Effect of autoxidation prior to deodorization on oxidative and flavor stability of soybean oil

Summary Oxidation prior to deodorization was shown to be detrimental to the flavor and oxidative stability of soybean oil. The increase in the nonvolatile carbonyl content of freshly deodorized oils was proportional to the peroxide value of the oils before deodorization. Rate of loss of flavor and oxidative stability of the oil were related to the extent of carbonyl development. All oils, whether or not they had been submitted to any known oxidation, contained some nonvolatile carbonyls. The loss in stability was not due to a loss of the antioxidant tocopherol. Oxidized soybean oil methyl esters were shown to develop nonvolatile carbonyl compounds upon heating at deodorization temperatures. The addition of isolated methyl ester peroxide decomposition products to deodorized soybean oil reduced its flavor and oxidative stability in proportion to the amount added. The results obtained were parallel and similar to those obtained by oxidizing soybean oil prior to deodorization. Flavor deterioration and undesirable flavors were typical of aging soybean oil whether or not the oils were oxidized before deodorization or whether an equivalent amount of nonvolatile thermal decomposition products was added to the oil. These oxidatively derived, nonvolatile carbonyl materials are believed to enter into the sequence of reactions that contribute to flavor instability and quality deterioration of soybean oil. The structure of these materials is not know. This work indicates the importance of minimizing autoxidation in soybean oil particularly before deodorization to insure good oxidative and flavor stability. Presented at fall meeting, American Oil Chemists’ Society, October 20–22, 1958, Chicago, Ill. This is a laboratory of the Northern Utilization Research and Development Division, Agricultural Research Service, U. S. Department of Agriculture.     

A review of soybean oil reversion flavor

Crude soybean oil has a characteristic “greenbeany” flavor, which during refining, bleaching and deodorization is eliminated to produce a bland tasting, light colored oil. However, flavor returns during storage and has been characteristically called the “reversion flavor” of soybean oil. This deleterious characteristics flavor has influenced the utilization of soybean oil and its fatty acids. Several theories for the cause of reversion flavor include: (a) oxidation of linolenic acid; (b) oxidation of isolinoleic acid of the 9,15-diene structure; (c) phosphatide reactions; (d) unsaponifiables; and (e) oxidative polymers. References are presented that support or contradict these theories. Recent publications concerning the isolation and characterization of the components of reversion flavor indicate slight oxidation of the fatty acids is the major cause. Techniques that are effective in increasing the flavor stability of soybean oil are presented.     

Tocopherol Content and Profile of Soybean: Genotypic Variability and Correlation Studies

Plant seed oils, including soybean seed oil, represent the major source of naturally derived tocopherols, the antioxidant molecules that act as free radical quenchers preventing lipid peroxidation in biological systems and vegetable oil products. All four isomers of tocopherols, i.e. α, β, γ, δ tocopherols that exist in nature are found in soybean seeds. The biological activity and the contribution of these isomers in improving the oxidative stability of vegetable oil are in reverse order. Because of the nutritive value and the importance for oil stability, enhancement of tocopherol content, through breeding programs, in soybean seeds has become a new and an important objective. Genotypic variability, which is the basis of every breeding program, is scarcely reported for tocopherol content and profile in soybean seeds. In the present investigation, the tocopherol content and profile in seed samples of 66 genotypes of Indian soybean were determined. The ratios observed between the lowest and the highest values for α, β, γ, δ, total tocopherol content were 1:13.6, 1:10.4, 1:7.5, 1:9.1, 1:7.9, respectively. The mean contents for α, β, γ, δ and total tocopherols were 269, 40, 855, 241 and 1,405 μg/g of oil, respectively. Total tocopherol content was the highest in ‘Co Soya2’ followed by ‘Ankur’. Concentration of α-tocopherol was the highest (27%) in ‘Ankur’ followed by ‘MACS124’ (26%) whereas gamma tocopherol concentration was the highest (69%) in ‘VLS1’ and ‘PK327’ followed by ‘MACS13’ (67%). In view of the fact that levels of unsaturated fatty acids, apart from tocopherols, also determine the oxidative stability of vegetable oils, the relationship of four isomers of tocopherols with each other as well as with different unsaturated fatty acids and oil content was also investigated in the present study. All the four isomers of tocopherols exhibited highly significant correlations with each other (p < 0.001) whereas γ-tocopherol and total tocopherol content showed a significant relationship with linoleic acid (p < 0.05).     

Oxidative and flavor stability of oil from lipoxygenase-free soybeans

Soybeans that lack or contain three lipoxygenase (LOX) isozymes, LOX-1, LOX-2, and LOX-3, were evaluated for oxidative and flavor stability at 60°C in the dark and at 35°C in the light. Although the two types of soybeans had a similar genetic background, there were significant differences (P ≤ 0.01) in fatty acid percentages between the lipoxygenase-free and normal oils before and after storage at both temperatures. The linolenic acid content of oil from LOX-free germplasm before storage averaged 7.2%, while normal lines averaged 6.6%. The linoleic acid content after storage averaged 6.9% for LOX-free and 6.6% for normal oils. LOX-free oil was not significantly different from normal oil in flavor, as judged by a sensory panel, or in concentrations of volatiles during storage at either storage condition. LOX-free oil had less hexanal than normal oil before storage, but had significantly greater (P ≤ 0.05) levels after storage for two weeks at 35°C. Peroxide values of oil from LOX-free soybeans were significantly greater (P ≤ 0.01) than oil from the normal soybean after storage at 60 and 35°C. LOX-free oil had significantly greater (P ≤ 0.01) levels of α-, β-, and γ-tocopherols. In general, oil from LOX-free soybeans did not have improved flavor or oxidative stability. Differences between the two oil types in peroxide value and in production of a few volatiles were probably a result of the differences in initial fatty acid composition.     

Oxidative and flavor stability of oil from lipoxygenase-free soybeans

Soybeans that lack or contain three lipoxygenase (LOX) isozymes, LOX-1, LOX-2, and LOX-3, were evaluated for oxidative and flavor stability at 60°C in the dark and at 35°C in the light. Although the two types of soybeans had a similar genetic background, there were significant differences (P≤0.01) in fatty acid percentages between the lipoxygenase-free and normal oils before and after storage at both temperatures. The linolenic acid content of oil from LOX-free germplasm before storage averaged 7.2%, while normal lines averaged 6.6%. The linoleic acid content after storage averaged 6.9% for LOX-free and 6.6% for normal oils. LOX-free oil was not significantly different from normal oil in flavor, as judged by a sensory panel, or in concentrations of volatiles during storage at either storage condition. LOX-free oil had less hexanal than normal oil before storage, but had significantly greater (P≤0.05) levels after storage for two weeks at 35°C. Peroxide values of oil from LOX-free soybeans were significantly greater (P≤0.01) than oil from the normal soybean after storage at 60 and 35°C. LOX-free oil had significantly greater (P≤0.01) levels of α-, β-, and γ-tocopherols. In general, oil from LOX-free soybeans did not have improved flavor or oxidative stability. Differences between the two oil types in peroxide value and in production of a few volatiles were probably a result of the differences in initial fatty acid composition.     

Soybean oil in dried egg mixes

Dried egg mixes prepared commercially with hydrogenated-winterized soybean, corn, or cottonseed oils were evaluated for initial flavor and for flavor storage stability. Quality evaluations were made on products from two processing plants; flavor, color, stability, and mix volumes were determined periodically during storage at 100 F for 1 year. All mixes contained 15% of the specified oil and were air-packaged in 6 oz laminated foil pouches. Replicated triangle flavor tests on reconstituted dried eggs (scrambled) indicated that neither an analytical-type taste panel nor a palatability panel could distinguish between the mixes containing the different vegetable oils. All samples, regardless of oil component, deteriorated at ca. the same rate when stored at elevated temperatures. Minor differences in flavor scores, color indices, and mix volumes were noted in samples stored at 100 F for 9 or 12 months. A dried egg mix made with hydrogenated-winterized soybean oil could not be distiguished, after 4 months’ aging at 100 F, from a fresh (unaged) mix made with corn oil. After 6 months’ storage at 100 F all aged mixes, regardless of the vegetable oil used in their preparation, could be distinguished from the fresh corn oil mix. Presented at AOCS Meeting, Mexico City, April 1974.     

Components of the hardening flavor present in hardened linseed oil and soybean oil

The characteristic hardening flavor which develops in hardened linseed and soybean oils during storage has been coned from hardened linseed oil by stripping with nitrogen. After separating the volatile substances by adsorption chromatography on silica, the fraction containing the hardening flavor has been converted into 2,4-dinitrophenylhydrazones (DNPHs) and separated by means of partition chromatography. On regeneration of the fractions of DNPSs obtained, the characteristic hardening flavor was observed in one specific band. Both by hydrogenation and by oxidation of the free carbonyl the carrier of the flavor was found to be an unsaturated aldehyde; however, not of the α-β unsaturated type. Further separation of the regenerated carbonyls by means of gas-liquid chromatography (GLC) points to a C₉-aldehyde. After synthesis of the 4-,5- and 6-cis- andtrans-nonenals, comparison made it probable that the carrier of the hardening flavor is a mixture of 6-cis and 6-trans-nonenal, the latter of which has the greatest share in the hardening flavor. In order to confirm the location of the double bond in the carrier of the hardening flavor a recent isolation technique was applied. The volatile substances from hardened linseed oil were first separated via GLC. After conversion of the carbonyls in question into their DNPHs, the latter have been separated by means of thin-layer chromatography (TLC). By means of IR-analysis and oxidation with osmium tetroxide of the pure derivative, the principal carrier of the hardening flavor has been identified as 6-trans-nonenal.     

Molecular markers associated with linolenic acid content in soybean

An altered FA profile with decreased linolenic (18∶3) acid in soybean germplasm was developed by crossing N97-3708-13, a soybean line with reduced 18∶3 (<5.4%) and ‘Anand’, a normal soybean cultivar (9.7% 18∶3). The resulting recombinant inbred lines are promising because they may promote healthier oil with improved oxidative stability and flavor. The objective of this study was to utilize the population N97-3708-13 × Anand to identify simple sequence repeat (SSR) markers associated with 18∶3 content. Two markers, Satt534 and Satt560, which are located approximately 10 cM apart from each other, near the Fan locus on linkage group B2, were identified as quantitative trait loci significantly associated with 18∶3 content (P=0.001, R ²=0.59, individually). The SSR markers identified in this study should be useful for implementation of marker-assisted selection for low-18∶3 genotypes in soybean breeding programs.     

The synthesis of 2-(1-Pentenyl) furan and its relationship to the reversion flavor of soybean oil

Cis and trans-2-(1-pentenyl) furan were postulated as possible contributors to the reversion flavor of soybean oil. These compounds were synthesized, and structures were confirmed by infrared, nuclear magnetic resonance, and mass spectroscopy. Organoleptic evaluation of them in oil suggested that they could contribute to the beany and grassy note of the reverted soybean oil. This approach to flavor problems of food was given a new name — reverse phase flavor chemistry. Paper of the Journal Series, New Jersey Agricultural Experiment Station, Cook College, Rutgers, The State University.     

Quantitation of flavor volatiles in oxidized soybean oil by dynamic headspace analysis

A dynamic headspace procedure was developed for isolating the volatiles from oxidized soybean oil and trapping them on an adsorbent under conditions that gave minimal decomposition of hydroperoxides (50°C for 30 min at a helium flow of 75 mL/min). The volatiles were desorbed from the adsorbent and separated by gas chromatography (GC) on a methyl silicone capillary column. Equations were derived from theoretical considerations that allowed the actual concentration of each flavor component in the oxidized oil to be calculated from the area of the GC peaks. The reliability of the method and calculations was demonstrated by recovery experiments. The concentration of 2-heptanone in a mineral oil emulsion, equivalent in flavor intensity to each component, was calculated and summed to estimate the overall flavor intensity of the samples. The overall estimations were compared with the concentrations of 2-heptanone observed to be equivalent in flavor intensity to the oxidized oil samples when these were tasted in emulsion. The concentrations of individual components calculated from the headspace volatiles data were all present at concentrations below their flavor thresholds, and the simple sum of the intensities of their flavors generally accounted for less than half of the flavor intensities of the oil samples. The differences in the headspace and sensory analyses might be attributed to the flavor of the unoxidized oil, synergistic interactions, and/or the presence of unmeasured flavors components.     

Effects of β-carotene and lycopene thermal degradation products on the oxidative stability of soybean oil

The relative oxidative stability of soybean oil samples containing either thermally degraded β-carotene or lycopene was determined by measuring peroxide value (PV) and headspace oxygen depletion (HOD) every 4 h for 24 h. Sobyean oil samples containing 50 ppm degraded β-carotene that were stored in the dark at 60°C displayed significantly (P<0.01) higher HOD values compared with controls. Lycopene degradation products (50 ppm) in soybean oil significantly (P<0.05) decreased HOD of samples when stored in the dark. PV and HOD values for samples containing 50 ppm of either β-carotene or lycopene degradation products stored under lighted conditions did not differ significantly from controls (P<0.05). However, soybean oil samples containing 50 ppm of unheated, all-trans β-carotene or lycopene stored under light showed significantly lower PV and HOD values than controls (P<0.01). These results indicated that during autoxidation of soybean oil held in the dark, β-carotene thermal degradation products acted as a prooxidant, while thermally degraded lycopene displayed antioxidant activity in similar soybean oil systems. In addition, β-carotene and lycopene degradation products exposed to singlet oxygen oxidation under light did not increase or decrease the oxidative stability of their respective soybean oil samples.     

Recent Developments in Breeding Soybeans for Improved Oil Quality

Soybean breeding has traditionally focused on improving crop productivity. While this is still the main objective of soybean breeding programs, there is increased emphasis on improving the intrinsic characteristics of the seeds. Breeding goals with respect to oil quality improvement have been threefold: (1) reducing lipoxygenase activity, (2) decreasing linolenic acid concentrations in soybean oil and more recently (3) reducing the saturated fatty acid, palmitate, in soybean oil. The first two are aimed at improving oil flavor and shelf life. The third has been added as a result of medical concerns about excess saturated fats in the human diet. Considerable progress has been made toward achieving all three objectives. Improved soybean germplasm has been released and is being used to develop high yielding cultivars improved oil quality. Null alleles for three lipoxygenase isozymes have been transferred into the cultivar ‘Century’ through backcrossing. Three different germplasm sources have been developed through mutagenesis and selection with low concentrations of linolenic acid (3.0 to 3.5%). This material has been released and is being widely used by public and private breeders to incorporate the low linolenic acid trait into high yielding cultivars. Cultivars with 2% germplasm with lower concentrations of palmitic acid in soybean oil by as much as 50%. In the U.S., soybean oil accounts for about 50% of the palmitic acid in diets. Therefore, reducing palmitic acid in soybean oil could significantly decrease saturated fat consumption without dietary change. Other research in progress includes studies of the genetic inheritance and control oil quality traits and efforts to use molecular genetics to improve oil quality.     

The impact of furanoid fatty acids and 3-methylnonane-2,4-dione on the flavor of oxidized soybean oil

Oils from soybeans with high or low contents of furanoid fatty acids were evaluated during storage for flavor intensity of soybean oil (SBO) off-flavor, but no significant differences were found. In addition, the compound 3-methylnonane-2,4-dione (MND), a breakdown product of furanoid fatty acids suggested by other researchers to contribute to reversion flavor of SBO, was evaluated for its contribution to off-flavor. The compound was synthesized in the laboratory and purified by gas chromatography (GC) on a Silar 10 C column. GC analysis of the purified MND on a nonpolar SPB-1 column showed two well-separated main peaks that have been suggested to represent keto and enol forms. Between these two peaks, a bridge of poorly resolved compounds may have represented various possible enol forms or an equilibration among these forms during the GC separation. MND had an intense straw-like and frulty odor when evaluated at the outlet of a gas chromatograph. Sensory evaluation of MND in a mineral oil/water emulsion system showed that its flavor intensity increased almost imperceptibly with increased concentration (from 0.09 to 2.56 ppm). An explanation for this unusual flavor response may be that, when molecularly dispersed in air, MND has an intense odor, but when placed in a mineral oil or soybean oil emulsion, MND may exist in a form with relatively low flavor intensity, or it may be bound by the media. The concentrations of MND in SBO at various peroxide values were measured at 0 to 0.804 ppb, which were far less than concentrations tested in mineral oil/water emulsions during sensory evaluation and below published odor threshold values for MND in oil. Therefore, these results do not support the theory that furanoid fatty acids or MND contribute strongly to the reversion flavor of SBO. This is Journal Paper J.17472 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa. Project No. 3396.     

Photochemical production of conjugated linoleic acid from soybean oil

Conjugated linoleic acid (CLA), an anticarcinogenic compound with numerous other health benefits, is present mainly in dairy and beef lipids. The main CLA isomer present in dairy and beef lipids is cis 9, trans 11 CLA at a 0.5% concentration. The typical minimum human dietary intake of CLA is 10 times less than the 3 g/d suggested requirement that has been extrapolated from animal and cell-line studies. The objectives of this study were to produce CLA isomers from soybean oil by photoisomerization of soybean oil linoleic acid and to study the oxidation status of the oil. Refined, bleached, and deodorized soybean oil with added iodine concentrations of 0, 0.1, 0.25, and 0.5% was exposed to a 100-W mercury lamp for 0 to 120 h. An SP-2560 fused-silica capillary GC column with FID was used to analyze the esterified CLA isomers in the photoisomerized oil. The CLA content of the individual isomers was optimized by response surface methodology. Attenuated total reflectance (ATR)-FTIR spectra in the 3400 to 3600 cm⁻¹ range and ¹H NMR spectra in the 8 to 12 ppm range of the photoisomerized soybean oil were obtained to follow hydroperoxide formation. The largest amount of cis 9, trans 11 CLA isomer in soybean oil was 0.6%, obtained with 0.25% iodine and 84 h of photoisomerization. Lipid hydroperoxide peaks in the ATR-FTIR spectra and aldehyde peaks in the ¹H NMR spectra were not observed in the photoisomerized soybean oil, and the spectra were similar to that of fresh soybean oil. This study shows that CLA isomers can be produced simply and inexpensively from soybean oil by photoisomerization.     

Spectrophotometric estimation of soybean oil in admixture with cottonseed and peanut oils

Conclusion A spectrophotometric method has been described for the determination of soybean oil in admixture with cottonseed oil. The method provides a simple and rapid means of detecting gross adulteration of one oil with another and permits an accurate determination of linolenic acid for use as a criterion of the economic value of an oil mixture and as a guide in oil processing. The factor limiting the precision of the method is variation in composition of the cottonseed and soybean oils in the mixtures to be analyzed. Variations in composition affect the proportion of measured triene conjugation, due to the linolenic acid content of the soybean oil and the apparent linolenic acid content of the cottonseed oil. Thus, for unknown mixtures only average value corrections can be made for apparent linolenic acid content and the accuracy of a particular analysis will depend upon how well the composition of the oils in the particular mixture follows those of the average mixture. The method described can be extended to mixtures other than those of soybean and cottonseed oils. Thus, soybean oil may be determined in admixture with a peanut oil. In general, any oil which has an unsaturated fatty acid capable of producing triene conjugation upon alkali isomerization can be determined in the presence of any other oil containing no appreciable quantity of unsaturated fatty acids which can produce triene conjugation by such treatment. Presented before The American Oil Chemists’ Society, New Orleans, Louisiana, May 10–12, 1944. One of the laboratories of the Bureau of Agricultural and Industrial Chemistry, Agricultural Research Administration, U. S. Department of Agriculture.     

Conversion of polyunsaturates in vegetable oils tocis-monounsaturates by homogeneous hydrogenation catalyzed with chromium carbonyls

Chromium carbonyl complex catalysts were used to selectively hydrogenate polyunsaturates in vegetable oils into products retaining 90% to 95%cis configuration and their liquid properties. The product from soybean oil contained 42–69% monoene, 10–40% diene and 0–4% triene. The product from safflower oil contained 73–82% monoene and 8–17% diene. About 45–55% of the double bonds in monoenes from hydrogenated soybean oil remained in the C₉ position, and the rest was distributed between C₁₀, C₁₁, and C₁₂. Preliminary oxidative and flavor stability evaluations showed that these hydrogenated soybean oils compared favorably with a commercial sample of hydrogenated-winterized soybean oil. Liquid fatty acids prepared by saponification of hydrogenated soybean and safflower oils (IV 90–100) had analyses about the same as those of commercial oleic acid. Presented before the Division of Agricultural and Food Chemistry, 156th American Chemical Society National Meeting, Atlantic City, N.J., September 1968.     

Preliminary gas chromatographic analysis of flavor compounds in mayonnaise

The flavor compounds in fresh, 3- and 6-month-old mayonnaise at room temperature have been analyzed by a gas chromatographic method. The results indicate that as the storage time of mayonnaise increased, the flavor compounds formed from oil in mayonnaise increased. However, the concentrations of allyl isothiocyanate which is the major flavor compound of mustard, and acetic acid and ethyl acetate which are the major compounds in vinegar did not change during the 6 months’ storage at room temperature. The analytical method described has shown good reproducibility in the analysis of mayonnaise flavor compounds and can be used as an instrumental analytical method to evaluate the mayonnaise flavor quality and to complement the sensory evaluation of mayonnaise. Ohio Agricultural Research and Development Center Journal Series Article 46–82.     

3-cis-hexenal,the “Green” reversion flavor of soybean oil

By means of gas-liquid-chromatography the neutral volatile products obtained by stripping oxidized soybean oil were fractionated. The fraction with “green bean” odor (a well-known offflavor of soybean oil) was investigated. By the application of various methods it was possible to isolate the flavor carrier proper and, in spite of its small amount, to complete the analysis to a point at which the structure 3-cis-hexenal seemed highly probable. Synthesis and subsequent comparison with the natural product confirmed the correctness of the assumption. n-Hexanal and 3-trans-hexenal, also found in this fraction, play no significant role in the reversion flavor.     

Photooxidation of soybean oils as affected by triacylglycerol composition and structure

The photooxidation of soybean oil was determined and correlated with triacylglycerol composition and structure. Purified triacylglycerols were photooxidized at room temperature under fluorescent light. Rates of peroxide formation and total headspace volatiles were related positively (P<0.5 significance) to oxidizability (r=0.75, r=0.76); content of linolenic acid (r=0.80, r=0.85) and linoleic acid (r=0.61, r=0.57); linoleic acid on carbon 2 (r=0.64, r=0.64); and average number of double bonds (r=0.76, r=0.76). Negative correlations were observed with respect to oleic acid (r=−0.70, r=−0.70). Soybean oil stability was decreased by linolenic acid-containing triacylglycerols and increased by oleic acid-containing triacylglycerols. Trilinoleoylglycerol and dilinoleoyl-oleoylglycerol were the most important oxidation product precursors. However, for high-linolenic acid soybean oil, dilinoleoyl-linolenoylglycerol and trilinoleoylglycerol were the most important oxidation product precursors. The most abundant volatile produced from thermal decomposition at 140°C of photooxidized triacylglycerols was 2-heptenal, except for high-linolenic acid oils, where the most abundant volatile was propanal. The photooxidative stability of soybean oil triacylglycerols with respect to composition and structure is of interest for the development of soybean varieties with oils of improved odor and flavor stability. Presented at the 20th ISF World Congress 83rd Annual American Oil Chemists’ Society Meeting, May 10–14, 1992, Toronto, Canada.     

Flavor evaluation of copper-hydrogenated soybean oils

The use of copper catalyst to reduce selectively the linolenate in soybean oil improves its flavor stability. As previously shown, the copper must be removed or properly inactivated to obtain an oil of high initial quality. In oven and heat tests, odor and flavor development in the hydrogenated soybean oil samples correlate surprisingly well with actual levels of linolenate, but there were some differences in overall responses among cottonseed oil, copper-reduced (0.0% linolenate) and nickel-reduced (3.0% linolenate) soybean oils. The taste panel generally scored the last three oils in the following order: cottonseed oil, copper-reduced and nickel-reduced soybean oil. One of 10 papers to be published from the Symposium “Hydrogenation”. presented at the AOCS Meeting, New Orleans, April 1970. No. Utiliz. Res. Dev. Div., ARS, USDA.