A CO2-switchable amidine monomer: synthesis and characterization

Abstract Smart system employed CO2 gas as new trigger has been attracting enormous attention in recent years, but few monomers that are capable of switching their hydrophobicity/hydrophility upon CO2 stimulation have been reported. A novel CO2 responsive monomer, 4-vinylbenzyl amidine, is designed and synthesized in this work with N,N-dimethylacetamide dimethyl acetal and 4-vinylbenzyl amine that is prepared through the Gabriel reaction. In bi-phase solvent of n-hexane and water, the monomer dissolves in n-hexane first and then transforms into water upon the CO2 treatment, indicating a hydrophobic to hydrophilic transition. This transformation is demonstrated as reversible by monitoring the conductivity variation of its wet dimethyl formamide solution during alternate bubbling/removing CO2. The protonation of 4-vinylbenzyl amidine upon CO2 treatment is demonstrated by 1H NMR which also accounts for the dissolubility change. The reversible addition-fragmentation chain-transfer polymerization of this monomer is also performed, finding the reaction only occurs in glacial acetic acid. The reason can be ascribed to the different radical structure produced in different solvent.


Introduction
CO 2 as a new trigger for smart systems has various advantages over its traditional counterparts such as pH, temperature, redox, light and voltage [1][2][3][4]. The first virtue is the good switchabilility, which means the responsive process can be repeated for much more times without significant attenuation or any contamination into the system [5]. The easy-removing feature of the gas state of CO 2 may account for this characteristic. Furthermore, the CO 2 is a metabolite of living cells rendering it good biocompatibility [6,7]. Last but not the least, CO 2 is a waste gas with wide availability and low price. All these advantages push CO 2 -responsive compounds and polymers into the spotlight [1].
Jessop team [5] first reported a long-chain alkyl amidine compound which worked as a CO 2 -switchable surfactant that can stabilize emulsion under exposing of CO 2 and break the emulsion by bubbling inert gas in higher temperature. Naturally, researchers want to figure out what would happen if the amidines groups are introduced into polymers. Some efforts have been made to achieve CO 2 -responsive polymers by post modification. Take our previous work as an example, we prepared poly(4chloromethylstyrene) (PCMS) first by reversible addition-fragmentation chain transfer (RAFT) polymerization.
Then the chloro groups were transferred into azido units in the second step. Finally, a homemade N′-propargyl-N,N-dimethylacetamidines (PDAA) were coupled with the azido units through a 'click' chemical process, yielding amidine-containing polymers [8,9]. Lowe and coworkers [10] synthesized polymers containing pentafluorophenyl acrylate (PFPA) and then substituted the pentafluorophenyl groups with amidine species histamine (HIS), resulting in amidine side chains. Both of these methods can target CO 2 -responsive polymers efficiently, but they seem tedious involving with functionalization after polymerization. Yuan and his coworkers [11,12] first synthesized monomer (N-amidino)dodecyl acrylamide (AD) that can be switched by CO 2 and developed 'breathable' vesicles with its block copolymer. Subsequently, similar acrylamide monomers containing amidine groups were synthesized with same route as well [13,14]. Meanwhile, the CO 2 -sensitivity of commercial available monomers containing tertiary amine and acid groups were discovered by Zhao's team [15][16][17][18]. However, the CO 2 -responsive monomer is rarely reported to date, resulting in limited choice can be made to fabricate CO 2 -sensitive polymers.

Synthesis of 4-vinylbenzyl amidine (compound 2)
6.6 g N,N-Dimethylacetamide dimethyl acetal (45 mmol) was added into 150 mL round-bottom flask, followed by dropping 5.0 g colorless transparent liquid obtained from last step (compound 1) under N 2 protection. The mixture was stirred for 15 min at room temperature and then heated to 65 °C. After 2 h reaction, the product was collected by removing side product methanol and excess N,N-Dimethylacetamide dimethyl acetal with rotary evaporation. 1

RAFT polymerization of 4-vinylbenzyl amidine
The RAFT polymerization is carried out as following typical procedure. 1.02 g 4-vinylbenzyl amidine (5.0 mmol), 14 mg CTPPA (0.05 mmol) and 3 mg initiator ACVA and 2 mL of solvent was added into a reaction tube. Then the system was deoxygenated with three freeze-thaw cycles or bubbling N 2 for 30 min followed by keeping at 70 °C for 24 h. After the reaction, the product was checked with 1 H NMR.

CO 2 -responsiveness of 4-vinylbenzyl amidine
As above-mentioned, we previously developed CO 2responsive polymers by combing RAFT polymerization and 'click' chemistry, which is not suitable for preparation of complex copolymers, such as multi-block and star structures [8,9]. Therefore, we try to synthesize a CO 2 -responsive monomer in this work. After we confirm the chemical structure of this monomer by NMR (see experimental part), its solubility is tested, finding it can be dissolved in common organic solvent including n-hexane, CH 2 Cl 2 , CHCl 3 , THF, DMF, DMSO. To check out the CO 2 -sensitivity, we then reaction. Then a CO 2 -responsive monomer, 4-vinylbenzyl amidine, is produced after a followed reaction of 4-Vinylbenzylamine with N,N-Dimethylacetamide dimethyl acetal (Scheme 1). The CO 2 -responsive feature of this monomer is demonstrated by visualizing the hydrophobility to hydrophility transition and monitoring the conductivity variation as well as NMR shifts of its solution under CO 2 -stimulation. The discovery of this work paves the way to development of new CO 2 -sensitive polymer materials.

Synthesis of 4-Vinylbenzylamine (compound 1)
In a 250 mL round-bottom flask, 22.9 g 4-vinylbenzyl chloride (0.15 mol) and 27.8 g phthalimide potassium salt (0.15 mol) were dissolved into 100 mL dimethyl formamide (DMF) solution and stored at 50 °C for 4 h with stirring. The solvent was removed by reduced pressure distillation after reaction. Then the mixture was dissolved in CHCl 3 and washed with NaOH solution (0.2 mol·L −1 ) and water successively. The raw product was concentrated and recrystallized in methanol for two times, yielding colorless transparent crystals. investigate whether it can transform from organic solvent into water phase, meaning whether it can transition from hydrophobicity to hydrophility. As shown in Figure 1, in a bi-phase solvent of n-hexane and water (v/v = 1:1), several drops monomer is added into the system. The monomer is dissolved into n-hexane in upper phase without CO 2 treatment, appearing as a yellow solution above water. Then CO 2 gas is bubbled into the solution, resulting in a yellow aqueous solution in the bottom. Meanwhile, the upper n-hexane becomes colorless and transparent. The n-hexane is a non-polar solvent, this transition implies the monomer can transform from hydrophobicity into hydrophilicity upon CO 2 stimulation.
To confirm the switchability of this monomer, we monitor the conductivity variation of its wet DMF solution during CO 2 is bubbling and removing alternately. As shown in Figure 2, the conductivity of the monomer solution jumps from 43.0 to 900.1 μS cm −1 after bubbling CO 2 for 4 min. Subsequently, it drops back to around 45.6 μS cm −1 after removing CO 2 by N 2 treatment. Furthermore, this jumping-dropping cycle can be repeated for four times without significant attenuation, indicating the CO 2 -responsiveness of this monomer is reversible. As a comparison, the conductivity of wet DMF solution without monomer inside only increases from 1.6 to 2.4 μS cm −1 under the stimulation of CO 2 , which further proves the switchable transition of the monomer under stimulation of CO 2 .
Beside the conductivity, we further investigate the reaction of the monomer with CO 2 by 1 H NMR characterization ( Figure 3). The monomers were dissolved into wet DMSO-d6 and characterized by 1 H NMR before and after the treatment of CO 2 . By comparing the main proton peak around the amidine group, one can easily indentify the significant chemical shift, i.e.    oxygen (entry 5, Table 1). Nevertheless, the product appears as cross-linked gel with yellow color rather than polymer dissolved in solution ( Figure 4). This gel cannot be dissolved into any solvents. Further improvements of this polymerization are still progressing in our group. Interestingly, the de-oxygenization method really affects the success of this reaction, if comparing entry 3 and 5 in Table 1. The reason may be ascribed to the highly volatile feature of HAc, which make it firstly filling the reaction tube in vacuum that restrains the discharge of oxygen in solution.
Then we come back to the problem, why the 4-vinylbenzyl amidine cannot be polymerized in neutral state without acetic acid? In the presentence of initiator, two kinds of radicals are possible produced; structure 3 and 4 ( Figure 5). The radical 3 is more stable then 4, because it forms a huge conjugated structure combining benzene ring and the donor amidine group. That means the 4-vinylbenzyl amidine transforms into radical 3 in organic reaction system, hindering the polymerization we desired. When the amidine groups are protonated by acetic acid, the possible huge conjugated structure 3 cannot form anymore and Figure 3, i.e. the amidine react with CO 2 and water producing bicarbonate. Actually, similar reaction of amidine and CO 2 in the presence of water has been recognized by other researchers [21]. This protonation produces a compound with positive charge [5] which also accounts for why the monomer can transform from hydrophobic to hydrophilic state under stimulation of CO 2 .

Polymerization of 4-vinylbenzyl amidine
After confirm the chemical structure and CO 2 -responsive feature of the monomer, we try to carry out the RAFT polymerization (Scheme 2). However, we get nothing when the reaction is done in organic solvent, neither in THF nor DMF. The reason, in our mind, might be the basicity of the amidine group (will discuss further). Long and his coworkers [22] reported RAFT polymerization of 4-vinylimidazole in glacial acetic acid (HAc), which is another basic monomer that is difficult to conduct controlled polymerization. Inspired by this discovery, we perform RAFT polymerization of 4-vinylbenzyl amidine using HAc as solvent (Scheme 2). The reaction conditions are illustrated in Table 1.
At first, we tried to polymerize 4-vinylbenzyl amidine in mixed solvent of DMF and HAc, but failed. Then various efforts were made, including utilization of pure HAc, increasing of the initiator ratio and optimization of deoxygenization method (Table 1). Finally, we find this monomer can be polymerized in conditions: (1) pure HAc, (2) initiator/CTA = 1, (3) bubbling N 2 for 30 min to remove   3 and 4) and reported amandine monomers by the groups of yuan (compound 5, ref. [11]) and yung (compound 6, ref. [14]).

Scheme 2. raFT polymerization of 4-vinylbenzyl amidine.
the polymerization occurs through the same mechanism of the radical polymerization of polystyrene. It should be point out that the reported amidine monomers (compounds 5 and 6 in Figure 5) can be polymerized through a radical mechanism in their natural state because they have no concerning to form stable structure like radical 3.

Conclusions
In conclusion, a CO 2 -responsive monomer, 4-vinylbenzyl amidine, was designed and synthesized in this work. It transforms from hydrophobic to hydrophilic state under the stimulation of CO 2 , and this transition is reversible. This monomer can be polymerized by RAFT technique in glacial acetic acid, yielding a gel-like homo-polymer. The results of this work provide another choice for preparing new CO 2 responsive polymers and pave a way to the development of novel CO 2 -sensitive 'smart' system.