Combining multilevel calculation protocol and QSAR model to predict the activity of phenolic antioxidants suitable for high-energy-density fuels at different temperatures

The thermal management of hypersonic vehicles critically depends on the oxidative stability of high-energy-density fuels used in regenerative cooling systems. While antioxidant addition represents an established approach to enhance this stability, the vast chemical space of potential antioxidants and their fuel-specific performance variations present significant challenges in identifying optimal candidates. This study introduces a quantum mechanics-driven computational framework that advances beyond traditional force field-based approaches by integrating GFNn-xTB semi-empirical methods with density functional theory (DFT) calculations. Unlike conventional single structure-based methodologies, our approach implements a comprehensive conformational sampling strategy, yielding more accurate predictions of two key performance indicators: antioxidative reaction rate and equilibrium constants. The advantages of this methodology become particularly evident when compared to traditional single structure-based calculations, demonstrating significant improvements in predictive accuracy. Building upon these computational results, we developed and validated robust quantitative structure-activity relationship (QSAR) models incorporating quantum chemical descriptors and temperature effects. These models provide reliable predictions of phenolic antioxidant activity across an extensive temperature range (25 °C–500 °C), offering valuable insights for rational antioxidant design. Our findings reveal that optimal antioxidant performance correlates strongly with electron-rich phenolic structures featuring steric hindrance.