Sustainable and recyclable super engineering thermoplastic from biorenewable monomer

Environmental and health concerns force the search for sustainable super engineering plastics (SEPs) that utilise bio-derived cyclic monomers, e.g. isosorbide instead of restricted petrochemicals. However, previously reported bio-derived thermosets or thermoplastics rarely offer thermal/mechanical properties, scalability, or recycling that match those of petrochemical SEPs. Here we use a phase transfer catalyst to synthesise an isosorbide-based polymer with a high molecular weight >100 kg mol−1, which is reproducible at a 1-kg-scale production. It is transparent and solvent/melt-processible for recycling, with a glass transition temperature of 212 °C, a tensile strength of 78 MPa, and a thermal expansion coefficient of 23.8 ppm K−1. Such a performance combination has not been reported before for bio-based thermoplastics, petrochemical SEPs, or thermosets. Interestingly, quantum chemical simulations show the alicyclic bicyclic ring structure of isosorbide imposes stronger geometric restraint to polymer chain than the aromatic group of bisphenol-A.


Supplementary Tables
Supplementary Figure 23. A jockey type-cutter for dog-bone shape specimen following ASTM D638.

Detailed information for quantum chemical simulation.
A cause of thermal expansion is well explained with an increase in the interatomic distances. 59 The bond length increase can be understood using an anharmonic potential energy curve (PEC) with explicit vibration energy levels. All molecules have vibration energy even at zero temperature, and every bond length oscillates with respect to its expectation value of position (green circles in Fig. 3a). Upon heating, the molecules obtain thermal energy, which is converted to molecular vibration energy, and this leads to the increased population in the vibrationally excited states (v ≥ 2). Due to the asymmetry of intermolecular potential energy, Theoretical models of 'repeating units' of each polymer are selected with an assumption that the CTE value of the polymer is not significantly different from the repeating unit's value. This is a reasonable assumption since the CTE value measures the relative change. The ground state geometries are obtained with density functional theory (DFT) using the B3LYP functional with 6-31G* basis sets. Based on the ground state geometry, each bond is elongated by 0.1% up to 2.7% for the SUPERBIO repeating unit and 2.3% for the BPA-SEP repeating unit, except the bonds involving hydrogen atoms. Each geometry is re-optimized to consider the relaxation effect from angle changes. The length of the repeating unit is measured between two non-hydrogen terminal atoms (left-end oxygen atom and rightend carbon atom). The ground state geometries of the SUPERBIO and BPA-SEP repeating units are described in Fig. 3b. All quantum chemical simulations were performed using Q-Chem 4.3. 60 Interestingly, a larger energy is required to elongate the SUPERBIO repeating unit compared with the BPA-SEP repeating unit to attain the same degree of geometric alternation (Fig. 3c). The potential energy along the relative length change of each repeating unit can be well described using a harmonic approximation within the relatively small distortion region. The second order polynomial fitting gives the coefficient of the square term, which represents the steepness of the potential wall as 94132 and 59999 for the SUPERBIO and BPA-SEP repeating units, respectively. As discussed above, this result implies that the SUPERBIO repeating unit features a larger vibration energy gap than that of the BPA-SEP repeating unit, which induces a smaller single molecular expansion value.
Considering that ISB consists of only single bonds, it was expected to require less energy to force bond lengthening compared with the BPA part, which contains benzene groups. However, the simulation outcome might indicate that the ring structure of ISB induces higher geometric restrain when it is forced to have The charge carrier transporting and green-coloured phosphorescence emitting layers were sequentially coated on the film by a thermal evaporator. Afterwards, cathode layers were formed into the letter shapes of KRICT by the thermal evaporator. Upon applying an electrical voltage of 8 V, the fabricated OLED device successfully emitted green light. Furthermore, the device maintained the light emission even when it was strongly bent.
Experimental, results, and discussion for Supplementary Figure 22.
The in vivo study was performed in laboratory animal center of Daegu-Gyeongbuk Medical Innovation Foundation (DGMIF, Korea) approved by the national institutional review board (IRB). Two Sprague-Dawley rats weighing 250-300 g were housed in a cage in a specific pathogen-free level-2 (SPF-2) facility; the cage was covered with soft bedding. The rats were freely accessible to food and water in a room at 23 °C and RH 50% at 8 AM and 8 PM. After a 7 d acclimation period, each rat was anesthetized with an intraperitoneal injection of 30 mg kg −1 of Zoletil and 10 mg kg −1 of Rompun, and the scalp was carefully incised. Each exposed calvarial was punctured to generate a critically sized defect using an 8 mm diameter trephine bur under sterile saline irrigation. The negative and positive controls were left as sham surgery and implanted of bioactive commercial collagen membrane (Bio-Gide, Geistlich, Switzerland), respectively. Whereas the experimental groups were filled with 70-μm-thick and 80-mm-long SUPERBIO film or 10 mg SUPERBIO granules (~0.5 mm in diameter). The incised periosteum and skin were closed in layers with 5-0 Vicryl sutures after the procedure. Then, the rats recovered for 12 w to confirm bioactive material-mediated tissue growth. Then, the rats were sacrificed using excessive CO2 inhalation to harvest calvarial tissue specimens and subsequently fixed using 10% neutralized formalin until analyses. The amount of newly formed tissue through the bioactivity of each specimen was measured after anesthesia after 12 w using microcomputed tomography (µCT, SkyScan 1172; Bruker-microCT, Kontich, Belgium). A total of 683 slices were imaged (10.89 mm per slice for a 590 ms exposure) for each specimen, and the pictures were analyzed and used to generate a 3D 2000 × 1048 pixel image using CT reconstruction software (NRecon v.1.4.4; Bruker-microCT). The regenerated bone area was quantified using an image analysis program (CTAn v.1.6.0; Bruker-microCT).
vivo bone regeneration experiments of rat calvarial defect model ( Supplementary Fig. 22). As an experimental group, a 8 mm diameter defect in a rat was plugged with SUPERBIO film or granule (several mm). Sham surgery and commercial collagen-plugged cases were negative and positive controls, respectively. After 12 weeks, the surgical sites were observed, and the defect sites were visualized using an X-ray microtomography. In the experimental groups of SUPERBIO, the surgical margins of skin fully healed up, and no noticeable pus, redness, and inflammation were observed. In the positive control group, the defect site was fully filled with regenerated bone. In contrast, the SUPERBIO film group exhibited the low healed bone volume as much as that of the negative control, and the granule group gave somewhat lower healed bone volume than the other groups probably because the granules freely floated with the movement of the rat and damaged the tissue. The film and granules of SUPERBIO were peeled easily from the tissues without adhesion, and the size and shape of the film remained unchanged.
This in vivo data indicates that SUPERBIO has bio-inertness of low interaction with tissues, which are important requirements of biomedical implants.