initiator = tBuMe2 SKA, catalyst = Tf 2 NSi t BuMe 2 (generated via the in situ reaction between tBuMe2 SKA and Tf 2 NH (Scheme S1, Supporting Information)), solvent = CH 2 Cl 2 . Two-stage polymerization was conducted. one-stage copolymerization was conducted, the conversion of the α-substituted acrylate (nBMA) was low (<10%). The molecular weights of the obtained polymers (M n = 49 300 g mol −1 for entry 16, M n = 46 700 g mol −1 for entry 17) were smaller than that of entry 12 (M n = 57 600 g mol −1 ), despite the polymerization process being conducted under conditions identical to those used for entries 16 and 17 (with the exception of the monomer feeding method). These results suggest that the living chain end derived from the β-substituted acrylate exhibits poor reactivity toward the αsubstituted acrylate monomer. In addition, it was suggested that the living chain end derived from a β-substituted acrylate is more susceptible to a termination reaction, such as a cyclization reaction, than that derived from an α-substituted acrylate [29,35,36] (Table S1 and Fig. S35, Supporting Information).
As shown in Fig. 5, the polymers obtained from the reactions outlined in entries 10-13 (Table 2) exhibit bimodal peaks in their GPC curves that correspond to lower-and higher molecular weight polymers. The peaks corresponding to a lower molecular weight were attributed to the homopolymers of the α-substituted acrylate generated by a termination reaction (e.g., cyclization reaction) of the living chain end in the first stage. On the other hand, the peaks corresponding to a higher-molecular weight were attributed to the block copolymers of M 1 and M 2 , which propagated from the living homopolymers of the α-substituted acrylate that were alive during the second feed of the β-substituted acrylate monomer (Table S2, Supporting Information).
The polymer corresponding to entry 14 was obtained by shortening the time of the first-stage polymerization of nBMA with respect to entry 12 (12 h for entry 14 c.f. 24 h for entry 12). As shown in Fig. 5, the polymer obtained by the reaction presented in entry 14 gave an almost unimodal peak by GPC. This result confirms the hypothesis that the observed bimodal peaks for entry 12 were caused by the terminated homopolymers of nBMA. In addition, this result indicates that the block copolymer content can be enhanced by optimizing the polymerization time in the first stage.

Thermal properties of the copolymers
We then investigated the thermal properties (i.e., T g , T d5 , and T d50 ) of the obtained copolymers by DMA and TGA, as summarized in Table 3. In this study, T g was defined as the temperature corresponding to the maximum tan δ value in the DMA measurements because differential scanning calorimetry did not afford a clear T g . Table 3 shows that the T g values of the copolymers obtained in the reactions outlined in entries 1-7 and prepared from two types of β-substituted acrylates were >130°C, indicating a relatively high heat stability. This high T g can be rationalized via the stiffness of the polymer main chain containing substituents at the α-position (ester group) and the β-position (methyl or phenyl group). A high T g was also confirmed for the homopolymers of β-substituted acrylates synthesized by GTP using Tf 2 NSi t BuMe 2 and tBuMe2 SKA (Table S3, Supporting Information). Comparing entry 4 (copolymer of MC with MCin) with entries 1-3 (copolymer of MC as M 1 with EC, iPC, or nBC, respectively, as M 2 ) shows that the copolymer obtained using the conditions outlined in entry 4 exhibited the highest T g . This result suggests that cinnamate is a better comonomer than crotonate for enhancing the T g of the resulting copolymers. Among entries 1-3, greater numbers of carbon atoms on the ester group of M 2 resulted in lower T g values for the resulting copolymers. Similarly, among entries 4-7 (copolymers synthesized using MC, EC, iPC, or nBC, respectively, as M 1 , and MCin as M 2 ), the ester group of crotonate (as M 1 ) significantly influenced the T g value of the resulting copolymers. These results indicate that the T g of the copolymers described herein can be varied over a relatively high-temperature range by selecting the optimal comonomer.  Table 2). The blue and red inverted triangles represent lower-and higher-molecular-weight polymers, respectively In terms of the block copolymers prepared from an α-substituted acrylate (MMA or nBMA) and a β-substituted acrylate (MC, nBC, or MCin), the T g corresponding to the poly(MMA) block was detected at approximately 116-32°C (entries 10 and 11), while the T g corresponding to the poly(nBMA) block was detected at approximately 51-58°C (entries 12, 13, and 15). It has been previously reported that the T g of poly(MMA) ranges from 51-122°C, while that of poly(nBMA) ranges from 15-50°C [42]. In contrast, the value of T g derived from the poly(β-substituted acrylate) block was confirmed only as an indistinguishable or a small region. Thus, the T g values of the block copolymers in this DMA study strongly reflected the characteristics of the poly (α-substituted acrylate).
Finally, in the context of their thermal decomposition, the copolymers obtained herein were found to exhibit relatively high heat resistance, with T d5 > 325°C and T d50 > 376°C. The exceptions to this rule were the copolymers obtained under the conditions presented in entries 6 ( Table 1) and 15 (Table 2), which had relatively low molecular weights of <20 kg mol −1 .

Conclusion
We herein reported the preparation of various copolymers incorporated with β-substituted acrylates, such as alkyl crotonates (MC, EC, iPC, and nBC), and MCin, via the GTP process using a silicon-based Lewis acid catalyst and SKAs possessing a relatively bulky silyl moiety, such as the tert-butyldimethylsilyl group. Our results indicated that using this organo-catalyzed GTP system, the copolymerization of β-substituted acrylates could be achieved by selecting a suitable comonomer. When multiple βsubstituted acrylates were used, a simultaneous monomer feed system was necessary to ensure smooth copolymerization. In addition, when an α-substituted acrylate was used as the comonomer, a multistage copolymerization method (where the α-substituted acrylate is homopolymerized in the first stage followed by a second feed of the β-substituted acrylate) resulted in the formation of a block copolymer incorporated with each component. This monomer feed method must consider the susceptibility of termination reactions and/or the copolymerizability of the living chain end derived from the monomers used. 1 H NMR spectroscopy and MALDI-TOF-MS were used to confirm the successful copolymerization reactions, and the thermal stabilities of the resulting copolymers were investigated by DMA and TGA. These results indicated that the T g values of the prepared copolymers can be varied by changing the comonomer. Future studies should focus on the use of additional comonomers, such as acrylamides and acrylonitriles, which are known to be suitable for application in GTP systems, to gain a better understanding of the comonomer scope and to develop high-performance polymeric materials.