Application of ESI and MALDI‐TOF MS for triacylglycerols analysis in edible oils

Triacylglycerols (TAG) are the most important group of compounds present in vegetable oils. These biomolecules, determining the physical, chemical and nutritional properties of the oils, are considered to be good fingerprints for quality and authenticity control. Therefore, TAGs characterization is a very important task in edible oil field, which has been undertaken by different analytical methods. The analysis of vegetable oils is still dominated by classic determinations, which are however laborious and time‐consuming and cannot be used routinely. More recently, advances in MS instrumentations coupled with online separation techniques and data processing have contributed to great expansion of MS in oil study, allowing the development of innovative analytical approaches that exhibit higher sensitivity, accuracy and rapidity in vegetable oils investigations. In the present contribution, a review of the most relevant applications of novel mass spectrometric techniques, such as ESI and MALDI, both alone and hyphenated with HPLC, used for analysis of the complex TAGs mixture of edible oils is illustrated.     

The estimation of the composition of edible oil mixtures

The fatty acid, sterol and tocopherol contents of edible oils were used to determine the composition of oil mixtures by means of a weighted least squares estimator with backwards elimination. The model was tested on 93 samples containing known amounts of sunflowerseed oil, groundnut oil, soybean oil, cottonseed oil, maize oil, olive oil and palm oil. Of these samples 75 were binary mixtures, seven were ternary mixtures, one contained seven oils and 10 were pure oils. Satisfactory results were obtained with 79 of the 93 samples (85%). The differences between the actual and estimated concentrations of the main components were greater than 2.7 times the standard error for six samples; the mean of the absolute differences was 4.7% for all 93 samples. The use of this model is considered superior to the matching of a fatty acid composition, but the model still needs a lot of development.     

Improving the oxidative stability of polyunsaturated vegetable oils by blending with high-oleic sunflower oil

Mixing different proportions of high-oleic sunflower oil (HOSO) with polyunsaturated vegetable oils provides a simple method to prepare more stable edible oils with a wide range of desired fatty acid composition. Oxidative stability of soybean, canola and corn oils, blended with different proportions of HOSO to lower the respective levels of linolenate and linoleate, was evaluated at 60°C. Oxidation was determined by two methods: peroxide value and volatiles (hexanal and propanal) by static headspace capillary gas chromatography. Determination of hexanal and propanal in mixtures of vegetable oils provided a sensitive index of linoleate and linolenate oxidation, respectively. Our evaluations demonstrated that all-cis oil compositions of improved oxidative stability can be formulated by blening soybean, canola and corn oils with different proportions of HOSO. On the basis of peroxide values, a partially hydrogenated soybean oil containing 4.5% linolenate was more stable than the mixture of soybean oil and HOSO containing 4.5% linolenate. However, on the basis of volatile analysis, mixtures of soybean and HOSO containing 2.0 and 4.5% linolenate were equivalent or better in oxidative stability than the hydrogenated soybean oil. Mixtures of canola oil and HOSO containing 1 and 2% linolenate had the same or better oxidative stability than did the hydrogenated canola oil containing 1% linolenate. These studies suggest that we can obviate catalytic hydrogenation of linolenate-containing vegetable oils by blending with HOSO. Presented at the AOCS/JOCS joint meeting, Anaheim, CA, April 25–29, 1993.     

Usefulness of the frequencies of some Fourier transform infrared spectroscopic bands for evaluating the composition of edible oil mixtures

Twelve sets of olive oil mixed with various edible seed oils in different proportions were made. The seed oils used were sunflower, corn, walnut, rapeseed, soybean, safflower, peanut, wheat germ, and sesame oil. These samples have very different proportions of saturated, mono- and polyunsaturated acyl groups, and of minor components. Fourier transformed infrared spectra of these blends were recorded from films of the oil samples between two discs of KBr. Taking into account the close relationships found previously between the frequency data of some specific bands and the composition of the oil samples, frequency data of all samples were collected and used in equations that relate frequency and composition predicting the percentage by weight of saturated, mono- and polyunsaturated acyl groups in the samples. The predicted values were compared with those derived from the combination of chemical and gas-chromatographic methods and a high degree of agreement was found. The presence of small amounts of seed oil in olive oil is shown by a small variation in the values of the frequencies of specific bands of the spectra, resulting from a smaller proportion of mono-unsaturated acyl groups than in pure olive oil. On the other hand, the frequency of the maximum absorbance between 915 cm^–1 and 904 cm^–1 also indicates the proportion of seed oil in the blend up to levels of 5% or up to 8% of seed oil in olive oil, depending on the nature of the seed oil. This methodology could give information about the composition of blends of edible oils in a very fast and simple way.     

Enzymatic Interesterification of Lauric Fat Blends Formulated by Grouping Triacylglycerol Melting Points

Lauric fat blends (appreciable amount of lauric fat with liquid oil and hard fat) initially formulated for shortening production by grouping triacylglycerol (TAG) melting points were further modified by enzymatic interesterification (EIE) to improve their key functionalities as plastic fats. At a similar fat blend formulation, only the high melting fat and medium melting fat were interesterified in binary‐EIE. Meanwhile, both fats and the liquid oil were interesterified in ternary‐EIE. The solid fat content (SFC) of all binary‐EIE blends was generally retained as similar in the temperature range between 0 and 20 °C when the amount of unsaturated TAGs was limited by excluding the liquid oil during EIE. However, the SFC was significantly reduced at temperatures above 20 °C compared to that of the initial blends. Furthermore, the melting point of binary‐EIE blends at BH50H15 formulation prepared with palm stearin and fully hydrogenated rapeseed oil as the hard fat was found to be drastically reduced from 54.6 to 35.3 °C and from 62.8 to 39.2 °C, respectively. In contrast, the SFC of ternary‐EIE blends was generally reduced when more unsaturated TAGs were available for EIE by including the liquid oil. However, higher SFC was noticed at temperatures around 10 °C in ternary‐EIE blends, as the amount of high‐melting fractions in their initial blends was increased from BH50H5 to BH50H15. Eventually, both binary and ternary‐EIE were also found to significantly alter the crystal microstructure of lauric fat blends, in terms of crystal morphology, size and network density.     

Enzymatic modification of canola/palm oil mixtures: Effects on the fluidity of the mixture

The bioreactor system to interesterify edible oils and fats at an ultra-micro aqueous phase of 100 ppm and less was investigated. The adsorption of lecithin, together with lipase onto a carrier, was effective for conducting the interesterifying reaction efficiently for oils and fats in micro aqueous phase. To improve the handling properties of palm oil at rather low temperature, palm oil was blended with canola or soybean oil, and then these blended oils were modified by enzymatic selective interesterification in a solvent-free, ultra-micro aqueous bioreactor system with an immobilized lipase that had 1,3-positional specificity. The effects of enzymatic interesterification were confirmed by triglyceride determination, by solid fat content profiles and by cloud point profiles, which were also compared to products of chemical interesterification. The improvement in the fluidity of blended oils with canola oil by the enzymatic reaction was bigger than with soybean oil, and chemical interesterification had no effects on the fluidity of blended oils.     

Long-term stability of blends of sesame oil or soybean oil with tuna oil under daily use conditions

The commercial feasibility of blending tuna oil into edible oil was studied from the perspective of stability under daily use conditions. A 210-day long-term simulation experiment was carried out on tuna oil blended with soybean or sesame oil at room temperature and cold storage (4°C). The bottle caps of all samples were opened manually and left open for 5 min every day to simulate the daily use of edible oil by consumers. The results indicate that cold storage can stabilize the blended oils containing tuna oil, and the peroxide and anisidine values of blended oil can be controlled at the recommended levels for at least 90 days by adding sesame oil. The polyunsaturated fatty acid content of all samples decreased by no more than 10% during the study term. The results of the sensory test indicated that in the daily use situation, the mixture of 20% tuna oil with 80% sesame oil could be stored at 4°C for up to 60 days without unacceptable quality and flavor changes. This study presents suggestions on how to design the packaging volume of the blended oil containing tuna oil, how to store the blended oil, and the term of best used before (once open) in practical commercial applications.     

Quantitative determination of diesel oil in contaminated edible oils using high-performance liquid chromatography

A simple and reliable high-performance liquid chromatography method for the analysis of diesel oil in contaminated edible oils is described. Analysis performed using a diol column with a mobile phase of heptane and isopropanol (94∶6, vol/vol). Although baseline separation between diesel and other background fluorescent components was not achieved, quantitation was still possible using baseline integration. The method is linear over the range of 5–1000 μg/g with a correlation coefficient (r ²) of 0.9984. Average recoveries from spiked edible oils were 94.4–101.3%, with a limit of quantitation (LOQ) of 5 μg/g for sunflower oil, palm olein, and groundnut oil. Corn oil has a higher content of ester components, thus, LOQ was slightly worse (40 μg/g). The applicability of the method was confirmed by gas chromatography-mass spectroscopic detection to show the presence of diesel hydrocarbons in the suspected contaminated crude palm oil. This procedure provides a simple and sensitive method for determining diesel oil concentration in contaminated edible oils without prior sample cleanup or extraction.     

Effects of microwave treatment on the oxidative stability of peanut (Arachis hypogaea) oils and the molecular species of their triacylglycerols

Peanut seeds (Arachis hypogaea) were roasted for 6, 12, 20, or 30 min at a frequency of 2450 MHz using a microwave oven. The quality characteristics and the compositions of the oils, i.e. their tocopherol distributions and the molecular species of the triacylglycerols (TAGs) were investigated. These results were compared with those of an unroasted oil sample. Only minor increases (p <0.05) in chemical and physical properties of the oils, such as the carbonyl value, the p‐anisidine value and the color development occurred after a prolonged roasting period. Compared to the original level, more than 92 wt‐% tocopherols remained after 30 min of roasting. A modified thin‐layer chromatography argentation procedure provided 12 different groups of TAGs, based on both the degrees of unsaturation and the total fatty acid chain‐length. Although significant increases (p <0.05) generated in these chemical and physical changes of the oils after 20 min of roasting, no significant loss (p >0.05) was observed in the molecular species of the TAGs during microwave roasting. These results indicate that phospholipids may be attributed to the quality characteristics of peanut oils during microwave roasting.     

Residual amounts of chlorophylls and pheophytins in refined edible oils

The quantities of chlorophyll (CHL) A and B, and pheophytin (PHY) A and B in 10 kinds of refined edible oils were estimated by the fluorometric method. The results revealed that CHL and PHY were present in commercial edible oils. PHY A showed the highest content at ca. 67% in total pigments. Compositional ratios of CHL and PHY were similar in different kinds of plant oils. Through the analysis of rapessed oils at every refining step, we determined that PHY is not formed during oil refining. In the autoxidation of soybean oils to which various amounts of CHL mixtures had been added, the peroxide value of tested oils increased in proportion to the total chlorophyll content. In addition, the compositional changes of the 4 components during autoxidation were investigated.     

The composition of edible oils modifies β-sitosterol/γ-oryzanol oleogels. Part I: Stripped triglyceride oils

The role of the fatty acid (FA) composition of triglycerides (TAGs) on sterol/sterol ester oleogels has been studied. Minor oil components of three vegetable oils with varying degrees of unsaturation (iodine values, IV) were removed. Typical oil quality parameters were determined before and after the treatment, and oleogels were produced using all six oils. Characteristic gel properties such as transition temperatures, mechanical properties and microstructure were tested. The results were compared regarding the impact of IV and the stripping procedure. Minor components were essentially removed during stripping, resulting in significantly different oil properties such as peroxide value, free FA and viscosity. However, peroxides formed rapidly in stripped flaxseed oil. Gel–sol transition temperatures and enthalpies were higher in gels from untreated oils and decreased with IV in samples with stripped oils. In contrast, the sol–gel transition was suppressed due to minor oil components in untreated oils. The effect of IV on gel formation was much less and linked to a lower solvent viscosity in more unsaturated oils. Nevertheless, gel firmness was significantly higher in oleogels from untreated oils and decreased slightly with IV in stripped oils. That was associated with differences in the arrangement of network building blocks, which was confirmed using atomic force microscopy. This study showed that the FA composition of TAGs has a limited effect on oleogel properties compared to those of minor oil components. The next part of this study focuses on modifying oleogel properties by adding selected minor components to stripped oils at varying concentrations.     

Changes in triacylglycerol molecular species and thermal properties of blended and interesterified mixtures of coconut oil or palm oil with rice bran oil or sesame oil

Blended oils were prepared by mixing appropriate amounts of coconut oil (CNO) or palm oil (PO) with rice bran oil (RBO) or sesame oil (SESO) to get approximately equal proportions of saturated/monounsaturated/polyunsaturated fatty acids in the oil. These blended oils were subjected to interesterification reactions using lipase to randomize the fatty acid distribution on the glycerol molecule. The fatty acid compositions of the modified oils were evaluated by gas chromatography while changes in triacylglycerol molecular species were followed by HPLC. The triacylglycerol molecular species of the blended oils reflected those present in the parent oil. Interesterification of the blended oils resulted in the exchange of fatty acids within and between the triacylglycerol molecules, resulting in alterations in the existing triacylglycerol molecules. Emergence of new triacylglycerol molecular species following interesterification was also observed. The thermal profiles of the native, blended and interesterified oils were determined by differential scanning calorimetry. Thermal behaviour, melting and crystallization properties of the modified oils showed significant changes reflecting the changes in the triacylglycerol molecules present in the oil. Therefore, interesterification of oils introduces significant changes in the physical properties of oils, even though the overall fatty acid composition of blended and interesterified oils remains the same.     

Authentication of Edible Vegetable Oil and Refined Recycled Cooking Oil Using a Micro-UV Spectrophotometer Based on Chemometrics

We have developed a new method to make a clear distinction between edible oil and refined recycled cooking oil by using a micro-ultraviolet spectrophotometer and analysis of spectra full data combined with principal component analysis (PCA) and cluster discriminant analysis (CDA). This method shows an excellent capability of distinguishing edible oil from refined recycled cooking oil, which is difficult to accomplish by previous methods using physical and chemical properties. Edible oils and refined recycled cooking oils have the different positions on the resulting plot of PCA and CDA. The oil samples are respectively concentrated relatively distribution and distinct from other kinds, with a certain amount of edge intersection between samples. But the edible oil and refined recycled cooking oil samples have a clear divisional interface. By the increase in sampling and the improvement of modeling, the edge intersection or overlap of commonality between samples can be reduced. Using this method, it is possible to determine qualitatively the identity of unknown samples.     

Analysis of triacylglycerols in refined edible oils by isocratic HPLC‐ESI‐MS

A simple, fast and reproducible reversed‐phase high performance liquid chromatography (HPLC) method coupled to electrospray ionization mass spectrometry (ESI‐MS) for the analysis of triacylglycerols (TAGs) species in the commercial edible oils has been developed. The TAGs species were separated using isocratic 18% isopropanol in methanol and a Phenomenex C18 column. The ESI‐MS conditions were optimized using flow injection analysis of standard TAG. Fifteen, fourteen, and sixteen TAGs were separated and identified in corn oil, rapeseed oil, and sunflower oil, respectively. The presence of intense protonated molecular (M + H⁺), ammonium (M + ${\rm NH}_{4}^{ + } $ ), and sodium (M + Na⁺) adducts ions and their respective diacylglycerols ions in the ESI‐MS spectra showed correct identification of TAGs. Some minor potassium adducts (M + K⁺) were also found. In addition, the identity of the fatty acid, position of each fatty acid, and the location of the double bond in the fatty acid moiety were explained. It was found that this isocratic method is useful for fast screening and identification of triacylglycerols in lipids.     

Optimization of biodiesel production from edible and non-edible vegetable oils

The non-edible vegetable oils such as Jatropha curcas and Pongamia glabra (karanja) and edible oils such as corn and canola were found to be good viable sources for producing biodiesel. Biodiesel production from different edible and non-edible vegetable oils was compared in order to optimize the biodiesel production process. The analysis of different oil properties, fuel properties and process parameter optimization of non-edible and edible vegetable oils were investigated in detail. A two-step and single-step transesterification process was used to produce biodiesel from high free fatty acid (FFA) non-edible oils and edible vegetable oils, respectively. This process gives yields of about 90-95% for J. curcas, 80-85% for P. glabra, 80-95% for canola, and 85-96% for corn using potassium hydroxide (KOH) as a catalyst. The fuel properties of biodiesel produced were compared with ASTM standards for biodiesel.     

Oleochemicals as a fuel: Mechanical and economic feasibility

The status of vegetable oils as diesel fuel substitutes is currently dubious. Although it is fair to consider them as short-term emergency fuels (or, more desirably, low proportion supplements to diesel fuels), they present mechanical problems in long-term use that have not yet been solved. It is preferable to use these oils blended in small proportions with diesel fuels. Indirect-injection diesel engines have had fewer problems than direct-injection engines, whether the tests were performed with pure vegetable oil fuel or with vegetable oil/diesel fuel blends. The economic prospect for these fuels is not promising. In general, they are not and have not been economical alternatives to diesel fuel. Exceptions appear to have occurred recently in Brazil and the Philippines where low local prices for vegetable oils combined with high petroleum prices encouraged officials to use low proportion vegetable oil/diesel fuel blends. Nonetheless, current and long-term trends in petroleum and oilseed prices indicate that these fuels will probably not be price competitive within the near future. Emergency disruption of petroleum supplies completely changes the economic situation. Vegetable oils would be worth much more as a fuel during disruptions than otherwise; thus incentives could be strong to include these oils in the fuel supply, diverting them from the food supply.     

Determination of Mineral Oil-Saturated Hydrocarbons (MOSH) in Vegetable Oils by Large Scale Off-Line SPE Combined with GC-FID

A method based on an off-line large-scale solid phase extraction (SPE) approach combined with conventional gas chromatographic-flame ionization detection (GC-FID) was developed to determine the mineral oil-saturated hydrocarbons (MOSH) in vegetable oils. A large-scale SPE column loaded with 10 g of activated silica gel impregnated with 1% silver nitrate which was used to retain lipids and olefins in vegetable oils and the MOSH in the oil samples was eluted with hexane. Then 2 μL concentrated solution was splitlessly injected into a common GC-FID instrument. The quantification limit reached 2.5 mg/kg when the MOSH fraction was concentrated to 0.1 mL. The accuracy of this procedure, as assessed by measuring the recoveries from spiked oil samples, was higher than 80%. This procedure was applied to analyze the MOSH in 38 commercial vegetable oils from Chinese market, which was the first survey of mineral oil contaminant in Chinese edible oils. The oil samples contaminated with different levels of MOSH, among which, 15 samples contained no mineral oils and 3 samples were contaminated with more than 50 mg/kg of MOSH. The highest contamination level was found in one of rice oils, in which the concentration of MOSH was up to 713.36 mg/kg. Of the 9 types of oils analyzed, camellia oil contained MOSH ranging between 6.76 and 78.49 mg/kg, averaging 46.72 mg/kg, indicating a higher contamination level than other types of oils. The results suggested that it is necessary to routinely detect mineral oil contamination in vegetable oils for food safety.     

Determination of solid fat content (SFC) of binary fat blends and use of these data to predict SFC of selected ternary fat blends containing low-erucic rapeseed oil

Several oils and fats often used for the industrial preparation of European shortenings were blended in binary systems. The equilibrium (after 48 h at 15°C) solid fat contents (SFC; determined by pulsed NMR spectroscopy) were measured and plotted against blend composition. SFC of the blends resulted from the SFC of each fat for the considered temperature as well as the type of interaction existing between those fats (namely, ideal behavior, monotectic interaction, eutectic interaction, and so on). The type of relationship fitted was dependent on the kind of interaction: Linear relationships were found for total compatibility between fats, and polynomial-type (order 2) relationships were found for fats exhibiting incompatibility. Some corresponding ternary oils and fats blends were also prepared and analyzed. Selected relationships (regression equations of the fitted curves) obtained for binary blends were combined in order to calculate the SFC of the corresponding ternary blends. Experimental values were generally close to predicted ones. The representation of SFC as a function of composition is interesting as it allows one to determine rapidly and easily the type of molecular interaction between two fats and also to determine equations that can be combined to calculate easily the SFC of corresponding ternary blends crystallized in the same way with a good accuracy. The texture (hardness) of several binary and ternary blends was also measured. The combination of the results obtained for SFC with the results obtained for the hardness of binary blends allows the prediction of the hardness of a corresponding ternary blend under the same conditions.     

罐式批次成品汽油调和配方集成建模方法

考虑罐式批次成品汽油调和过程中罐底余油对产品质量指标的影响,本文提出了一种基于改进多核模糊C均值（multi-kernel fuzzy C-means,MKFCM）与极端梯度提升树（extreme gradient boosting,XGBoost）集成的成品汽油调和的通用配方建模方法。该方法首先考虑各调和组分的差异性,提出一种自适应核参数计算方法对MKFCM改进,并将其用于罐底油聚类分析,旨在最大程度划分出性质相近的罐底油批次类型;在此基础上择优选用XGBoost算法,以各批次罐底余油的组分比例和产品预期质量指标作为输入,建立各批次子配方模型;在配方生成时,基于改进MKFCM求得当前罐底余油的隶属度向量,并以此为权值对子配方模型进行加权融合,最终以多模型集成的方式得到了成品汽油调和的通用配方。经使用某企业实际工业数据进行实验分析,结果表明,较单一模型或未改进MKFCM的集成模型,基于改进MKFCM-XGBoost的多模型集成配方,预测精度和泛化能力均更优,更适合罐式批次成品汽油调和过程。     

黑种草籽油的超临界CO_2萃取及其GC-MS分析

以新疆瘤果黑种草籽为原料,采用超临界CO2萃取技术(SFE-CO2)研究了瘤果黑种草籽油的萃取工艺,并对其化学成分进行了GC-MS分析。得到瘤果黑种草籽油较适宜的工艺条件为:萃取压力20 MPa,萃取温度35℃,萃取时间2 h,CO2流量20 kg/h。在该工艺条件下,黑种草籽油的得油率达36.33%。GC-MS检测出6种脂肪酸成分,主要为不饱和脂肪酸,其中,亚油酸质量分数60.95%,油酸质量分数20.54%,8,11-二十碳二烯酸质量分数2.43%,不饱和脂肪酸的质量分数近84%。