E. C. Baker

Development of a pilot-plant process for the extraction of soy flakes with aqueous isopropyl alcohol

Soy flakes were extracted with aqueous isopropyl alcohol (IPA) at 77 C in a Kennedy countercurrent continuous extractor at a retention time of 71 min. IPA concentration was varied from 85.0 to 90.5% w/w and included the 87.7% IPA-water azeotrope. Solvent to meal ratios were varied from 1.5 to 3.0. The oil-IPA miscella leaving the extractor was chilled and coalesced to yield an oil phase and an IPA phase. The IPA phase was recycled to the extractor without being distilled. Excess IPA was expressed from the defatted flakes, and this also was recycled to the extractor. IPA recovered by distillation in the evaporator-stripper and desolventizer-toaster accounted for less than 10% of the total. Refined deodorized oils from the IPA extraction process compared favorably with their hexane counterpart in color, peroxide value and phosphorous and free fatty acid contents. Desolventized meals from the IPA process compared favorably with their hexane-extracted counterpart in protein, ash and fiber content.

Processing ofcrambe abyssinica seed in commercial extraction facilities

Crambe abyssinica seed was processed in four commercial oilseed crushing facilities, two utilizing prepress solvent extraction and two utilizing straight solvent extraction techniques. Mill capacities ranged from several T/day to 200 T/day. Crambe throughput in the larger facilities ranged from 30 to 150 T/day. Seed, press cake, flakes and finished meal samples were collected and analyzed during and following the runs. On-site testing included measurements of moisture, oil, glucosinolate and temperature and estimates of thioglucosidase enzyme activity. Three to 7 T of defatted meal were produced for each of four beef cattle feeding studies, and oil produced was blended into commercial erucic acid production streams. Except for one run, thioglucosidase inactivation had to be completed in the desolventizing/toasting (DT) unit, and the high temperatures required resulted in destruction of glucosinolate and the formation of aglucon product, 1-cyano-2-hydroxy-3-butene, and glucose in the finished meals. Protein solubility and lysine levels decreased with excessive heat. Regression analysis was used to examine some of the data for relationships between temperature, moisture, glucosinolate, nitrogen solubility and aglucon products. The results of these runs further demonstrate the feasibility of processing Crambe in commercial oilseed facilities.

Critical processing factors in desolventizing-toasting soybean meal for feed

Even though it is well established that both underheated and overheated meals are of inferior nutritive value, comparatively little is known of the fundamental nature of the changes brought about in the protein and how these correlate with the processing conditions during toasting. In the present study we examined the interrelation of several factors in the commercial desolventizing-toasting process for toasting soybean meal and determined how these relate to protein quality of the meal. A total of 48 test runs were made in the pilot plant from two cultivars of soybeans (one high and one low in protein) that were dehulled, flaked, and defatted in a continuous extractor using hexane. The solvent-wet flakes were desolventized and toasted under a variety of conditions. In a simulation of commercial operation, independent variables such as moisture, temperature and time of toasting were mathematically converted to equations for computer fitting of the data, which were used to predict several dependent measurements. Quality of the meal was improved by increasing heating time, jacket steam pressure and moisture content. Moisture level in the toasting operation was directly affected by the hexane level in the feed material to the toaster.

Crambe processing: Glucosinolate removal by water washing on a continuous filter

Abstract and SummaryCrambe seed was dehulled and screw pressed to remove approximately two-thirds of the oil, and then it was hexane-extracted to remove the rest. The defatted meal was toasted in the presence of moisture to form a crisped meal possessing fast drainage characteristics required for continuous filtration. The crisped meal was slurried with four parts of water, filtered, and washed on a continuous pilot-plant filter. Water washing removed about one-fourth of the meal solids, which contained 92-96% of the glucosinolates. Estimated processing costs for water-washing crambe meal are 22-23 dollars per ton of unwashed defatted meal, in addition to the cost of crushing the seed to oil and meal.

Biological evaluation of crambe meals detoxified by water extraction on a continuous filter.

Abstract and SummaryCrambe meals prepared by water extraction on a continuous filter when fed to rats gave protein efficiency ratios that were equal to or higher than the casein control, indicating that the water washing produced a palatable, nutritious meal. In a 4-week chick-feeding study, crambe was fed at 20% of the total diet. The diets containing crambe had somewhat lower gains (83-87% of control) and feed efficiency (94-95%) compared to the basal control group. Livers and kidneys appeared normal for all groups. There was some very slight gizzard erosion in the crambefed group. In a 90-day rat-feeding study, water-washed crambe was fed at 30% of the total diet, and body and organ weights were determined. Growth was slightly less than with the 30% soy control. There were no significant differences among relative organ weights for all groups. Results of feeding studies in rats and chicks indicate that the process of water extraction on a continuous filter can successfully prepare crambe meals with greatly reduced toxicity.

Sensory characteristics and oxidative stability of soybean oil and flour extracted with aqueous isopropyl alcohol

Soybean flakes extracted with hexane or aqueous isopropyl alcohol (85%, 87.7% and 90.5% IPA by weight) were processed to toasted flours and the miscellas to refined soybean oils. These products were evaluated for sensory characteristics and oxidative stability. Sensory analyses of initial oils and flours indicated good quality products. Initial flavor scores of IPA-extracted oils and flours were not significantly different from those of hexane-extracted oil and flour. Flour samples aged at 49 C for 1 mo and 37 C for 3 mo were rated slightly lower in flavor score than the initial flours. Flavor scores of oils decreased after aging but remained acceptable. Oils extracted with aqueous IPA concentrations of 85% and 90.5% received significantly lower scores than oils extracted with hexane or 87.7% IPA after 8 hr of fluorescent light exposure. Oxidative stability measured by the induction of weight increases of the oils during aging was similar. Residual solvent flavors were slightly detectable in unaged IPA flours and in those aged 3 mo at 37 C.

The preparation of soy products with different levels of native phytate for zinc bioavailability studies

Soy products with low, intermediate and normal phytate levels were prepared in the pilot plant for subsequent rat-feeding experiments to evaluate zinc bioavailability. The low level (0.29%) phytate product was made by precipitation of the protein curd at pH 5.5, whereas the normal level (1.05%) phytate product was produced by a similar process except that the phytate previously isolated from the whey fraction was added back to the original curd as native phytate. The intermediate level (0.73%) phytate product was also produced by acid precipitation, but at pH 4.5. The pH 5.5 precipitation process yielded a large quantity of whey in which the ratio of water content to phytate was over 1,000 parts to 1. However, ca. 75% of the water was subsequently removed by reverse osmosis (RO), which increased the concentration of phytate in the whey fraction and facilitated its isolation. Protein was first removed from the whey by precipitation with trichloracetic acid, then phytate was precipitated in the supernatant with ferric chloride. Another series of experiments was run to find optimal conditions to convert ferric phytate to the more soluble sodium phytate form, using a minimal amount of sodium hydroxide so that the phytate could be recycled back to the curd without causing a large increase in sodium content of the product. There were only minor differences in the protein, lipid and mineral contents of the three products.

Development of a pilot plant process for the preparation of a soy trypsin inhibitor concentrate

A pilot-plant procedure was developed to prepare a soy trypsin inhibitor (TI) concentrate in sufficient quantities to support a lifetime (2-yr) feeding trial in which diets containing varying amounts of TI would be fed to rats to assess the physiological effects on the pancreas and other organs. Starting with water dispersions of commercial defatted soy flour, separation of TI (MW<21,500) from non-TI protein (MW 180,000–350,000) by virtue of their MW difference was attempted using ultrafiltration techniques but was not successful. However, good separation was obtained when selective acid precipitation coupled with “salting in” of the TI with 0.1 N sodium chloride was employed. Low MW components were separated successfully by ultrafiltration using a 1,000 MW cutoff membrane. The final soy TI concentrate obtained by freeze drying exhibited a 9-fold increase in TI activity.

Desolventizing-toasting of extracted soybean flakes: Development of pilot plant equipment and operational procedure

A pilot plant batch desolventizer-toaster (D-T) was designed and built with the intent of producing soybean meals of varied composition, as well as to simulate meals produced in a continuous commercial D-T unit. Trial runs were made first to determine workable loading levels, temperature control and sparge steam generation. Moisture levels after the steam sparge were influenced by the residual hexane content of defatted hexane-wet flakes reaching the D-T. Two moisture levels were used in testing the effectiveness of the toasting operations in producing flakes with low urease activity and trypsin inhibitor levels. The trial runs reported here also provide basic data for current work designed to optimize toasting procedures to produce suitable meals for ongoing animal nutrition studies.