Tailoring Bacterial Cellulose Membranes via Plastic-Derived Carbon Sources: A Structural and Performance Characterization

Mixed polyethylene terephthalate (PET) and poly(lactic acid) (PLA) waste is a recognised dead end for mechanical recycling, since PLA is difficult to separate from PET and even minor contamination degrades recyclate quality. This study investigated the bioconversion of terephthalic acid (TPA), a degradation product of the petrochemical-derived plastic PET, and lactic acid (LA), the monomer of the bioplastic PLA, into bacterial cellulose (BC) by the bacterium Komagataeibacter xylinus DSM2004 under static cultivation. To the best of our knowledge, this is the first report of BC synthesis by K. xylinus from LA as sole carbon source under static cultivation, and the first assessment of a mixed TPA/LA feed that mimics depolymerised PET/PLA waste streams. Cultivation on LA resulted in the highest BC production (1.14 ± 0.02 g/L), whereas TPA supported lower BC production (0.62 ± 0.02 g/L), although it resulted in the highest substrate-based yield (0.41 ± 0.12 g/g). Co-feeding both monomers sustained BC production (1.08 ± 0.01 g/L), indicating that K. xylinus was able to maintain BC, with a production comparable to that obtained with LA alone. The carbon source alone acted as a lever over membrane properties: thickness varied from 96 ± 11 to 866 ± 126 µm, water holding capacity from 150 ± 9 to 247 ± 35 g_water_/g_BC_, CO₂ permeability from 159 ± 11 to 1602 ± 80 barrer and O₂ permeability from 35 ± 8 to 190 ± 10 barrer, all without any post-synthesis modification. Tensile strength was higher for LA-containing media (0.53–0.58 MPa) than for TPA alone (0.30 ± 0.06 MPa), while stiffness and elongation at break were statistically similar. These properties fall within the ranges reported for commercial wound-care materials, which are necessary but not sufficient criteria, as biocompatibility, cytotoxicity and antibacterial performance were outside the scope of this work.