Reaction Network Database for Fast Pyrolysis of Vanillyl Alcohol Based on Molecular Wavefunction Descriptors

• • The database documents fast pyrolysis of vanillyl alcohol at 823.15 K, providing thermodynamic energy barriers and primary reaction pathways that enable quantitative prediction of product selectivity, addressing the 10%–20% phenolic yield bottleneck in industrial lignin pyrolysis. • • Atomic-level electronic fingerprints, including Fukui indices (f−, f+, f0) and condensed dual descriptors (CDD), are provided for key intermediates; for example, SR-1 shows CDD values ranging from −0.0598 to 0.0824, enabling identification of radical attack sites for catalyst design. • • Spin population analysis for radical SR-2 reveals that oxygen atom O9 carries 33.70% of total spin density, while carbon atoms C1 and C3 carry 21.17% and 18.35%, respectively, indicating preferential reaction sites for radical coupling and secondary reactions. • • The standardized DFT workflow and quality control ensure reproducibility, making the database suitable for feature engineering in machine-learning models for reaction prediction and inverse design of pyrolysis catalysts.