Multi-objective optimization and modeling of microwave-infrared pretreatment on drying and quality characteristics of cannabis (Cannabis sativa L.) using response surface methodology and artificial neural network

Electromagnetic drying of cannabis is a fast and energy-efficient method, but prolonged exposure may impact product quality. The study aimed to explore short-time microwave-infrared (MI) pretreatment of cannabis before controlled environmental drying at 25 degrees C and 50% RH. Using a Box-Behnken design and response surface methodology (RSM), pretreatment time (2-5 min), infrared (75-225 W), and microwave (70-210 W) power were optimized to maximize drying rate and cannabinoid contents, with minimizing color change and energy consumption. Results showed that the drying rate, color changes and tetrahydrocannabinol (THC) of dried inflorescences increased significantly (p < 0.05), whereas the energy consumption and tetrahydrocannabinolic acid (THCA) decreased due to MI pretreatment, without affecting the total THC. The optimal parameters were determined to be 225 W infrared and 210 W microwave pretreatment for 3.36 min. Comparing to untreated cannabis drying, MI pretreatment of cannabis at optimized conditions and drying resulted in shorter drying time and lower moisture content, >65% energy savings, 43% reduction of terpenes and more porous microstructure. Artificial neural network (ANN) modeling with a 3-9-6 structure outperformed RSM in predicting the response variables. Overall, this study identified that short-time MI pretreatment improved cannabis drying efficiency and neutral cannabinoids, with ANN modeling offering accurate predictions.