Isoconversional kinetics, mechanical property loss and quantum computing prediction of kinetic parameters for naturally aged NBR

Nitrile butadiene rubber (NBR) stored for 2 years under natural ambient conditions (28 ± 3°C, 55 ± 10% RH) was compared with fresh NBR by thermogravimetry at 5, 10 and 15°C min ^‒1 under nitrogen and by tensile, hardness and density measurements. After separating the main backbone-decomposition stage from overlapping processes, Friedman and Vyazovkin analyses gave mean activation energies (α = 0.40–0.90) of 274 and 242 kJ mol ^‒1 for fresh and aged NBR, and the Kissinger method gave 251 ± 12 and 201 ± 3 kJ mol ^‒1 . The best-fit reaction model changed from three-dimensional diffusion (D3) to first order (F1), and tensile strength and modulus decreased by 41–46%. Quantum amplitude estimation, quantum kernel ridge regression and a quantum neural network, trained on the two measured states, were used to forecast the kinetic parameters at 1, 3 and 5 yr and compared with linear interpolation. All methods agree at 1 yr (257–258 kJ mol ^‒1 ) but diverge beyond 2 yr (195–263 kJ mol ^‒1 ) within propagated uncertainties of ± 62–110 kJ mol ^‒1 .