Butane-1,4-diol, octane-1,8-diol, decane-1,10-diol, dodecane-1,12-diol, diphenyl carbonate, 12-amino-1-dodecanol and 1,2-dimethoxybenzene were purchased from Tokyo Kasei Kogyo Co., Inc. (Tokyo, Japan). Hexaethylene glycol, octaethylene glycol and anisole (anhydrous) were purchased from Aldrich Chem. Co., Inc. (Milwaukee, WI, USA). 12-Aminododecanoic acid and toluene (anhydrous) were purchased from Wako Chemical Co., Inc (Tokyo, Japan). Immobilized lipase from Candida antarctica was kindly supplied by Novozymes Japan Ltd. (Chiba, Japan); CALB: Novozym 435, a lipase (lipase B) from C. antarctica produced by submerged fermentation of a genetically modified Aspergillus oryzae microorganism and absorbed on a macroporous acrylic resin (10,000 propyl laurate units per gram; lipase activity based on ester synthesis). The enzyme was dried under vacuum over P₂O₅ at 25 °C for 6 h. Thermally deactivated CALB was prepared by heating the enzyme for 15 min in steam using an autoclave at 120 °C and then freeze-drying under vacuum.

The weight-average (M_w) and number-average (M_n) molecular weight and the molecular weight distribution (M_w/M_n) of the polymer were measured by means of size exclusion chromatography (SEC) using SEC columns (Shodex K-805L + K-800D, Showa Denko Co., Ltd., Tokyo, Japan) with a refractive index detector ; chloroform was used as the eluent at 1.0 mL·min⁻¹. The M_w, M_n and M_w/M_n values of the oligomers were also measured by means of SEC (Shodex K-802 + K-800D, Showa Denko Co., Ltd., Tokyo, Japan) under the same conditions. The SEC system was calibrated with polystyrene standards of narrow molecular weight distribution. The molecular weight of oligomers was also measured by the matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectra using a Bruker Ultraflex mass spectrometer equipped with a nitrogen laser. The detection was performed in the reflection mode, a cobalt matrix was used for oligomers, sodium bromide was used as the cation source, and positive ionization was used. Supercritical carbon dioxide fluid (scCO₂) chromatography (SFC) was used for the fractionation of the oligomer mixture, using a JASCO SFC-201 chromatograph equipped with a 4.6 mm i.d. × 250 mm preparative column packed with silica gel (SFCpak SIL-5, JASCO Ltd., Tokyo, Japan), using scCO₂ as the mobile phase and ethanol as a modifier. The scCO₂ pressure was controlled at 25 MPa and the column temperature was maintained at 70 °C. Chromatograms were recorded using a UV detector operating at a wavelength of 215 nm. ¹H NMR spectra were recorded on a Varian 300 Fourier transform spectrometer (Varian Inc., CA, USA) operating at 300 MHz, and ¹³C NMR spectra were recorded on a ECA-500 Fourier transform spectrometer (JEOL Ltd., Tokyo, Japan) operating at 125 MHz.

The glass-transition temperatures (T_g), crystallization temperatures (T_c) and melting temperatures (T_m) of the polymers were determined by differential scanning calorimetry (DSC-60, Shimadzu Co., Kyoto, Japan). The measurements were made with a sample (3–5 mg) on a DSC plate. The polymer samples were heated at the rate of 10 °C·min⁻¹ from 30 °C to 150 °C (first scan), rapidly cooled to −50 °C at the rate of −20 °C·min⁻¹, and then scanned at the same heating rate of 10 °C over the temperature range of −50 °C to 150 °C (second scan). The Young’s modulus of film samples produced by the hot-press method was measured using an Autograph instrument (Shimadzu Co., Kyoto, Japan) at the cross-head speed of 30 mm·min⁻¹. The static water contact angle of film samples produced by the solvent-cast method was measured using a contact angle meter (DMs-400, Kyowa Interface Science Co., Ltd., Saitama, Japan).

Preparation of PEEU and PEU are outlined in Scheme I.

Preparation of PEDU is outlined in Scheme II.

Two novel types of polyurethanes containing ester linkages, namely a diol-diacid-type poly(ester-ether-urethane) (PEEU) and poly(ester-urethane) (PEU), and a hydroxy acid type poly(ester-diurethane) (PEDU), were designed and synthesized by a green method that avoided the use of diisocyanate (Scheme I and Scheme II). The urethane linkage was formed by the reaction of amine and phenyl carbonate, while ester linkages were periodically introduced into the main chain of the polymer by lipase to afford efficient chemical recycling segments.

First, tetramethylene bis(phenyl carbonate) (1) was prepared for the synthesis of the urethane linkage. Conventionally, 1 is produced by the reaction of phenyl chloroformate and butane-1,4-diol in the presence of a base catalyst [15,16]. In spite of the diisocyanate-free method for the synthesis of the urethane linkage, however, phenyl chloroformate is produced by the reaction of phenol and phosgene. Therefore, in order to avoid the use of materials derived from phosgene, a new route to 1 was carried out by the reaction of diphenyl carbonate (17 mmol) and butane-1,4-diol (0.85 mmol) with CALB in toluene solution. An excess amount of diphenyl carbonate was required to minimize oligomer formation. Next, the diurethane-containing dicarboxylic acid (2) was prepared by the reaction of 1 and sodium 12-aminododecanoate in chloroform/methanol. The urethane bond formed quickly without any catalyst.

The direct polycondensation of 2 and diol was carried out using CALB in 1,2-dimethoxybenzene as a solvent at 100 °C for 48 h. In the absence of CALB or in the presence of thermally deactivated CALB, no polycondensation of 2 and diol was observed. These results indicate that CALB is the active catalyst for the polycondensation process. Table 1 shows the polymerization conditions and some analytical data of the polymers.

The polymerization behavior of 2 with various diols was similar, irrespective of the oxyethylene or the methylene chain length, such that the M_ws of the polymers were around 100,000 g·mol⁻¹. The yields were 80–90% after purification by reprecipitation. Figure 1 shows the effect of temperature on M_w and M_w/M_n of PEEU (6), prepared by the polycondensation of 2 and hexaethylene glycol. The M_w values of the polymer increased with reaction temperature from 90 to 100 °C, and were almost constant at around 110,000 g·mol⁻¹ at temperatures between 100 and 120 °C. Reaction temperatures over 100 °C are higher than that used in aqueous solutions. However, it has been reported that under relatively dry conditions, enzymes, such as lipases, proteases and esterases, remain catalytically active at temperatures between 90–120 °C [17–20]. With temperatures over 120 °C, the M_w value gradually decreased due to thermal deactivation of the enzyme, and at 140 °C practically no polymer was produced. Therefore, further studies were carried out at 100 °C.

The substrate concentration is the decisive factor for polymerization and depolymerization. In general, in a more concentrated solution, the equilibrium shifts toward polymerization. Figure 2 shows the effects of monomer concentration of 2 on M_w and M_w/M_n of the polymer in 1,2-dimethoxybenzene solution. Polymerization occurred quickly and a high M_w value was produced at a monomer concentration around 200 mg·mL⁻¹. The M_w of the polymer decreased with monomer concentrations higher than 200 mg·mL⁻¹ due to the solidification of the polymerization system. The enzyme concentration also affected the M_w value of the polymer. Figure 3 shows the effect of enzyme concentration (relative to 2) on M_w and M_w/M_n of the polymer. Although CALB was immobilized on an acrylic resin, a relatively large amount of immobilized lipase was required for rapid polycondensation. The highest M_w was obtained for the polymerization at a lipase concentration of 60 wt%. However, the M_w of the polymer decreased with higher amounts of enzyme due to the increasing amount of enzyme containing water, which was responsible for both initiation and termination of the polymerization. Based on these results, further studies were carried out at 60 wt% lipase concentration. The M_w and M_w/M_n values increased gradually with time. Figure 4 shows the time course for the polycondensation of 2 and hexaethylene glycol. M_w and M_w/M_n values increased gradually by transesterification and reached equilibrium after 48 h.

In order to produce a polyurethane with a higher T_m compared to that of diol-diacid type PEEU(x) and PEU(y), a hydroxy acid type polyurethane (PEDU) with a higher content of urethane moieties was prepared. Diurethane containing hydroxy acid (4) was prepared by the reaction of 1 with 12-amino-1-dodecanol, followed by the reaction with sodium 12-aminododecanoate. It was found that 4 polymerized readily by CALB in anisole to produce PEDU with high M_w (M_w = 109,000 g·mol⁻¹, Table 1).

The mechanism for the proposed lipase-catalyzed polymerization of hydroxy acid is basically the same as that for lactones [21,22]. Scheme III shows the proposed mechanism for lipase-catalyzed polycondensation of hydroxy acid-type monomer as a typical example. The initiation step is the formation of acyl-enzyme intermediate (EAM) via an enzyme-monomer complex. The propagation starts with nucleophilic attack of the EAM by the hydroxy group of the monomer, followed by successive nucleophilic attacks on the EAM by the terminal hydroxy group of the growing oligomer chain.

Some thermal properties of PEEU(x), PEU(y) and PEDU were measured using DSC, and the results are summarized in Table 2. T_g could not be obtained by DSC analysis under the test conditions in this report. In Table 2, PDEU, a poly(diester-urethane) with an unsymmetrical molecular structure, previously reported [13], is shown as a reference (entry 7). PEEU(x), PEU(y) and PEDU are polymers with homogenous molecular structures, since 2 and 4 contain a symmetrical diurethane moiety. Due to the symmetrical structure of the polymers, PEEU(x), PEU(y) and PEDU showed higher T_m, Young’s modulus and tensile strength compared to the previously reported poly(diester-urethane) (PDEU) [13].

Fusion enthalpy (ΔH_u) was also determined by DSC measurements and the fusion entropy (ΔS_u) was calculated from the T_m and ΔH_u values for PEEU(x) and PEU(y). The ΔH_u and ΔS_u values for PEEU(x) were 31.4–31.5 (kJ·mol⁻¹) and 87.4–89.2 (J·mol⁻¹·K⁻¹), and for PEU(y) were 26.9–28.7 (kJ·mol⁻¹) and 69.3–74.4 (J·mol⁻¹·K⁻¹), respectively (Appendix Figure A1). Ether-containing polymers [PEEU(x)] tended to show a lower T_m when compared to the corresponding methylene chain polymers [PEU(y)]. Since PEEU(x) had a higher ΔS_u (a parameter related to chain flexibility: an increase in ΔS_u indicates that the polymer is more flexible [23]) than PEU(y), the lower T_m was due to the greater flexibility of the ether moiety in the polymer chain (entries 1–5). The T_m was also affected by the content of the urethane moiety. T_m increased with the increasing urethane content, thus PEDU showed the highest T_m value among the poly(ester-urethane)s synthesized in this study.

PEEU(x) showed relatively lower values of Young’s modulus, elongation at break and tensile strength among the tested poly(ester-urethane)s. This is partially due to the flexibility of the ether moiety [24,25]. By extending the ether moiety (oxyethylene chain), the polymer became softer, but weaker. In contrast, PEU(y) and PEDU containing no ether moieties were harder and tougher than PEEU(x). It appeared that PEU(y) and PEDU were more similar to linear low-density polyethylene (LLDPE) [26–28] than some thermoplastic elastomers based on diphenylmethane diisocyanate (MDI) and polyols [29].

The static water contact angle is one of the factors affected by the polymer structure. The static water contact angle was measured by dropping distilled water on a glass slide coated with the polymers. The uncoated glass slide was used for the control (47.2°). PEEU(x) showed the lowest water contact angle (entries 1 and 2), indicating that the polymer surface is the most hydrophilic among the tested polymers. Since the composition of the urethane moiety is the same for PEEU(x) and PEU(y), the hydrophilicity is partially due to the ether moiety in the soft segment [24]. On the other hand, the water contact angles of poly(ester-urethane)s without the ether moiety were almost the same, between 83.8–85.9°, despite the length of the soft segment or the chemical structure of the polyurethane (entries 3–7).

The chemical recycling of diol-diacid-type PEEU and PEU was evaluated by their degradation into cyclic oligomers using a lipase in organic solvent. The concentration of the substrate is the decisive factor in the depolymerization, which is accelerated under dilute conditions (Scheme I). As a typical example, PEEU(6) was enzymatically degraded in anisole at 110 °C for 72 h in dilute conditions (3 mg·mL⁻¹). The polymer degraded quickly to afford oligomers and no unreacted polymer was detected by SEC (Figure 5a). Figure 6 (1a) shows the MALDI-TOF mass spectrum of the crude degradation products. Molecular masses (m/z) of [818.55 m + Na⁺] and [818.55m + K⁺] corresponded to the structure of cyclic(ester-ether-urethane) (6) [CEEU(6)] oligomer. The cyclic monomer [m = 1 in Figure 6 (1a)] and a smaller amount of the cyclic dimer (m = 2) were the main components. The linear ester-ether-urethane oligomer, with a molecular mass 18 units higher than that of the corresponding cyclic oligomer fraction, was barely detected. In order to verify the cyclic structure, the crude product was purified by reprecipitation using chloroform (good solvent) and ethanol (poor solvent). The ethanol soluble part was collected and fractionated by using a preparative SFC with scCO₂ and ethanol to obtain the molecularly pure CEEU(6) monomer (Appendix Figure A2). Figure 6 (1b) shows the MALDI-TOF mass spectrum of the isolated CEEU(6) monomer. The cyclic structure of the monomer was also confirmed by the absence in the ¹H NMR spectra of peaks at δ = 2.15 ppm corresponding to the terminal CH₂COOH moiety of 2 and at δ = 3.60 and 3.73 ppm corresponding to the terminal OCH₂CH₂OH moiety of the hexaethylene glycol moiety (Appendix Figure A3). The produced CEEU(6) monomer was repolymerized by CALB in a concentrated 1,2-dimethoxybenzene solution to produce PEEU(6), having similar M_w values as the initial polymer (Figure 5a). The chemical recycling of PEU(12) was also carried out by the same procedure as described for PEEU(6) (Figure 5b). These results confirm the chemical recyclability of PEEU(6) and PEU(12) by lipase.