A Deeper Learning approach in exploring the eventualities of solvent-free microwave-based extraction of lemongrass essential oil: Understanding milky emulsion, biomass temperature, first drop appearance, oil quality and biorefinery

The current research work is a deep learning initiative to explore the various eventualities of solvent-free microwave-based extraction (SFME) of lemongrass essential oil and strategize recovery of more than two distinct classes of fractions (volatile and non-volatile) in a single-step extraction. The model does not make use of any solvent and extraction is based on rupturing of oil glands due to internal thermal stress followed by draining of the oil under gravity. From the various microwave power level attempted (170 W–510 W), 255 W was found to be optimal, producing the highest Citral content of 68.32% in 56 min with a total yield of EO of 1.6%. Results were comparable to the Citral content of EO obtained from 4 h of SD (67.81%) with a total yield of 1.56%. Comparable antioxidant activity zones (SFME-22323 & SD-21782) were obtained as well. A zone of milky emulsion was reported when the temperature of biomass exceeded 90 °C during SFME which can adversely affect the oil quality. The advantage of steam distillation was in terms of saving time and making the biomass available for re-extraction of phenolic principles thus making the biomass available for the simultaneous extraction of volatile oil and non-volatile principles. [Display omitted] •Presenting solvent-free microwave extraction (SFME) as a green tool for the extraction of lemongrass essential oil.•Decoding the sequence of events occurring during SFME.•Investigating the biorefinery aspect of the biomass after SFME.•Unrevealing important correlations among various factors.•Presenting a complete operational profile of SFME.