Materials and methods
Materials and chemicals
The refined peanut oil and degummed crude soybean oil were obtained from Kaifeng Longda Vegetable Oil Co. (China). The stainless steel oil storage tanks are manufactured by Henan Sunshine Oil and Fats Co. (Zhengzhou, China). TBHQ was purchased from Guangdong Food Industry Research Institute (purity≥99.0%);fatty acid methyl ester standards, β -sitosterol (purity≥99.0%),stigmasterol(purity≥95.0%),campesterol (purity≥99.5%),Sitosterol(purity≥99.0%),α -cholesterol (purity≥95.0%), N,O-bis-(trimethylsilyl)trifluoroacetamide (BSTFA) containing 1% trimethylchlorosilane (TMCS) was purchased from Sigma-Aldrich (St. Louis, MO) and used without further purification; Tocopherols (α -, β -, γ - and δ -) standards (purity≥99.0%)Sigma-Aldrich (St. Louis, MO). N-hexane, 1-Methylethanol were HPLC grade which were purchased from VBS (St. Louis, MO, USA), potassium hydroxide (KOH), Chloroform (CHCl3) and all other solvents and chemicals used in this study were of analytical grade and used without further purification.
Experimental design
A temperature real-time detection recorder is set up on the above ground tanks and the underground tanks. The analog data received from the temperature probes were converted to digital data by the YA200R recorder and stored in flash memory. The YA200R recorder has an LCD sensor status display for real-time visualization (Fig. 1).
Oil was mixed in a 800 L stainless steel tank and equally distributed to the four 200 L stainless steel tanks. The headspace volume was equal to 10% of the total volume of the each tank. Four different conditions: static atmosphere tank, added TBHQ tank, nitrogen-filled tank and underground tank.
Tank A and E was not filled with N2 and stored outdoors (static atmosphere tank) with peanut and soybean oil; Tank B and F was not filled with N2 and stored outdoors, but was added 0.02% TBHQ of oil weight storage (added TBHQ tank) with peanut and soybean oil; Tank C and G was was filled with N2 and stored outdoors (nitrogen-filled tank) with peanut and soybean oil; Tank D and H was was not filled with N2 but stored underground (underground tank) with peanut and soybean oil.
Analysis of fatty acid composition
Fatty acid methyl esters (FAMEs) were prepared according to the method [18] described by GC analysis was performed on an Agilent 7890B Gas Chromatograph (Agilent Technologies, USA), equipped with a flame ionization detector and a HP-88 capillary column (100 mnologies,0.20 μm, Agilent Technologies, USA). The column temperature program was as follow: the initial column temperature was 140 ℃, and it was maintained for 5 min, and then from 140 ℃ to 240 ℃ at 4 ℃/min, and held for 10 min at 240 ℃. The injector temperature was 250 ℃ and the detector temperature was 280 ℃. Nitrogen was used as the carried gas at a flow rate of 1.0 mL/min, and a split ratio of 50:1 was used. The fatty acids peaks were identified by comparison of retention times with those of standards. Fatty acid methyl esters were quantified as percentages of the total methyl ester peak areas. Eight fatty acids were considered in this study. They are palmitic acid (C16:0), stearic acid (C18:0), oleic acid (C18:1), linoleic acid (C18:2), linolenic acid (C18:3), arachidic acid (C20:0), behenic acid (C22:0) and tetracosanoic acid (C24:0) expressed as percentages of fatty acid methyl esters.
Analysis of phytosterol
The phytosterol content of the oils was determined following ISO 12228-1:2014 method [19] and was analyzed using gas chromatography (GC) system (Agilent, 7890, Santa Clara, U.S.A.), equipped with a flame ionization detector (FID). Separation of the sterols was performed using HP-5 column (30 m×320 μm×0.25 μm, Agilent J&W Scientific, Santa Clara, U.S.A.). Further parameters were as follows: nitrogen as carrier gas; split injection (1:20) was used, and the injection volume was 1 μl; flow rate was 1.0 mL/min; oven temperature was held for 20 min; injector temperature, ; detector temperature, . The compounds were quantified by adding an internal pattern (5α-cholestanol). All samples were analysed in triplicate and the sterol/oil ratio was expressed in mg/100g.
Analysis of tocopherols
0.5 g of oil was dissolved in 10 mL of HPLC-grade n-hexane, then filtered using a 0.45 μm organic membrane filter for HPLC analysis. The analysis of tocopherols and tocotrienols were performed using a Waters e2695 HPLC system equipped with a Waters 2475 fluorescence detector (Waters Corporation, Milford, USA) and an Elite NH2 column (250 mm×4.6 mm×5 μm, Dalian Elite Analytical Co., Ltd, Dalian, China). The column temperature was 35 °C, isopropanol and n-hexane (1:99, v/v), flow rate 0.8 mL/min. The excitation wavelength was 298 nm and the emission wavelength was 325 nm. The tocopherols and tocotrienols were identified by comparing their retention times with authentic standards and quantified based on the peak areas compared with the external standards. All samples were analysed in triplicate and the tocopherol/oil ratio was expressed in mg/kg.
Analysis of volatile flavor compounds
The headspace was sampled using a (50/30μm DVB/CAR/PDMS) SPME fiber and fiber assembly (Supelco, Inc., U.S.A.) for each peanut oil sample, allowing the fiber to equilibrate in the headspace of the 50-mL screw top vial heated by a water bath (Agilent, Santa Clara, U.S.A.) in order to acheive headspace equilibrium more quickly. Headspace equilibration of samples was determined by sampling the headspace using SPME over fixed periods of time. Equilibrium was reached after 30 min in a water bath. After the SPME fiber reached equilibrium, thermal desorption for 3 min in the injection port was required in order to remove all compounds on the fiber, as determined by preliminary studies. The volatile aroma compounds were analysed on Agilent 5977 Network-mass selective detector (MSD) equipped with an Agilent 7890B gas chromatograph (both Agilent Technologies, Inc., Palo Alto, CA, USA). Furthermore, The volatiles were separated by HP-5MS (Agilent J&W GC Columns., U.S.A.) Capillary Column (30m length×0.25mm i.d.×0.25μm thickness). Flavor compounds were separated by increasing the temperature from (3 min hold time) to (8-min hold time), ramped at per min. Injection port temperatures were . The transfer line temperature was 240 °C. The mass detector was operated at 230 °C in electron impact mode at 70 eV, and the ion source temperature was 230 °C. The chromatograms were recorded by monitoring the total ion currents (TIC) in a mass range of 30–500 m/z. Qualitative and quantitative analysis: The mass spectrometry information of each component was matched and qualitatively matched with the NIST 17 mass spectrometer library, and only the results of positive and negative matching degrees greater than 80 (maximum 100) were reported. The volatile silane impurities contained in the instrument itself are removed, and the relative contents of the various compounds are calculated by the peak area normalization method.
Analysis of Gas chromatography-olfactometry (GC–O)
A GC–O analysis was performed on an gilent 7890B gas chromatograph equipped with a sniffing port (Sniffer 9000, Brechbühler Scientific Analytical Solutions Inc., Switzerland).
A Gerstel ODP-3 sniffing port using deactivated capillary column (30 cm×0.3 mm) heated at 200 °C. This system which can obtain an MS signal for the identification and the odor characteristics of each compound detected by sniffing port at once. GC effluent was split 1:1 among the MSD and olfactometer via a switch. By smelling and recording the odour descriptions, 3 welltrained panellists performed the GC–O analysis. The perceived aroma intensity was evaluated and recorded by using the 1-9 scores. Each sample was sniffed twice by each panellist.
Physicochemical analysis
GB 5009.227-2016 [20] was used to determine the peroxide value (PV, milliequivalents oxygen (meq O2)/kg oil). GB 5009.299-2016 [21] was used to determine the acid value.
Statistical analysis
Statistical analysis each laboratory test on the oil was carried out in three replicates. Data were expressed as mean ± sd. Variables were compared using one-way ANOVA. Data processing was performed using the data analysis and graphing software of Origin 9.0 and Microsoft Office Excel 2013, Hierarchical Cluster ananlysis (HCA) was finished by IBM SPSS Statistics software (version 20, SPSS, Inc., Chicago, IL).