Supercritical carbon dioxide extraction of oils from Andean lupin beans: Lab‐scale performance, process scale‐up, and economic evaluation

Abstract

Oil extraction from Andean lupin beans (Lupinus mutabilis SWEET) via supercritical carbon dioxide (scCO 2 ) was studied on both lab scale and pilot scale. On the lab scale, the effect of pressure, solvent-to-feed ratio (S/F), sample particle size and temperature on oil yield were evaluated. The oil quality (fatty acid [FA] composition and tocopherol content) were investigated. Five-hour scCO 2 extraction yielded about 86% oil of Soxhlet extraction (using hexane as solvent). The fraction of unsaturated FA rose with extraction pressure at specific time. High tocopherol contents were detected in oils extracted at low pressure. An increase in temperature was unfavorable to oil and tocopherol yield, thereby confirming the validity for preserving oil extract quality under a mild scCO 2 extraction condition. Oil quality and yield did not have identical optimum settings, opening up possibilities for producing different qualities of oils. Pilot-scale extraction offered comparable oil yield to lab-scale extraction at similar S/F ratio. Economic evaluation showed that it is promising to implement industrial scale scCO 2 process for lupin oil extraction. It was predicted that, at a specific industrial scale of extraction (2 Â 1000 L, 550 bar, 40 C and S/F of 24), the manufacturing cost of oils got close to actual commercial production cost. Practical ApplicationThis study applied an environment-friendly high-pressure extraction method, supercritical carbon dioxide (scCO 2 ) extraction, to separate oils from Andean lupin beans (Lupinus mutabilis SWEET). ScCO 2 extraction can serve as an alternative oil extraction method to conventional ones that use fossil-derived organic solvents as the extractant. Up-scaled scCO 2 processing was estimated to be economically viable for commercial lupin oil production.

Lipid extraction of high‐moisture sour cherry (Prunus cerasus L.) stones by supercritical carbon dioxide

“…Also investigated was whether the extraction process favors certain lipid compositions, which means that different FAME compositions were observed at the beginning of the experiment from those at a later point of the experiment. However, such behavior was not found; in contrast to other matrices, 31,36 the FAME composition kept the same proportions over the total extraction time (see supporting information, S1). Other FAMEs were detected but were not quantified, due to their limited presence.…”

“…An amount of 50 g of cherry material was measured and placed in the extraction vessel of the unit. After 30 min temperature equilibrium time, the experiments were carried out with a CO 2 flow rate of 25 g min −1 for 6 h. Unless otherwise mentioned, the extraction was performed with a pressure of 350 bar and a temperature of 40 °C, which showed good performance in preliminary testing and other biomatrices 20,21,31 . The extract was collected every 30 min from the collection vessel (60 bar, 40 °C) and the carbon dioxide was continuously recycled.…”

“…The FAME determination was based on transesterification as described by Lepage and Roy 32 and adapted according to Yu et al 31 . The internal standard (IS) solution was prepared by dissolving 40 mg of tetradecane (99%; Acros Organics, Geel, Belgium), 40 mg of methyl nonanoate (98%; Sigma Aldrich, St Louis, USA) and 40 mg of nonadecanoic acid (≥98%; Sigma Aldrich) in 10 mL of n ‐hexane (>99%; Acros Organics).…”

“…The FAME determination was based on transesterification as described by Lepage and Roy 32 and adapted according to Yu et al 31 The internal standard (IS) solution was prepared by dissolving 40 mg of tetradecane (99%; Acros Organics, Geel, Belgium), 40 mg of methyl nonanoate (98%; Sigma Aldrich, St Louis, USA) and 40 mg of nonadecanoic acid (≥98%; Sigma Aldrich) in 10 mL of nhexane (>99%; Acros Organics). Additionally, a FAME solution containing 20 mg of methyl palmitate (≥99%; Sigma Aldrich), 150 mg of methyl linoleate (≥99%; Sigma Aldrich), 120 mg of methyl oleate (Sigma Aldrich) and 10 mg of methyl stearate (99%; Sigma Aldrich) in 50 mL of n-hexane was prepared.…”

“…The saturated shape results probably through the limited mass transfer of lipids from the pores of the biomatrix to the scCO 2 phase, which was also reported elsewhere 20 and is in agreement with results for other lipid-containing matrices. 31 While the yield limitation for the lipid extractions for drier samples might result from a limited diffusion process (SK-RH20, SK-D40), the limitation for samples with higher moisture content (SK-RH75, SK-Original) might result from blockage of pores through water or inhibited partition of lipids of the scCO 2 phase. An increased moisture content showed therefore smaller lipid extraction rate per fresh and dry weight, which is not related to the maximum solubility of lipids inside the scCO 2 stream as shown in Fig.…”