Volatile compounds in heated oleic acid-esterified propoxylated glycerol

Static headspace and capillary gas chromatography, in combination with infrared and mass spectrometry, were used to collect, separate, identify, and quantitate the oxidative and thermal decomposition products in two heated model compound triacylglycerols, oleic acid-esterified propoxylated glycerol (EPG-08 oleate) and triolein. The EPG-08 oleate and triolein were each heated in a deep-fat fryer at 192±8°C for 12 h each day until the oil contained ≥20% polymeric material. Most of the volatile compounds identified in heated triolein and EPG-08 oleate were associated with the oxidative and thermal decomposition of oleates, such as short-chain alkanes, aldehydes, and alcohols. The major volatile compounds (concentration >20 ppm) in heated EPG-08 oleate and triolein were heptane, octane, heptanal, octanal, trans-2-decenal, nonanal, and trans-2-undecenal. However, a few minor volatile compounds, not previously found in heated fats and oils, were identified in the heated EPG-08 oleate samples. These compounds included 1,2-propanediol, 1-hydroxy-2-propanone (hydroxyacetone), 1-acetoxy-2-propanone (acetoxyacetone), 2-ethyl-4-methyl-1,3-dioxolane, 2,2,4-trimethyl-1,3-dioxolane, 2-heptyl-4-methyl-1,3-dioxolane, and 4-methyl-2-octyl-1,3-dioxolane. Alkyl-substituted dioxolanes have been found in heated methyl linoleate.     

Nonvolatile components produced in triolein during deep-fat frying

Triolein was heated at 190°C (375°F) in a deep-fat fryer for 12 h/day until high-performance size-exclusion chromatography indicated polymer formation had exceeded 20%. Increases in the free fatty acid, total acid value, food oil sensor andp-anisidine values upon heating indicated that thermal oxidation and degradation of triolein had occurred. After the initial sample (day 0), the peroxide values decreased to very low values. The amount of polymeric triacylglycerol material increased during heating. Linear regression analysis of percent polymer vs. heating time indicated that the sample would contain ≥20% polymers after 51.1 h of heating. Capillary supercritical fluid chromatography (SFC) was used to determine the percentage of triolein remaining after 12, 24, 36, 48 and 60 h of heating, which was 68.6, 53.9, 35.9, 33.0 and 19.0%, respectively. The average reaction rate constant (apparent first-order) for the change in triolein concentration, SFC, during heating was 0.0256±0.0011 h⁻¹.     

Bleaching of vegetable oils: I. Conversions in soybean oil,triolein and trilinolein

Soybean oil, triolein and trilinolein were treated with acid-activated earth (2–15%) at 25–120 C. With low initial peroxides and in the absence of air, little additional conjugation was found. With high initial peroxides or in the presence of air, conjugated dienes and trienes were formed particularly with high amounts of earth. At higher temperatures and longer bleaching times the amount of conjugated compounds also increased. It can be concluded from the bleaching experiments with triolein and trilinolein that conjugated dienes, trienes and tetraenes are formed much more easily with linoleate than with oleate.     

Pan-heating of low-linolenic acid and partially hydrogenated soybean oils

Genetically modified low-linolenic acid soybean oil (LL-SBO) was compared to partially hydrogenated soybean oil (PH-SBO). Samples were heated on a Teflon pan at ∼180°C until a selected end point of ≥20% polymer content was reached. High-performance size-exclusion chromatography analysis indicated the PH-SBO contained >20% polymer after 20 min of heating, whereas the LL-SBO sample contained >20% polymer after 10 min. Supercritical fluid chromatography analysis indicated degradation rates of 0.161±0.011 min⁻¹ for LL-SBO and 0.086±0.004 min⁻¹ for PH-SBO. The volatile compounds were identified and quantitated with static head-space-GC-MS. 1-Heptene (239.9 ppm) and hexanal (1486.1 ppm) were present at the greates concentration among the volatile compounds in LL-SBO. The volatile compounds present in the greatest concentrations in heated PH-SBO were hexanal (376.9 ppm) and pentane (82.1 ppm). After 10 min of heating, the LL-SBO oil FFA value (2.66%), p-anisidine value (386.5 abs/g oil), Food Oil Sensor reading (18.75), and color intensity (Y=4.0, R=1.0) were significantly greater than those of PH-SBO after 14 min of heating (4.28%, 298.5 abs/g oil, 16.08, Y=1.0, R=0.1, respectively). There was a significant difference in the degradation rates between LL-SBO and PH-SBO (P<0.05). The PH-SBO was more stable than the LL-SBO.     

Pan-frying stability of NuSun oil,a mid-oleic sunflower oil

Pan-frying is a popular frying method at home and in many restaurants. Pan-frying stabilities of two frying oils with similar iodine values (IV)—mid-oleic sunflower oil (NuSun oil; IV=103.9) and a commercial canola oil (IV=103.4)—were compared. Each oil sample was heated as a thin film on a Teflon-coated frying pan at ∼180°C to a target end point of ≥20% polymer. High-performance size-exclusion chromatography analysis of the mid-oleic sunflower and canola oil samples indicated that the heated samples contained 20% polymer after approximately 18 and 22 min of heating, respectively. The food oil sensor values increased from zero to 19.9 for the canola sample and from zero to 19.8 for the mid-oleic sunflower sample after 24 min of heating. The apparent first-order degradation rate for the mid-oleic sunflower sample was 0.102±0.008 min⁻¹, whereas the rate for the canola sample was 0.092±0.010 min⁻¹. The acid value increased from approximately zero prior to heating to 1.3 for the canola sample and from zero to 1.0 for the mid-oleic sunflower sample after 24 min of heating. In addition, sensory and volatile analyses of the fried hash browns obtained from both oils indicated there were no significant differences between the two fried potato samples.     

Kinetics of geometrical isomerization of unsaturated FA in soybean oil

Laboratory treatment of soybean oil were carried out at the following conditions: atmospheric pressure in the presence of air or nitrogen at different temperatures ranging from 160 to 250°C for 12 to 72 h. These conditions were used to study geometric isomerization of cis,cis-linoleic and cis,cis,cis-linolenic acid in the presence or in the absence of oxidative degradation reactions. Based on these experiments, a model of consecutive, parallel reactions was developed to describe the reaction steps occurring in the soybean oil during heating at constant temperature. For both cis,cis-linoleic and cis,cis,cis-linolenic acid, the reaction of formation isomers followed a first-order reaction, and the rate constant of isomerization varied according to the Arrhenius law. The isomerization rate constant for linoleic acid was 9.57×10⁻³±0.50 h⁻¹ in the presence of oxygen and 7.39×10⁻³±0.39 h⁻¹ in its absence, and the isomerization rate constant for linolenic acid was 1.18×10⁻¹±0.10 h⁻¹ in the presence of oxygen and 0.87×10⁻¹±0.07 h⁻¹ in its absence (all obtained at 250°C).     

Evolution of short-chain glycerol-bound compounds during thermoxidation of FAME and monoacid TAG

Methyl heptanoate and octanoate, methyl 8-oxooctanoate and 9-oxononanoate, and dimethyl suberate and azelate were quantitated in model systems of FAME, i.e., methyl oleate and linoleate, and monoacid TAG, i.e., triolein and trilinolein, heated at 180°C for 5, 10, and 15h. Polar compounds, ranging from 20.2 to 53.2% in TAG systems, were also quantitated as a reference for the level of degradation of the samples analyzed. Base-catalyzed techniques were used for methylation as they did not modify the structure of the compounds of interest. The maximum amounts of short-chain FAME and of aldehydic FAME obtained were of the same order (3–4 mg/g sample in methyl linoleate and trilinolein heated for 15 h). After a second methylation step with diazomethane, amounts of dimethyl suberate and azelate were also quantitated. The total amounts of dimethyl esters in TAG model systems ranged from 1.07 to 3.35 mg/g of heated sample and were similar to those found for total aldehydic FAME (from 1.03 to 3.15 mg/g of heated sample), suggesting that oxidation of the aldehydic acyl group to an acid acyl group was a major reaction.     

Odor significance of undesirable degradation compounds in heated triolein and trilinolein

To better understand production of undesirable or negative odors such as fruity, plastic, and waxy that are characteristic of higher oleic acid-containing oils, model heated oil systems of triolein and trilinolein were studied. Identification of the odor significance of volatile compounds produced by fractionated and nonfractionated triolein and trilinolein was done by purge and trap-gas chromatography-ion trap mass spectrometry-olfactometry. The predominant odors of the triolein heated 1, 3, and 6 h at 190°C were fruity and plastic, with other negative odors of acrid and grassy. Some of the volatile compounds that produced negative odors in heated triolein, in order of increasing concentration, were hexanal (grassy), octanal (fruity), (E)-2-decenal (plastic), nonanal (fruity), and (E)-2-undecenal (plastic). Some of the negative odor compounds in trilinolein heated for 1, 3 and 6 h, in order of increasing concentration, included (E)-2-nonenal (plastic), pentanal (grassy), and hexanal (grassy). However, the amount of volatile compounds produced and the intensity levels of the odors were lower in trilinolein than in triolein. Formation of many of the volatiles was explained after identification of the volatile precursors, including epoxy, keto, and dimer oxidation products that were produced during heating. Presented at the American Oil Chemists’ Society 90th Annual Meeting & Expo, Orlando, FL, May 9–12, 1999.     

Dietary fatty acids: Their metabolic fate and influence on fatty acid biosynthesis

Adult rats fed a low-fat diet or diets containing 15% of either tripalmitin, triolein or trilinolein were injected intraperitoneally with H³-labeled acetate. Those which received fat were also given by mouth, simultaneously with acetate, the 1-C¹⁴-labeled sodium salt of the respective dietary fatty acid. The fate of the tagged material was followed by time-spaced biopsies of subcutaneous adipose tissue and by collection of the expired C¹⁴O₂. After 72 hr, 51, 64, and 52% of the dietary palmitate, oleate, and linoleate, respectively, were catabolized, as indicated by the corresponding percentages of the label having been excreted as C¹⁴O₂. Dietary linoleate was relatively less incorporated into body triglycerides than palmitate and oleate. Animals ingesting diets of 15% triolein had only about one-half the amount of phospholipids in their tissues as had the other groups. The distribution of both the C¹⁴ and H³ labels in the tissue triglycerides showed that all diets containing fats decreased fatty acid synthesis but did not inhibit conversion of palmitate to oleate. Conversions of oleate or linoleate appeared to be through acetate. As a result of these factors, the fatty acid composition of the tissue triglycerides after 3 months’ ingestion of tripalmitin was essentially the same as that of the low-fat group, whereas the ingestion of triolein produced triglycerides with a very high content of oleic acid. Trilinolein ingestion produced effects similar to triolein but to a less pronounced degree. Both the respiratory C¹⁴O₂ and the C¹⁴- and H³-labeled fatty acids in subcutaneous adipose tissue exhibited a second rise in specific activity 12 to 24 hours after the administration of the label. Presented at the AOCS meeting, Chicago, 1964.     

Novel method for rapid monitoring of lipid oxidation by FTIR spectroscopy using disposable IR cards

A new FTIR approach was investigated for assessing edible oil oxidative stability with the use of polymer IR (PIR) cards as sample holders. This approach allows oil oxidation to be monitored at moderate temperatures owing to the fairly rapid rate at which unsaturated oils oxidize on the PIR cards. To assess the FTIR/PIR card method, pure TAG—triolein, trilinolein, and trilinolenin—were loaded onto cards and placed in a chamber where warm air (55°C) flowed over them continuously to facilitate oxidation. At periodic intervals, individual cards were removed and their FTIR spectra scanned, after which they were replaced in the aeration chamber. All spectra were normalized to compensate for variations in PIR card path lengths or oil loadings, and for each card the initial spectrum recorded (t=0) was subtracted from all subsequent spectra taken over time to produce differential spectra. With the use of a peak-find algorithm, the absorbance minimum in the cis region (3017–3000 cm⁻¹) and the absorbance maxima in the hydroperoxide (3550–3200 cm⁻¹), isolated trans (977–957 cm⁻¹), and conjugated trans regions (995–983 cm⁻¹) were measured in the differential spectra and plotted as a function of time. For all three TAG, the loss of cis double bonds was linearly related to the development of hydroperoxides and isolated trans bonds for much of the oxidation process, whereas for the polyunsaturated TAG a similar relationship also existed for conjugated trans species. Based on experimentally determined hydroperoxide (ROOH) absorbance slope factor (0.06 mAbs/PV), ROOH absorbance changes were converted to PV, allowing direct PV monitoring as a function of time using the PIR cards. Trilinolenin, trilinolein, and triolein attained a PV of 100 mequiv/kg oil after, 43,98, and 2889 min, respectively, their relative reaction rates being similar to ratios published in the literature. The assessment of the FTIR/PIR card method using TAG indicates that it may be a practical and rapid means of oxidizing lipids and tracking their oxidative state in terms of PV so as to provide a measure of their oxidative stability.     

Plasma kinetics of a cholesterol-rich emulsion in young, middle-aged, and elderly subjects

Low density lipoprotein (LDL) plasma concentration is increased in the elderly. In this group, the incidence of coronary artery disease (CAD) is greater and LDL remains an important risk factor for CAD development. In this study, the plasma kinetics of a cholesterol-rich emulsion that binds to LDL receptors was studied in 10-subject groups of the elderly (70±4 yr), middle-aged (42±5 yr) and young (23±2 yr). All were normolipidemic, nonobese, nondiabetic subjects who did not have CAD. The emulsion was labeled with ¹⁴C-cholesteryl oleate and injected intravenously into the subjects. Blood samples were drawn at regular intervals over 24 h to determine the plasma decay curve of the emulsion radioactive label and to estimate its plasma fractional clearance rate (FCR, in h⁻¹). FCR of the emulsion label was smaller in elderly compared to young subjects (0.032±0.035 and 0.071±0.049 h⁻¹, respectively; mean±SD, P<0.05). FCR of the middle-aged subjects (0.050±0.071 h⁻¹) was intermediate between the values of the elderly and young subjects, although not statistically different from them. A negative correlation was found between the emulsion FCR and subjects’ age (r=−0.47, P=0.008). We conclude that aging is accompanied by progressively diminished clearance of the emulsion cholesterol esters and, by analogy, of the native LDL.     

Quantitation of volatile compounds in heated trilinolein by static headspace-capillary gas chromatography/infrared spectroscopy-mass spectrometry

Static headspace and capillary gas chromatography/infrared spectroscopy-mass spectrometry were used to collect, separate, detect and quantitate the oxidative and thermal decomposition products in heated trilinolein. Trilinolein without added antioxidants was heated in a deep-fat fryer at 190°C for 12 h each day for a total of 24 h until the amount of polymeric material in the oil exceeded 20%. The concentration of the volatile compounds in trilinolein, as well as thep-anisidine values and polymer content, increased with an increase in heating time. After 24 h of heating, the concentrations of the major volatile compounds (>10 ppm) had increased as follows: pentane (33.4–398.7 ppm), pentanal (11.7–73.5 ppm), 1-pentanol (5.2–46.7 ppm), hexanal (67.6–346.1 ppm), (E)-2-hexenal (12.9–87.6 ppm), (E)-2-heptenal (309.6–894.1 ppm), 1-octen-3-ol (35.2–150.2 ppm), formic acid (0–18.7 ppm), hexanoic acid (0–83.8 ppm), heptanal (2.5–29.5 ppm), 2-pentylfuran (29.2–146.9 ppm), (E)-2-octenal (26.4–347.2 ppm), (E)-2-octenal (26.4–347.2 ppm), (E)-2-nonenal (31.6–163.2 ppm), (E,Z)-2,4-decadienal (15.0–320 ppm) and (E,E)-2,4-decadienal (43.1–1137.5 ppm).     

Triolein and Trilinolein Ameliorate Oxidized Low-Density Lipoprotein-Induced Oxidative Stress in Endothelial Cells

Uptake of oxidized low-density lipoprotein by endothelial cells is a critical step for the initiation of atherosclerosis. Triacylglycerol uptake in these cells is understood to be a part of the process. The present investigation, comparison among the effects of simple acylglycerol, including tristearin, triolein, and trilinolein, upon oxidized low-density lipoprotein -induced oxidative stress was undertaken. Results indicated that trilinolein (78 % ± 0.02) and triolein (90 % ± 0.01) increased cell viability of endothelial cells exposed to oxidized low-density lipoprotein, whereas tristearin decreased the cell viability (55 % ± 0.03) (P < 0.05). Oxidized low-density lipoprotein treatment significantly increased apoptosis (23 %), compared to cells simultaneously exposed to trilinolein (19 %) or triolein (16 %), where apoptosis was reduced (P < 0.05). On the other hand, exposure to tristearin further increased oxidized low-density lipoprotein -induced cell apoptosis (34 %). Treatment with trilinolein or triolein on oxidized low-density lipoprotein -stimulated endothelial cells inhibited the expression of ICAM-1 and E-selectin mRNA. Moreover, both trilinolein and triolein demonstrated a strong antioxidant response to oxidative stress caused by oxidized low-density lipoprotein. Taken together, the results indicate trilinolein and triolein possess anti-inflammatory properties, which are mediated via the antioxidant defense system.     

Autoxidation of methyl linoleate initiated by the ozonide of allylbenzene

Allylbenzene ozonide (ABO), a model for polyunsaturated fatty acid (PUFA) ozonides, initiates the autoxidation of methyl linoleate (18∶2 ME) at 37°C under 760 torr of oxygen. This process is inhibited by d-α-tocopherol (α-T) and 2,6-di-ert-butyl-4-methylphenol (BHT). The autoxidation was followed by the appearance of conjugated diene (CD), as well as by oxygen-uptake. The rates of autoxidation are proportional to the square root of ABO concentration, implying that the usual free radical autoxidation rate law is obeyed. Activation parameters for the thermal decomposition of ABO were determined under N₂ in the presence of radical scavengers and found to be E_a=28.2 ±0.3 kcal mol⁻¹ and log A=13.6±0.2; k_d (37°C) is calculated to be (5.1±0.3)×10⁻⁷ sec⁻¹. Autoxidation data are also reported for ozonides of 18∶2 ME and methyl oleate (18∶1 ME).     

In Situ Rheo-NMR Measurements of Solid Fat Content

The properties of crystallized fats depend on their solid fat content (SFC) and their fractal structures. The SFC and the structures are dramatically affected during crystallization under shear flow. A mini-Couette cell was developed to crystallize fat samples under shear. The cell was tested with blends of canola stearin (CS) in canola oil (CO) in a 20-MHz NMR spectrometer. The blends were placed in the cell, melted at 80 °C, and then crystallized under different shear rates (58–460 s⁻¹) at 40 °C inside the spectrometer for 4 h. Time averaged NMR free induction decay (FID) curves were captured at 20 s intervals. SFC values were calculated using parameters determined by a calibration procedure. The SFC values determined by the direct method with and without the shaft of the Couette device were reasonably close. Similar results were observed with and without shear in the Couette device. The FID curves did not show a significant difference either. Therefore this system is accurate for in-situ time-resolved determination of SFC under shear flow. Furthermore, a combination of the direct and the indirect methods was successfully used to estimate the temperature increase due to viscous heating. The system developed will help in understanding the effects of shear flow on SFC of nanostructured lipid multicomponent systems. This will permit the optimization of the manufacturing processes.     

Quantitation of volatile compounds in heated triolein by static headspace capillary gas chromatography/infrared spectroscopy-mass spectrometry

Static headspace and capillary gas chromatography/infrared spectroscopy-mass spectrometry were used to collect, separate, identify and quantitate the oxidative and thermal decomposition products in heated triolein. Approximately 4 L of triolein was heated in a deep-fat fryer at 190°C for 12 h each day for a total of 60 h, until it contained>-20% polymeric material. The concentration of heptane and octane increased after 12 h heating, then decreased upon further heating, while the concentration of volatile aldehydes decreased gradually during heating. After 12 h of heating, the concentrations of the major volatiles changed as follows: heptane (16.36 to 30.17 ppm), octane (24.48 to 40.77 ppm), heptanal (21.68 ppm to 9.11 ppm), octanal (35.53 to 36.64 ppm), nonanal (68.95 to 43.16 ppm), (E)-2-decenal (135.00 to 89.00 ppm) and (E)-2-undecenal (88.50 to 61.00 ppm). After 60 h of heating, when the oil contained over 20% polymeric material, the concentrations of heptane, octane, heptanal, octanol, nonanal, (E)-2-decenal and (E)-2-undecenal were 5.16, 4.39, 5.45, 7.02, 18.07, 12.50 and 6.00 ppm, respectively.     

13C nuclear magnetic resonance spectroscopic analysis of the triacylglycerol composition ofBiota orientalis and carrot seed oil

¹³C nuclear magnetic resonance (NMR) spectroscopic analysis of the whole oil (triacylglycerols) ofBiota orientalis seeds confirms the presence of oleate [18:1(9Z)], linoleate [18:2(9Z, 12Z)], linolenate [18:3((9Z, 12Z, 15Z)], 20:3 (5Z, 11Z, 14Z), 20:4(5Z, 11Z, 14Z, 17Z), and saturated fatty acids in the acyl groups by comparing the observed carbon shifts with previously established shift data for model triacylglycerols. This technique shows that the saturated, 20:3 and 20:4 fatty acids are distributed mainly in the α-acyl positions, whereas oleate, linoleate, and linolenate are randomly acylated to the α- and β-positions of the glycerol “backbone”. Stereospecific hydrolysis of theBiota oil with pancreatic lipase, followed by chromatographic analysis of fatty esters, reveals the presence of trace amounts of 16:0(0.7%), 18:0(0.5%), 20:3 (0.4%), and 20:4 (1.3%) in the β-position of the glycerol “backbone”, which are undetectable by¹³C NMR technique on the whole oil. Semiquantitative assessment of the¹³C NMR signal intensities gives the relative percentages of the fatty acid distribution as: saturated 16:0, 18:0 (12.0% α-acyl), oleate (7.7% α-acyl 8.7% β-acyl), total linoleate and linolenate (31.7% α-acyl; 24.2% βacyl), total 20:3 and 20:4 (15.7% α-acyl). The¹³C NMR spectroscopic analysis of carrot seed oil identifies the presence of saturated (18:0), 18:1(6Z), 18:1(9Z), and 18:2(9Z, 12Z). The saturated fatty acid is found in the α-acyl positions. Semi-quantitative assessment of the signal intensities gives the relative percentages of the fatty acids as: 18:0 (4.5% α-acyl), 18:1(6Z) (49.6% α-acyl; 19.7% β-acyl), oleate (6.5% α-acyl; 8.6% β-acyl) and linoleate (5.2% α-acyl; 6.9% β-acyl).     

The structure of the triacylglycerols of meadowfoam oil

The triacylglycerols of meadowfoam oil have been resolved by HPLC in the silver ion and reversed-phase modes, and by the two techniques used in a complementary fashion. The fractions obtained were collected and quantified by gas chromatography of their methyl esters in the presence of an internal standard. Silver ion chromatography gave a distinctive resolution in which fractions differing solely in the position and chain-length of a single monoenoic fatty acyl group were resolved, the order of elution being 11−20∶1, 5−20∶1, 13−22∶1, 5−18∶1 and 9−18∶1. Reversed-phase chromatography also gave fractions containing single positional isomers, (11−20∶1<5−20∶1<13−22∶1<5−22∶1), but the pattern was more difficult to discern since fractions containing 22∶2 tended to overlap with those containing 20∶1. The species (5−20∶1)(5−20∶1)(22∶2), (5−20∶1)(5−20∶1)(5−20∶1) and (5−20∶1)(5−20∶1)(13−20∶1) were found to be the most abundant, and together comprised 67% of the total. A small but significant trilinolein fraction was detected and its presence may have biosynthetic implications.     

Restructing butterfat through blending and chemical interesterification. 1. Melting behavior and triacylglycerol modifications

Chemical interesterification of butterfat-canola oil blends, ranging from 100% butterfat to 100% canola oil in 10% increments, decreased solid fat content (SFC) of all blends in a nonlinear fashion in the temperature range of 5 to 40°C except for butterfat and the 90∶10 butterfat/canola oil blend, whose SFC increased between 20 and 40°C. The sharp melting associated with butterfat at 15–20°C disappeared upon interesterification. Heats of fusion for butterfat to the 60∶40 butterfat/canola oil blend decreased from 75 to 60 J/g. Blends with >50% canola oil displayed a much sharper drop in enthalpy. Heats of fusion were 30–50% lower on average for interesterified blends than for their noninteresterified counterparts. Both noninteresterified and interesterified blends deviated substantially from ideal solubility, with greater deviation as the proportion of canola oil increased. The change in the entropy of melting was consistently higher for noninteresterified blends than for interesterified blends. Chemical interesterification generated statistically significant differences for all triacylglycerol carbon species (C) from C₃₀ to C_56′ except for C_42′ and in SFC at most temperatures for all blends.     

Effect of temperature on linolenic acid loss and 18:3 Δ9-cis, Δ12-cis, Δ15-trans formation in soybean oil

Oil was extracted from soybeans, degummed, alkalirefined and bleached. The oil was heated at 160, 180, 200, 220 and 240°C for up to 156 h. Fatty acid methyl esters were prepared by boron trifluoride-catalyzed transesterification. Gas-liquid chromatography with a cyanopropyl CPSil88 column was used to separate and quantitate fatty acid methyl esters. Fatty acids were identified by comparison of retention times with standards and were calculated as area % and mg/g oil based on 17:0 internal standard. The rates of 18:3ω3 loss and 18:3 Δ9-cis, Δ12-cis, Δ15-trans (18:3c,c,t) formation were determined, and the activation energies were calculated from Arrhenius plots. Freshly prepared soy oil had 10.1% 18:3ω3 and no detectable 18:3c,c,t. Loss of 18:3ω3 followed apparent first-order kinetics. The first-order rate constants ranged from .0018±.00014 min⁻¹ at 160°C to .083±.0033 min⁻¹ at 240°C. The formation of 18:3c,c,t did not follow simple kinetics, and initial rates were estimated. The initial rates (mg per g oil per h) of 18:3c,c,t formation ranged from 0.0031±0.0006 at 160°C to 2.4±.24 at 240°C. The Arrhenius activation energy for 18:3ω3 loss was 82.1±7.2 kJ mol⁻¹. The apparent Arrhenius activation energy for 18:3c,c,t formation was 146.0±13.0 kJ mol⁻¹. The results indicate that small differences in heating temperature can have a profound affect on 18:3c,c,t formation. Selection of appropriate deodorization conditions could limit the amount of 18:3c,c,t produced.