Abstract
Abstract
Polyethylene terephthalate (PET) waste was depolymerized by microwave-assisted alkaline hydrolysis in two separate reagent systems, one based on NaOH and one on KOH, with the aim of recovering terephthalic acid (TPA) and characterizing each system at the process level. The two systems were optimized independently, each within its own operating window, using a Taguchi L9 (3
4
) orthogonal array of four factors at three levels, with four replicates per run (36 runs per system). Because the two campaigns used different catalyst variables, TiO
2
loading in the NaOH system and catalyst type (none, zinc acetate, zinc oxide) in the KOH system, the results are reported as process-oriented characterizations of each system rather than as a catalyst-matched comparison of the two cations. Analysis of variance, signal-to-noise (
S
/
N
) analysis, main-effect analysis and general linear model (GLM) regression all identified temperature as the controlling factor in both systems, followed by alkali concentration and reaction time. PET conversion approached completion at 195 °C/30 min/12 wt.% NaOH and at 200 °C/36 min/18 wt.% KOH. The GLM models were predictive (
R
2
⩾ 98%; leave-one-out predicted
R
2
within 1.6% of
R
2
), and residual analysis supported the model assumptions. The main methodological contribution is a pair of severity-normalized descriptors, a performance index and an apparent rate index adapted from the combined severity factor of biomass pretreatment, which read efficiency per unit of applied thermal and temporal severity rather than by conversion alone. A preliminary process assessment quantified the alkali stoichiometric excess (≈4.3–4.6-fold) and the inorganic salt by-product stream, both of which identify reagent efficiency as the main obstacle to scale-up. TPA recovery was confirmed by Fourier-transform infrared spectroscopy and
1
H NMR, and product morphology was examined by scanning electron microscopy.