RESUMO
BACKGROUND: Lipids are minor components of flours, but are major determinants of baking properties and end-product quality. To the best of our knowledge, there is no single solvent system currently known that efficiently extracts all non-starch lipids from all flours without the risk of chemical, mechanical or thermal damage. This paper compares nine ambient solvent systems (monophasic and biphasic) with varying polarities: Bligh and Dyer (BD); modified Bligh and Dyer using HCl (BDHCL); modified BD using NaCl (BDNaCl); methanol-chloroform-hexane (3:2:1, v/v); Hara and Radin (hexane-isopropanol, 3:2, v/v); water-saturated n-butanol; chloroform; methanol and hexane for their ability to extract total non-starch lipids (separated by lipid classes) from wheat flour (Triticum aestivum L.). Seven ambient extraction protocols were further compared for their ability to extract total non-starch lipids from three alternative samples: barley flour (Hordeum vulgare L.), maize starch (Zea mays L.) and tapioca starch (Manihot esculenta Crantz). RESULTS: For wheat flour the original BD method and those containing HCl or NaCl tended to extract the maximum lipid and a significant correlation between lipid extraction yield (especially the glycolipids and phospholipids) and the polarity of the solvent was observed. For the wider range of samples BD and BD HCl repeatedly offered the maximum extraction yield and using pooled standardized (by sample) data from all flours, total non-starch lipid extraction yield was positively correlated with solvent polarity (r = 0.5682, P < 0.05) and water ratio in the solvent mixture (r = 0.5299, P < 0.05). CONCLUSION: In general, BD-based methods showed better extraction yields compared to methods without the addition of water and, most interestingly, there was much greater method dependence of lipid yields in the starches when compared to the flour samples, which is due to the differences in lipid profiles between the two sample types (flours and starches).