1-Thioglycerol (98%), epichlorohydrin, photoinitiator 819, sodium hydroxide, tetrabutylammonium iodide (TBAI) were purchased from Macklin Inc. and used without further purification. 4-PA was purchased from Wuhan Lanabai Pharmaceutical & Chemical Co. and purified by reduced pressure distillation. Tetrahydrofuran (THF) was distilled under argon atmosphere from sodium/benzophenone. CO₂ (99.999%) was used directly as received. (R,R)-(−)-N,N-Bis(3,5-ditert-butylsalicylidene)-1,2-cyclohexanediaminocobalt (III) chloride (SalenCoCl) and 5,10,15,20-Tetraphentlporphine aluminum (III) chloride (TPPAlCl) were prepared according to the previous literature (Aida and Inoue, 1983). Bis-(triphenylphosphine)iminium chloride (PPNCl) was purchased from Tianjin Sienna Biochemical Technology Co. All chemicals were reagent grade unless otherwise noted.

Fourier transform infrared (FT-IR) spectroscopy was performed on a Shimadzu IRAffinity-1S FTIR spectrophotometer. ¹H NMR spectra of the products were performed on a Bruker AV-400 MHz NMR spectrometer, chemical shift values were referenced to TMS as internal standard at 0.0 ppm, the solvent was CDCl₃ or C₄D₈O. The molecular weight (M _n) and polydispersity (Ɖ) of the polymer was determined by gel permeation chromatography (GPC) on a Waters 2414 binary system with a refractive index detector, calibrated with polystyrene standards. The column temperature was maintained at 35 °C during the test using THF as the eluent at a flow rate of 1.0 mL/min. Differential scanning calorimetric (DSC) tests were conducted with the heating rate of 10 K/min from 20°C to 180 °C under N₂ atmosphere. Samples for thermal analyses were all purified. The water contact angle was determined using the JC2000 contact angle measuring instrument.

4-PA (50.00 g) was dissolved in acetone (300.00 mL) with stirring at room temperature for 10 min. Then NaOH solution (74.00 g, 30 wt%) was dropwise added with vigorous stirring for 4 h. After filtration, the filtrate was dried at 45°C under vacuum for 24 h to obtain 4-PA sodium salt (4-PANa).

GPA with epoxide groups were obtained through the reaction of the prepared 4-PANa (0.37 mol, 45.00 g) and epichlorohydrin (ECH) (2.14 mol, 198.00 g) at 130°C for 0.5 h. Then, TBAI (2.80 mmol, 1.05 g) was added, and the reaction mixture was stirred for another 1 h. The reaction mixture was further cooled to ambient temperature and centrifuged for 5 min at 6,000 rpm to remove the unreacted 4-PANa from the suspension solution, and then the clean GPA monomers were obtained after excess ECH was removed using a vacuum rotary evaporator and decompression distillation.

The synthesis of polycarbonates from GPA (0.02 mol, 3.30 g) and CO₂ (3 MPa) was performed in a pre-dried 10 mL autoclave in the presence of catalyst SalenCoCl (0.04 mmol, 25.00 mg) and cocatalyst PPNCl (0.04 mmol, 23.0 mg). The autoclave was filled with CO₂ at a pressure of 3 MPa. The reactions were performed at different temperatures and reaction times with magnetic stirring at a stirring speed of 1,000 rpm. The crude products were dissolved in dichloromethane (DCM, 5.00 mL) and further precipitated through dropwise addition into 50.00 mL ethanol to remove the catalysts, unreacted raw materials, and products with low molecular weight. The purified polycarbonates were obtained after vacuum drying at 25 °C for 48 h.

The synthetic procedure for the terpolymerization of the GPA, PO, and CO₂ was similar to the above copolymerization procedure. The reactions were conducted with three mole ratios of GPA/PO, 3:7, 5:5, and 7:3. Correspondingly, the volumes of the GPA monomers in these three experiments were 0.90 mL (6.40 mmol), 1.50 mL (10.70 mmol) and 2.10 mL (15.00 mmol), and the volumes of the PO were 1.05 mL (14.90 mmol), 0.75 mL (10.70 mmol) and 0.45 mL (6.40 mmol), respectively. After the reaction, the products were dissolved in DCM, precipitated in ethanol, and then dried to obtain the purified polycarbonates.

Photoinitiator 819 (8.40 mg, 0.02 mmol), polycarbonate of Table 1, entry 3 (4.18 g), and 1-mercaptoglycerol (2.17 g, 0.02 mol) were dissolved in anhydrous THF, and further irradiated by ultraviolet light for 4 h. After the reaction, the solution was concentrated using a rotary evaporator and then precipitated in cold ethyl ether. After filtration, the filtrate was dried at 45°C under vacuum overnight to obtain the post-modification of polycarbonate.

GPA was synthesized by the nucleophilic substitution reaction between epichlorohydrin and the sodium salt of 4-PA. 4-PANa was prepared by the saponification reaction between pentenoic acid and aqueous NaOH solution in acetone, and further reacted with epichlorohydrin using the phase transfer catalyst tetrabutylammonium iodide (TBAI) to obtain 4-PA glycidyl ester in 91% yield.

As depicted in Figure 3, the produced GPA monomers were analyzed using FTIR spectra. Compared to the spectra of 4-PA, the spectra of GPA monomer displayed a new characteristic absorbance peak at 855 cm⁻¹, which was attributed to epoxy groups and suggested that the epoxy groups were successfully inserted into the terminal of the fatty acid chains. The disappearance from the typical signals of hydroxyl group (-OH) of -COOH between 2700 and 2500 cm⁻¹ indicates that the carboxyl group of 4-PA was properly converted. Identifiable signals at 1,640 cm⁻¹ and 3,008 cm⁻¹, corresponding to the (CH-CH) and (CH = CH) of vinyl groups on the unsaturated fatty acids chains, were also detected, suggesting that the original vinyl groups of the unsaturated fatty acids chains were maintained. These characteristic chemical structures were also demonstrated by ¹HNMR spectroscopy. As shown in Figure 4, the characteristic hydrogen atoms of the GPA monomers were as follows: b, 5.7 ppm; a, ∼4.8 ppm; e, ∼4.3 ppm and 3.7 ppm; f, ∼3.1 ppm; g, ∼2.5 ppm and 2.7 ppm; d, ∼2.4 ppm; and c, ∼2.3 ppm. These expected spectral results further confirmed the successful introduction of the oxirane group and demonstrated the interior double bonds retained on the linear fatty acid chains, offering a promising potential site for further synthetic modification.

To achieve high selectivity of polycarbonates, two different catalysts involving TPPAlCl/PPNCl and SalenCoCl/PPNCl were used to catalyze the copolymerization. TPPAlCl/PPNCl is a highly efficient catalytic system for copolymerization of PO with CO₂ (Jung et al., 1999). The active center of the system is aluminum, which is considered to be green catalyst with no harm to the soil and living creatures (Wang et al., 2018). The catalytic reaction was carried out at 60°C for 24 h using GPA/TPPAlCl/PPNCl with a molar ratio of 500:1:1. FTIR spectra showed characteristic absorption peaks for cyclic carbonate and polycarbonate at 1800 cm⁻¹ and 1750 cm⁻¹ (Figure 5), respectively, and further analysis by ¹H NMR spectroscopy revealed up to 35% cyclic by-products (Figure 6).

To improve the polymer selectivity of the copolymerization reaction, another catalytic system SalenCoCl/PPNCl, which proved to be an efficient binary catalyst for CO₂ copolymerization with high reactivity and selectivity (Lu and Wang, 2004; Cohen et al., 2005; Lu et al., 2006), was used to catalyze the copolymerization of GPA with CO₂. FTIR spectra showed characteristic absorption peak for cyclic carbonate at 1800 cm⁻¹ was still present (Figure 5). Further calculation the conversion of GPA and the polymer selectivity by ¹H NMR spectrum. The reaction catalyzed by SalenCoCl/PPNCl achieved nearly 100% conversion of GPA within 8 h, demonstrating much higher activity than TPPAlCl/PPNCl. The polymer selectivity of the reaction was calculated to be more than 95%. The absence of ether linkage peak at 3.5 ppm suggests that no polyether was produced during the reaction (Figure 6).

For crude products obtained from GPA/CO₂ copolymerization catalyzed by SalenCoCl/PPNCl system, the peaks at 5.1 and 4.2 ppm were attributed to the protons on the methyl and methylene of the carbonate linkage, respectively, while the peaks of 4.9, 4.6 and 4.3 ppm were attributed to the protons on the methyl and methylene of the five-membered cyclic carbonate. The peaks at 3.5–3.8 ppm were attributed to the protons in the polyether chain segment (Figure 6). The ¹H NMR spectrum of the purified product showed that the peaks of cyclic carbonate (j, k, l) at 4.9 ppm, 4.6 ppm, and 4.3 ppm completely disappeared, and the pure polycarbonate was obtained, as shown in Figure 7.

As shown in Table 1, GPA/CO₂ copolymerization catalyzed by SalenCoCl/PPNCl was conducted under various reaction times and temperatures. The conversion of the epoxides increased with extending reaction time (entries 1 and 2), reaching nearly 100% after 12 h. The polymer selectivity of the reaction was 92%, and almost no polyether linkage was formed. When the reaction time was extended to 24 h, the polymer selectivity decreased to 79% (entry 4), which was mainly due to the unstable structure of Salen (III)CoCl during the reaction. A portion of SalenCo(III) in the system was reduced to SalenCo(II) which might catalyze the cycloaddition reaction to form cyclic carbonate instead of the copolymerization to form polycarbonate, resulting in a decrease in polymer selectivity. In addition, the molecular weight of the polycarbonates declined with the prolongation of the reaction time, falling from 15.9 kg/mol at 12 h to 6.3 kg/mol at 48 h (entries 2–5). The decrease of the molecular weight was attributed to the back-biting reaction of the polycarbonate chain segment to form cyclic carbonate. Increasing the reaction temperature to 40°C (entry 8) decreased the polymer selectivity to 23% after 8 h of reaction, indicating that the stability of SalenCoCl declines as the temperature rises, leading to the generation of a large number of cyclic carbonate by-products.

To prevent side reactions, we lowered the reaction temperature to 0°C (entries 6, 7), and the polymer selectivity of the reaction increased to 95%. However, the catalytic activity of SalenCoCl/PPNCl was also reduced under 0°C, and the conversion of epoxide was only 76% after 48 h. Similarly, the molecular weight of the polymers also decreased, but it showed an increasing tendency with the prolongation of the reaction time. The molecular weights of polycarbonates gradually increased and reached 17.1 kg/mol at 5 MPa (entries 3, 9 and 10).

Thermal analysis of GPA/CO₂ showed that the glass transition temperature (T _g) of polycarbonate was −36°C, which was significantly lower than that of CO₂/PO copolymer, most likely due to internal plasticization effects or micro-Brownian motion of the long alkyl side chains of GPA with ester groups (Okada et al., 2011). The terpolymerization of PO, GPA and CO₂ was an effective method for achieving a balance between the hard and soft polymer segments. To adjust the T _g of CO₂-based polycarbonate, a series of PO/GPA/CO₂ terpolymer were synthesized via terpolymerization. When the mole ratio of GPA/PO changed from 7:3 to 3:7, the glass transition temperature (T _g ) of the polymers rose from 2°C to 19 °C (Figure 8).

The vinyl groups in the side chain of the polymers offer the possibility of constructing functional polycarbonates through post-modification. 1-thioglycerol was successfully linked to the side chain of the polycarbonates via a radical-mediated thiol-ene click reaction, which was confirmed by ¹H NMR spectrum (Figure 9). The peaks attributed to the vinyl groups at 5.2 ppm (a) and 5.8 ppm (b) completely disappeared along with appearance of new peaks which were attributed to 1-thioglycerol at 3.5 ppm (p, q), indicating complete conversion of the alkene groups via thiol-ene click reaction. In general, the thiol-ene click reaction occurs predominately with anti-Markovnikov regioselectivity. However, a hydrogen signal attributed to the methyl at 1.1 ppm (r) was found in the post-modification products, indicating the presence of partial Markovnikov regioselectivity in the thiol-ene click process.

Contact angle measurements were conducted on GPA polycarbonates before and after the click reaction (Figure 10). It was observed that the contact angle of the unmodified polycarbonate increased from 92.1° to 87.2° after post-modification. This increase in contact angle indicates that the modified polymers became more hydrophilic compared to the unmodified polymers. The introduction of the hydroxyl group from 1-thioglycerol contributed to this increase in hydrophilicity. Therefore, the post-modification of polycarbonates with vinyl side groups has extensive potential applications in the preparation of biobased functional materials, such as antibacterial materials.