A glucose-fed E. coli strain matched a commercial hot-melt adhesive in a steel lap-shear test

Poly(4HB-co-PhLA) edged the commercial EVA control in the steel test. It is also the weaker producer: the terpolymer left the fermenter at about 5.2 times its titre ^[15]. On wood, the terpolymer performed comparably to EVA ^[6]. Both titres come from laboratory fed-batch runs, in which nutrients are supplied continuously through the cultivation ^[18]. The steel margin over the EVA control was about 9 percent ^[16]. That turns a chemistry question into a process-economics one, decided by how much of the glucose ends up as polymer.

Making several monomers in one cell creates metabolic imbalances. The fixes the KAIST group reports are incremental ones. They adjusted the strength and timing of gene expression and added a CoA transferase to help the linking reactions along. They then used a genome-scale metabolic model to find the steps where precursor supply ran short ^[9]. The strain takes up glucose, makes both 4HB and PhLA itself, and links them into polymer inside the cell ^[10].

4HB supplies softness and adhesion, PhLA supplies rigidity and heat resistance, and shifting the ratio moved both the thermal behaviour and the bond ^[7]. The band where adhesion was best is about ten mole points wide ^[19]. In a fermenter the culture sets that ratio, so I think hitting that band batch after batch is harder than raising titre.

EVA is the incumbent chemistry for hot melts, which is why it was the comparison ^[11]. Hot melts are applied molten and set as they cool, with no solvent and a fast bond ^[13]. On a line, that step depends on melt viscosity at the nozzle and on open time before the two substrates meet. A lap-shear coupon measures neither. The phys.org account reports lap-shear strength, thermal properties and adhesive strength retained after repeated melting, without naming the EVA grade or the number of melt cycles ^[14][17].