Meifang Cai

Objective: To establish a gas chromatographic method combined with computational modeling for the determination and prediction of hexamethylcyclotrisiloxane (D3) residues in lyophilized live attenuated varicella vaccine, aiming to enhance vaccine packaging safety and support consumer health risk assessment. Methods: Vaccine samples were reconstituted with water and extracted using ethyl acetate. Separation was performed on an Agilent HP-5 column (30 m × 0.32 mm × 0.25 μm) with nitrogen as carrier gas at 1.0 mL/min. Detection was conducted using a flame ionization detector (FID). In parallel, molecular dynamics simulations and machine learning models were employed to predict the migration behavior of D3 from halogenated butyl rubber stoppers into vaccine matrices. Results: D3 exhibited good linearity in the range of 1.0–10.0 μg/mL (r = 0.999). The repeatability RSD was 0.6% (n=6), and the average recovery was 99.2% with an RSD of 1.7% (n=9). The limit of quantification was 1.0 μg/mL (S/N > 10). Computational models successfully simulated the diffusion coefficients and interaction energies of D3 within the vaccine matrix, providing mechanistic insights into its migration potential. Conclusion: The proposed method, integrating experimental analysis with computational modeling, offers a simple, accurate, and reproducible approach for detecting D3 residues in lyophilized varicella vaccines. It provides a robust tool for evaluating packaging compatibility and advancing consumer health protection through predictive analytics.