Conditioning of oil-bearing materials for solvent extraction by extrusion

A new procedure, based on extrusion, was developed to prepare soybeans for direct solvent extraction. The procedure eliminates the need of controlled cracking, partial shell remotion, heating and flaking, requiring only an adequate communition to particles passing Sieve USS 14 followed by extrusion in a specially developed extruder. Extruded soybeans were obtained in the form of hard and porous agglomerates, without any dusting tendency, higher extraction, draining and percolation rates, lower solvent hold up and higher apparent density than the best possible flakes. Extruded soybeans were processed by solvent extraction in batch and continuous industrial installations, enabling substantial increases in capacity (up to 100%) and reduction in overall steam consumption (up to 30%) when compared with flakes processing. Extracted oil from extruded soybeans seems to show lower levels of gums and toasted meal at higher soluble protein levels for the same urease activity. Peanut press-cakes were preconditioned by extrusion before solvent extraction in batch-type installations, enabling a substantial reduction in residual oil for similar conditions with nonextruded press-cake. For existing plants, the new procedure allows a substantial increase in plant capacity and lower steam consumption. For new plants, the new procedure allows substantial capital savings in equipment and space. The new procedure enables soybean processing by direct solvent extraction in small batch-type installation, which may be very convenient for some underdeveloped areas.     

Analysis of the oil extraction process in soybeans: A new continuous procedure

Soybean oil extraction techniques were studied in which solvent was recirculated or pumped once through a suspension of soybean tissue. Both refractive index and ultraviolet absorbance were used to monitor the extraction continuously. Slices of soybean tissue showed rapid extraction from damaged tissue, followed by slow extraction from intact tissue. When soybean flakes were extracted, a continuously decreasing rate was noted. When solvent was forced through flakes, extraction was more rapid than when solvent was allowed to diffuse in and out of flakes. Reextraction of partially defatted flakes showed that the last soybean oil to be extracted was not inherently resistant to extraction. The adsorption of soybean oil to defatted flakes may account for slow removal of small quantities of oil at the end of the extraction.     

Origin of the nonhydratable soybean phosphatides: Whole beans or extraction?

Whole soybeans and flakes, tempered to normal (10%) and elevated (14%) moisture levels, were stored and then extracted under a variety of conditions in both the presence or absence of phospholipase activity. Crude oils were degummed, and the resulting nonhydratable phosphatide (NHP) content was determined. Extractions performed on flakes at ambient temperature or at the boiling point of hexane showed that at normal (10%) moisture levels the temperature of extraction had little effect on the magnitude of NHP formation; whereas at 14% moisture, considerably higher levels of NHP were observed at the higher extraction temperature. Studies performed with 10- and 14%-moisture whole beans stored at 40°C for extended periods, with or without inactivation of enzymes, showed that at normal 10% moisture levels little deterioration occurs after one week of storage; however, after four weeks considerable NHP is formed. At 14% moisture, NHP formation was rapid during the first week of storage, and complete destruction of the phospholipid occurred after four weeks’ storage at 40°C. The results of these experiments indicate that the adverse effects of storage conditions, excessive moisture levels and elevated temperatures cannot be overcome by inactivation of phospholipase D prior to solvent extraction of the flakes.     

1,1,2-trichloro-1,2,2-trifluoroethane for oilseed extractions: Demonstration of a viscosity effect

Extractions of full fat soy flake and meal were carried out at 5 to 70 C with hexane and 1,1,2-trichloro-l,2,2-trifluoroethane (FC-113) and their corresponding miscellas in order to evaluate FC-113 as a process solvent. In flake extraction, the rate of oil extraction by FC-113 was markedly improved with increasing temperature while extraction by hexane was relatively independent of temperature. In contrast, extraction of flake by the miscellas from both solvents and extraction of meal by fresh solvents gave similar results. A comparison of solvent properties indicates that the differences can be related to the viscosity dependent diffusion into the microporous flakes and suggests similar efficiencies for FC-113 and hexane in countercurrent flow extractors operating at elevated temperatures. Contribution No. 561 of the Research and Development Division, Jackson Laboratory, Organic Chemicals Department, E.I. du Pont de Nemours & Company, Wilmington, DE 19898     

Experiments with solvent extraction of glandless cottonseed and glanded cottonseed

Preliminary bench scale tests indicated that the rate of extraction of oil from glandless cottonseed flakes is about the same as for glanded flakes. Pilot plant experiments, using 20 lb. batches of flakes in baskets 8×8 in. in cross section, showed that the usual percolation rates can be used and will produce the same results with glandless flakes as with glanded. Solvents used were commercial hexane, nearly pure normal hexane, and a mixed solvent of acetone, hexane and water. In some runs raw flakes were extracted; in others, the meats were tempered by heating to various temperatures. Refine and bleach tests were run on the resulting oil. Colors were much lower than generally obtained with oil from glanded seed, most samples testing below one red Lovibond on a spectrophotometer. The meals from the extractions were used in rat feeding tests. The mixed solvent meal seemed to be a cut above the hexane-extracted meals in protein quality, showing up equivalent to a casein based diet. The meal from meats which had been heated to 230 F seemed very slightly inferior to those which had undergone less drastic heating. All glandless meals were much superior to a commercial prepress solvent meal which was run for the purpose of making comparisons. Deceased. So. Utiliz. Res. Dev. Div., ARS, USDA.     

Aqueous Extraction of Oil and Protein from Soybeans with Subcritical Water

Aqueous extraction using subcritical water is an environmentally friendly alternative to extracting oil and protein from oilseeds with flammable organic solvents. The effects of solids-to-liquid ratio (1:3.3–1:11.7), temperature (66–234 °C), and extraction time (13–47 min) were evaluated on the extraction of oil and protein from soybean flakes and from extruded soybeans flakes with subcritical water. A central composite design (2³) with three center points and six axial points was used. Subcritical water extractions were carried out in a 1-L high-pressure batch reactor with constant stirring (300 rpm) at 0.03–3.86 MPa. In general, oil extraction was greater for extruded soybean flakes than with soybean flakes. More complete oil extraction for extruded soybean flakes was achieved at around 150 °C and extraction was not affected by solids-to-liquid ratios over the range tested, while oil extraction from soybean flakes was more complete at 66 °C and low solids-to-liquid ratio (1:11.7). Protein extraction yields from flakes were generally greater than from extruded flakes. Protein extraction yields from extruded flakes increased as temperature increased and solids-to-liquid ratio decreased, while greater protein extraction yields from soybean flakes were achieved when using low temperatures and low solids-to-liquid ratio.     

Effect of equilibrium oil extraction on the chemical composition and sensory quality of soy flour and concentrates

Previous studies have shown that ambient-temperature equilibrium, hexane extraction of soy flour yielded the same amount of oil as was extracted from soy flakes by conventional high-temperature processing. The oil obtained at ambient temperatures contained less phospholipid than commercial crude oils obtained by traditional processing. In this study, chemical composition, flavor and odor of soy flour obtained after oil extraction by the equilibrium procedure were evaluated before and after toasting. Results were compared with those obtained for commercial untoasted food-grade soy flakes. Chemical and sensory analyses were performed on soy protein concentrates (SPC) prepared from defatted flour, defatted toasted flour and commercial defatted white food-grade flakes. SPC were made by acid and ethanol-extraction methods. Ethanol extraction of soy flour produced SPC with similar protein, lipid and sensory qualities to those obtained from commercial flakes. Acid extraction produced SPC with more lipid than was obtained by ethanol extraction. Toasted soy flour and flakes had similar sensory properties, as did the SPC prepared from them.     

Relationships between moisture contents of cottonseed flakes and water in miscellas from extractions with acetone-cyclohexane azeotropes

A method for determining the composition of acetone, hexane and water mixtures was adapted for use with acetone, cyclohexane and water mixtures. Techniques for drying cottonseed flakes to different moisture contents for extraction with acetone-cyclohexane azeotropes were devised. Tests for possible gossypol binding occurring during the drying operation were made. Bench scale extractions yielded data which show the water resulting in miscellas when extracting cottonseed flakes at various moisture levels. It is envisioned that in a commercial process utilizing an acetone-cyclohexane-water mixture to produce cottonseed protein products for food uses, the water in the extracting mixture would be stablized at a level somewhere between the 0 and 3% by wt water levels of the two azeotropes used in these studies. Efforts are currently being devoted to defining the optimum practical water level that could be easily attained in commercial usage. Additional products evaluations will also be made. A laboratory of the Cotton Research Committee of Texas operated by the Texas Engineering Experiment Station.     

Cottonseed extraction with mixtures of acetone and hexane

Cottonseed flakes were extracted with mixtures of n-hexane and acetone, with the concentration of acetone varying between 10 and 75%. Adding small amounts of acetone (≤25%) to n-hexane significantly increased the extraction of free and total gossypol from cottonseed flakes. Sensory testing detected no difference in the odor of cottonseed meals produced either by extraction with 100% n-hexane or by extraction with a 10∶90 (vol/vol) mixture of acetone/hexane. More than 80% of the free gossypol was removed by the 10∶90 mixture of acetone/hexane, whereas pure n-hexane extracted about 47% of the free gossypol from cottonseed flakes. A solvent mixture containing 25% acetone removed nearly 90% of the free gossypol that was removable by extraction with pure acetone; the residual meal had only a minimal increase in odor. In contrast, cottonseed meals produced by extraction with pure acetone had a much higher odor intensity. The composition of the cottonseed crude oil was insignificantly affected by the acetone concentration of the extraction solvent. The results indicate that mixtures of acetone and n-hexane can be used as extraction solvents to produce cottonseed crude oil without the concomitant development of odorous meals.     

Solubilities of cottonseed oil and the phase distribution of fatty acids in methyl alcohol at elevated temperature and pressure

A study was made to determine whether methyl alcohol could be used for the selective extraction of fatty acids from crude cottonseed oil under high temperature and high pressure conditions. Equilibrium data and phase relationships were obtained at several temperatures, and corresponding triangular phase diagrams were prepared. From these, it was concluded that the proposed process would not be commercially attractive inasmuch as excessive amounts of solvent would be needed. The research described in this article was conducted in cooperation with the Cotton Research Committee of Texas.     

Immersion vs. percolation in the extraction of oil from oleaginous seeds

The influence of the two contact modes, percolation and immersion, during the extraction of oilseeds by means of a solvent are presented. Experiments were performed in lab-scale equipment with soybean flakes, arranged in beds that reproduce these two contact modes. The extractions were carried out with hexane at constant temperature. To simulate the performance of shallow-and deep-bed extractors, two different bed height/diameter ratios were used. The experimental results are explained in terms of the basic transfer phenomena that occur during extraction. These phenomena are addressed to develop a mathematical model, which is used to simulate extraction under both contact modes. The immersion scheme yielded greater efficiency than the percolation mode to extract soybean flakes for the two bed height/diameter ratios studied. The mathematical model predicts very well the experimental findings. It also predicts the solvent retained by the solid mass after extracting the oil.     

Measurement of extractability of raw and cooked cottonseed flakes

A batch co-current laboratory method for measuring comparative extraction rates and extraction efficiencies of oleaginous materials in solvent is described. The method, a modification of that by Winward and Shand, was carefully tested with raw and cooked cottonseed flakes of various thicknesses and in various hexane miscella concentrations. It enables measurement of intrinsic extraction rates and extractabilities of materials, unaffected by diffusional effects in the liquid medium, and yields accurate and concordant results even with extracting miscellas of considerably high concentration. It is equally applicable for evaluating and predicting the effect upon extractability of different material preparation operations, particle sizes, moisture contents, temperature, solvents, etc. The method was used in this investigation to compare the rate and degree of extraction under the specified testing conditions of raw and cooked cotton-seed flakes of .005-in., .015-in., and .025-in. thicknesses in miscella concentration of 0%, 25% and 50% oil. The results may be summed up as follows:

1. the extractability of both raw and cooked flakes in each of the miscella concentrations decreases as the flake thickness increases.
2. the cooked material prepared from the medium and thick flakes extracted at a more rapid rate and to a greater degree in all miscella concentrations than the raw flakes of comparative thicknesses, but the rate and degree of extraction were about equal for the very thin flakes.
3. the effect of increasing miscella concentration for both the raw and the cooked flakes of medium and thick sizes was to slow down the initial extraction rate; but for the very thin flakes the effect was negligible.
4. the effect of increasing miscella concentration in extracting the cooked material, regardless of flake thickness, was to increase the degree of extraction. For the raw flakes the effect was to increase the degree of extraction only of the very thin flakes.

     

Solvent extraction of oil from soybean flour I— extraction rate,a countercurrent extraction system,and oil quality

Measurements of rates of oil extraction from either fine flour or soybean flakes in a column showed that oil extraction from flour was dependent on the volume of solvent, but oil extraction from flakes depended on time of contact rather than volume of solvent. We interpreted the data to mean that oil was being washed out of the fine flour with little diffusion involved, whereas in flakes, the limit on rate was diffusion of the solvent into and out of the tissue. Fine full fat flour worked well in a batchwise countercurrent extraction system with mixing and centrifugal separation. Because the oil dissolved immediately and reached equilibrium rapidly, the actual material balance was close to calculated values. However, due to the large hold-up volume, the separation of miscella from the meal required several mixing and separation stages. The oil resulting from this countercurrent extraction system had a superior quality with 37 ppm phosphorus, 0.08% free fatty acids, and a light color.     

Extraction of oil from meadowfoam flakes

As part of a program to improve meadowfoam seed processing, the authors examined the effects of seed moisture, seed temperature, and flaking roll opening on oil extraction efficiency in meadowfoam flakes. Flakes were prepared using a Wolf Mill with dual horizontal, unheated 12-in. diameter rolls. Roll openings of 0.005, 0.013, and 0.020 in. (0.127, 0.330, and 0.508 mm, respectively) gave average flake thicknesses of 0.013, 0.021, and 0.031 in., respectively (0.330, 0.533, and 0.787 mm). Seed moistures of 9, 12, and 15% and seed temperatures of 65, 190, and 210°F (18, 88, and 99°C) chosen for flaking were known to provide a range of conditions suitable for enzyme inactivation during seed cooking prior to flaking. Experimental flakes were examined for extractable oil content (petroleum ether extraction); this was compared to total oil content (31.5%) determined on finely ground flakes. Roll opening was the dominant variable determining flake thickness, the primary parameter affecting oil extraction efficiency. Thus, the thinnest flakes at 0.013 in. were only slightly less extractable (29.8%) than finely ground flakes (31.5%), but intermediate (0.021 in.) and thick (0.031 in.) flakes were significantly less extractable (28.0 and 26.0%, respectively). There was a slight but significant (P<0.01) trend toward thicker flakes with increasing seed moisture (15>12>9%) during flaking. A similar trend to thicker flakes with increasing temperature was significant (P<0.01) only for the thickest flakes produced at the largest roll opening (0.020 in.). Lower seed moisture and higher seed temperature significantly impacted extractable oil content of the thickest flakes, but negligibly affected extractability of the thinnest flakes. The authors conclude that meadowfoam flakes must be as thin as possible (e.g., <0.015 in.) for efficient oil extraction. Further, seed cooking temperatures >190°F at moistures >10% and <15% that are adequate for efficient enzyme inactivation in the whole seed are also suitable for seed flaking.     

Extraction of Phospholipids from Egg Yolk Flakes Using Aqueous Alcohols

Two alcohols, ethanol and butanol, with different water contents were evaluated for phospholipids (PL) sequential extraction from drum dried egg yolk flakes. It showed that butanol was more effective in extracting total yolk lipids compared to ethanol, but the PL in the extract had the same concentration as in the original yolk total lipid. The use of aqueous ethanol of 95 and 75% resulted in lipid extracts with higher PL concentration during the initial stages of the sequential extraction. When ethanol was further diluted to a concentration of 55%, the solvent lost its PL extraction ability, and the total lipid recovery also decreased dramatically. When both the PL purity and recovery were considered, 75% ethanol was the most effective aqueous alcohol for PL extraction and enrichment from the yolk flakes. In the first stage of extraction using such a solvent, 67% of the total PL in the original yolk was recovered in a lipid fraction with a PL purity of 75%. This study identified the optimal ethanol concentration for PL extraction from dried egg yolk. With this information, the best solid:solvent ratio can be designed to extract and enrich the polar lipids from lipid-bearing materials with known moisture content using a renewable or “green” solvent, ethanol.     

Effect of processing on sphingolipid content in soybean products

Soybeans are believed to be a rich source of sphingolipids, a class of polar lipids that has received attention for their possible cancer-inhibiting activities. The effect of processing on the sphingolipid content of various soybean products has not been determined. Glucosylceramide (GlcCer), the major sphingolipid type in soybeans, was measured in several processed soybean products to illustrate which product(s) GlcCer is partitioned into during processing and where it is lost. Whole soybeans were processed into full-fat flakes, from which crude oil was extracted. Crude oil was refined by conventional methods, and defatted soy flakes were further processed into alcohol-washed and acid-washed soy protein concentrates (SPC) and soy protein isolates (SPI) by laboratory-scale methods that simulated industrial practices. GlcCer was isolated from the samples by solvent extraction, solvent partition, and TLC and was quantified by HPLC. GlcCer remained mostly within the defatted soy flakes (91%) rather than in the oil (9%) after oil extraction. Only 52, 42, and 26% of GlcCer from defatted soy flakes was recovered in the acid-washed SPC, alcohol-washed SPC, and SPI products, respectively. All protein products had a similar GlcCer concentration of about 281 nmol/g (dry wt basis). The minor quantity of GlcCer in the crude oil was almost completely removed by water degumming.     

Factors promoting the formation of nonhydratable soybean phosphatides

Whole, cracked and flaked soybeans were stored under a variety of conditions. After extraction with hexane, the crude oils were degummed in the laboratory, and the nonhydratable phospholipid (NHP) content was estimated from the phosphorus content of the degummed oil. Results showed that four interrelated factors promote NHP formation. These include (i) moisture content of beans or flakes entering the extraction process; (ii) phospholipase D activity; (iii) heat applied to beans or flakes prior to, and during, extraction; (iv) disruption of the cellular structure by cracking and/or flaking. Results from this study suggest that NHP formation can be minimized by control of the moisture of beans and/or flakes entering the extraction process, inactivation of phospholipase D enzyme, and optimizing temperatures during the conditioning of cracked beans or flakes. Presented at the 82nd Annual Meeting of the American Oil Chemists’ Society, Chicago, IL, May 12–16, 1991.     

Flaking and extrusion as mechanical treatments for enzyme-assisted aqueous extraction of oil from soybeans

Flaking and extruding dehulled soybeans were evaluated as a means of enhancing oil extraction efficiency during enzyme-assisted aqueous processing of soybeans. Cellulase, protease, and their combination were evaluated for effectiveness in achieving high oil extraction recovery from extruded flakes. Aqueous extraction of extruded full-fat soy flakes gave 68% recovery of the total available oil without using enzymes. A 0.5% wt/wt protease treatment after flaking and extruding dehulled soybeans increased oil extraction recovery to 88% of the total available oil. Flaking and extruding enhanced protease hydrolysis of proteins freeing more oil. Treating extruded flakes with cellulase, however, did not enhance oil extraction either alone or in combination with protease. Discrepancies in oil extraction recoveries were encountered when merely considering crude free fat because some oil became bound to denatured protein during extrusion and/or sample drying. Bound fat was unavailable for determination by using the hexane extraction method, but was accounted for by using the acid hydrolysis method for total oil determination. Oil extraction recovery from extruded soybean flakes was affected by oil determination methods, which was not the case for unextruded full-fat soy flour.     

Comparative absolute yields of crude and neutral oils from variously prepared cottonseed meats

Summary Experiments utilizing cottonseed meats of diverse origin and composition were conducted for the purpose of determining the effect of the method of meats preparation on the yields of crude and neutral oil obtainable from differently prepared, comparable meats by solvent extraction. Three methods of meats preparation were employed,i. e., simple flaking of raw meats “as is”, tempering of cracked meats prior to flaking, and cooking by the modified hydraulic method developed for use with the filtration-extraction process. Commercial hexane was used as the extraction solvent. The experiments were carried out by procedures which eliminated the effects of any variables other than the method of preparing the meats for extraction The results of the studies showed that the method used in preparing cottonseed meats for extraction had a significant effect on the yields of crude oil obtained but that the yields of neutral oil, the valuable constituent of crude oils, were virtually unaffected. Analyses of the crude oils showed that the differences in crude oil yields were caused by the relative amounts of non-neutral oil materials in the crudes from the differently prepared meats. The greatest yields of crude oil were obtained from raw flakes, intermediate yields from tempered flakes, and the smallest yields from cooked flakes. The impurities content in the respective crude oils followed the same order,i. e., crudes from raw flakes were highest in impurities and lowest in neutral oil, crudes from tempered flakes were lower in impurities and higher in neutral oil, and the crudes from the cooked meats were outstandingly low in impurities and high in neutral oil. Virtually equal amounts of neutral oil were obtained from equivalent quantities of comparable meats regardless of the method used in preparing the meats for extraction. Presented at the meeting of the American Oil Chemists’ Society, Chicago, Ill., September 24–26, 1956. One of the laboratories of the Southern Utilization Research Branch, Agricultural Research Service, U.S. Department of Agriculture.     

Water accumulation in the alcohol extraction of cottonseed

The critical moisture content of cottonseed flakes extracted with an aqueous alcohol solvent can be defined as that flake moisture level at which the flakes lose no moisture during extraction. This study shows that the critical moisture content for aqueous ethanol (92%, w/w) is 3%. For aqueous isopropanol (88%, w/w) this value is 6%. If the moisture contents of the flakes are above these levels, then the solvents pick up moisture. For moisture contents below this level, the flakes adsorb moisture and actually dry the alcohol. It is proposed that this latter capability can be used as a basis for a method to control water accumulation in aqueous alcohol solvent extractions.