Application of steam explosion in oil extraction of camellia seed (Camellia oleifera Abel.) and evaluation of its physicochemical properties, fatty acid, and antioxidant activities

Abstract This study evaluated the physicochemical properties of oils extracted from steam‐exploded camellia seed (Camellia oleifera Abel.). Steam pressure, resident time, fatty acid composition, total phenolics, tocopherol, squalene, and sterol contents, and volatile compounds were determined. 1H NMR and FTIR spectra were performed for the structure of camellia seed oil. This study has found the highest yield of oil was 86.56% and was obtained when steam explosion pretreatment was at 1.6 MPa 30 s. Oil extracted by steam explosion pretreatment exhibited favorable physicochemical properties and stronger antioxidant activities compared to untreated oil. The compositions of fatty acid were similar between treated and untreated camellia seed oil. According to the 1H NMR and FTIR analyses, the functional groups of the oils were not significantly affected by the steam explosion pretreatment. Furans such as 2‐pentyl‐furan, 2‐furanmethanol, and 3‐methyl‐furan were produced from stream‐exploded camellia seed. Scanning electron microscopy revealed that steam explosion pretreatment efficiently promoted the release of oil by destroying the cell structure of camellia seed. Therefore, steam explosion can be an effective method for the camellia seed oil extraction.

and beneficial bioactive compounds such as sterols, squalene, and tocopherols (Xiao et al., 2017), which have been characterized as having health-promoting effects include lowering blood pressure, atherosclerosis, and scavenging free radicals (Lee & Yen, 2006). It also exerts potent antiulcer effects against oxidative damage in the stomach and intestine induced by ketoprofen (Cheng et al., 2014).
Also, camelliaa seed oil could be a valuable raw material and functional product to the food industry. For example, camellia oil is used as a natural antioxidant to improve the stability of many oil, and it effectively delays the development of pericarp browning and the loss of red color in litchi fruit because it contains high levels of antioxidants and vitamins such as phenolics .
Oil extraction from Camellia oleifera seed is a difficult process and camellia seed contains tea saponin, which is a foam-stabilizing and excellent emulsifying agent (Zhang, Han, et al., 2012;Zhang, Zhang, & Chen, 2012). Because of this, the difficulties of separating oils have made it challenging to provide an efficient yield. Until now, many techniques such as organic solvent extraction (Lee & Yen, 2006), subcritical water extraction (Wu et al., 2018), ultrasoundassisted (Wu & Li, 2011), microwave puffing-pretreated (Zhang & Jin, 2011), microwave puffing and aqueous enzymatic extraction (Zhang, Han, et al., 2012;Zhang, Zhang, & Chen, 2012), and aqueous enzymatic process (Fang et al., 2016) have been employed to extract oil from camellia seed. However, among them are used of chemicals (such as, Petroleum ethe and hexane) which can lead to environmental pollution and may require additional processes to remove or neutralize the chemicals used. Aqueous enzymatic extraction is a promising method and has been developed and performed in the laboratory which can improve the free fatty acid, vitamin E, and squalene contents, and physicochemical properties of oil (Fang et al., 2016). However, aqueous enzymatic method comes at a high cost. Another reason is that the aqueous enzymatic method cannot produce the desired aroma of oil. Consequently, searching for an efficient technique with low cost and high extraction of oil is vitally important for the extensive utilization of camellia seed.
Steam explosion is an innovational, economical, and effective method, which is explored extensively used for the pretreatment of cellulose (Deepa et al., 2011), hemicellulose , and lignin (Chang et al., 2012). The steam explosion has also been used to extract protein (Zhang, Yang, Zhao, Xiao, & Zhang, 2013), flavonoids (Song et al., 2014), polyphenols (Chen & Chen, 2011a), antioxidant compounds (Gong, Huang, & Zhang, 2012), and ethanol (Rocha, Gonçalves, Oliveira, Olivares, & Rossell, 2012). The principle of the treatment is the use of steam hydrolysis at high temperature and pressure, followed by sudden decomposition of the materials biomass with low molecular weight substance produced. In a steam explosion process, high saturated steam pressure is rapidly released to ambient, in 0.00875 s; meanwhile, high temperature decreased quickly and cooling the materials in the process. This is different from other normal thermal pretreatment (Gong et al., 2012). Compared with other pretreatments, the advantages of steam explosion include a significantly lower energy cost and environmental friendly (Zhang et al., 2013). In recent years, this technology has been used for the extraction process in studies regarding oil extraction, such as that of Ni, Zhao, Zhang, Gasmalla, and Yang (2016) which reported that steam explosion is a highly effective method for extracting oil from corn germ. And used steam explosion to extract oil from sumac fruit, their results showed that the oil yield at equilibrium increased to 16.04%, approximately fourfold higher than that of the raw sample (Chen & Chen, 2011b). Also, the oil extraction yield was improved, and the mass transfer coefficient according to the kinetics of the oil extraction of sesame seed was decreased in sesame seed (Sesamum indicum L.). In other words, steam explosion enhanced the oil extraction efficiency and it has proven to be an efficient method for releasing oil from vegetable seed.
However, using steam explosion in oilseeds for enhancing free oil yield is exceedingly limited, the application of steam explosion pretreatment for camellia seed, to the best of our knowledge, has not yet been reported. Therefore, the fatty acid compositions, physicochemical properties, total phenolic, total tocopherols, squalene, and sterol contents are evaluated and determined using 1 H NMR and FTIR. In addition, the microscopic structures of material before and after steam explosion are observed to clarify extraction mechanisms.

| Materials and reagents
The camellia seeds (Camellia oleifera Abel.) were provided by a local forest farm (Haikou country, Hainan province, China). The raw material (dry camellia seeds) contained 9.38% moisture content. The seeds were sealed in plastic containers and stored in a refrigerator at 4°C until extraction.

| Steam explosion and oil extraction
All steam explosion experiments were carried out in the QBS-80B SE device with a 0.4 L chamber from Gentle Bioenergy, Henan, China. 300 g of camellia seeds was placed inside the vessel and exposed to the saturated steam. The steam pressures were set at 0-2.3 MPa.
The resident time was performed in the range of 0-120 s, and finally termination by explosive decompression. The exploded materials were collected and dried in a ventilated drying oven for 5 hr at 50°C.
The dried camellia seeds were ground in a high-speed medicine grinder (Y-800, Kemanshi, China). The ground camellia seeds were then coarsely grounded, after passing through a set of standardmesh sieves (40 mesh), the milled camellia seeds were stored in 4°C fresh keeping cabinet until used for the extraction experiments.
Camellia seed oil was extracted from milled camellia seed using an aqueous as solvent. For each extraction, 5 g of camellia seed powder and extraction solvent of specified volume (solid-liquid ratio was set as 1:4.5 [w/v, kg/L]) were added into a flask, the pH value of system was adjusted to 9.0 using 1 M NaOH and 0.5 M HCI solution, and shaking horizontally was done for 2.5 hr at 75°C in a rotary shaker. The mixture was centrifuging at 6986 g for 10 min, then it was kept frozen at −20°C for 12 hr and thawed at room temperature. After demulsification, the slurry was centrifuging at 4,000 rpm for 10 min. The oil content of the entire fruit and its isolated parts was determined by Soxhlet extraction using petroleum ether.
Three replicates of the experiments were conducted. The amount of extracted oil was calculated gravimetrically after collection, and the free oil yield is expressed as follows:

| Physicochemical properties
The acid, peroxide, iodine, and saponification values were deter-

| Determination of the fatty acid compositions
To analyze the fatty acid composition of camellia seed oil, the oil (60 mg) was firstly converted into fatty acid methyl esters (FAME) using 2 ml NaOMe (0.5 M). The mixture was bathed in water for 30 min at 65°C, then 2 ml methanolic boron trifluoride (15%) was added in the mixture, continuously bathing for 5 min. Lastly, 1 ml saturated sodium chloride solution and 4 ml n-hexane were added immediately followed by vigorous shaking for 30 s. The anhydrous sodium sulfate was added; after stratification, the upper isooctane layer (1 ml) was filtered at 0.45 μm.
The fatty acid composition of oil was determined with a gas chromatograph (Agilent 7890B) equipped with a FID and a HP-5 column (30 m, 0.32 mm i.d., 0.25 μm film thickness; Supelco, USA). The nitrogen was used as a carrier gas at a flow rate of 1 ml/min. Sample was injected (1 μl) with a split mode (ratio 10:1). Injector temperature and detector temperature were set at 250 and 300°C, respectively.
Oven temperature increased from 100°C (1 min) to 190°C at a rate of 5°C/min and was further increased at a rate of 1°C/min to a final temperature of 220°C. Fatty acids were identificated with retention times obtained from commercial FAME standards (Sigma Chemical, St. Louis, MO). All experiments were carried out in triplicate sets.
The relative amount of each fatty acid was calculated from the integrated area of each peak and expressed as a percentage of the total area of all peaks.

| Total tocopherols (TT) and total phenolics (TP) contents
The TP content of camellia seed oil was extracted by methanolwater solution (80%:20% v/v) and determined by Folin-Ciocalteu method according to the colorimetric method described previously by Delfan-Hosseini, Nayebzadeh, Mirmoghtadaie, Kavosi, and Hosseini (2017). A calibration curve of gallic acid in methanol was carried out in the concentration ranges of 0.04-0.40 mg/ml. The results were expressed as μg gallic acid equivalent per gram of oil samples. Triplicate test was performed for each sample.
Tocopherols content of the extracted oils was determined using UPLC method with fluorescence detection. Waters liquid chromatography system equipped with a column heater, a photodiode array detector ACQ-FLR, controlled by Waters Empower chromatographic software. In all analyses, an Acquity UPLC Waters BEH C18 column of 1.7 μm (2.1 × 50 mm) was used. The analysis was carried out at 35°C temperature under isothermal condition, and the mobile phase was composed of 100% acetonitrile. The volume of injection is 10 μl, and the flow rate was 0.5 ml/min. Using FLR to detect and quantify α-tocopherol, the excitation wavelengths and emission wavelengths are 294 nm and 338 nm, respectively. A calibration curve of α-tocopherol in toluene was performed in the concentration ranges of 0-300 mg/ml. Results were expressed in mg of α-tocopherol per kilogram of oil.

| Determination of squalene and sterol contents
The content and composition of the sterols and squalene were determined by gas chromatography (GC) following procedure reported by Wang et al. (2017) with some modification. Camellia seed oil (1.5 g) was saponified with 50 ml of 1 M methanolic potassium hydroxide at 85°C for 1 hr in a reflux condenser. After cooling, 50 ml of saturated sodium chloride was added, and the unsaponifiable matter was extracted with 3 × 50 ml of n-hexane. The combined n-hexane fractions were washed 3-5 times with distilled water (30 ml). Anhydrous sodium sulfate was added to the n-hexane layer (in order to eliminate aqueous residues), the organic layer was then evaporated at 40°C on a rotary evaporator, and the residue was redissolved in 5 ml of n-hexane. Finally, the extract was filtered through a 0.45 μm syringe filter and stored at −20°C until analysis.
The samples were analyzed on a GC (Agilent 7890B) equipped with a FID and a HP-5 column (30 m × 0.32 mm × 0.25 μm; Supelco, USA). Helium was used as the carrier gas at a flow rate of 1 ml/min. A sample of 1.0 μl was injected in a splitless mode with an injector temperature of 250°C. The column temperature was held at 100°C for 5 min, then increased to 180°C at a rate of 25°C/min, held for 1 min, then further increased at a rate of 10°C/min to a final temperature of 280°C. The detector temperature was set at 300°C.
According to the retention times of reference samples of sterols (5α-cholestane) and squalene, results were expressed as mg/100 g of oil. All samples were analyzed in triplicate and means of the results are reported.

| Antioxidant activity evaluation
The seed oil obtained under the optimum conditions was subjected to screening for its possible antioxidant activity. The antioxidant were expressed by IC 50 values which corresponded to the concentration of oil (mg/ml) neutralizing 50% of DPPH radicals.

| Scanning electron micrographs (SEM) observation
In order to investigate the influence of steam explosion, microstructure observations of raw and the optimum conditions camellia seed oil were carried out via SEM (JEOL JSM-7500F, Tokyo, Japan).
Samples were dried, fixed, and coated by gold, and then examined under high vacuum condition at an accelerating voltage of 10.0 kV (20 μm, 1,800 magnification).

| Acquiring 1 H NMR spectra analysis
1 H NMR spectroscopy was used to obtain maximum possible information on positional distribution of fatty acids in camellia oil.
15-25 mg of sample was dissolved in a mixture of 0.5 ml of CDCl 3 .
These mixtures were thoroughly mixed well and then transferred to NMR tubes for 1 H NMR spectra analysis. The spectra were recorded at 25°C on a Bruker Av500 NMR spectrometer (Bruker BioSpin GmbH, Germany, 1 H frequency 500.13 MHz) equipped with inverse detection ( 1 H-13 C-15 N) system. The resulting spectra were processed using Bruker Topspin 3.5 software (Bruker Biospin, Rheinstetten, Germany). The data were processed without zero-filling and by using exponential multiplication with a line-broadening of 1.0 Hz.

| Acquiring Fourier transform infrared (FTIR) spectra
A FTIR spectrophotometer (AVATAR 370 FIR, Thermo Nicolet) was utilized to record the percent transmittance in the absorption mode 400 to 4,000 cm −1 at a resolution of 4 cm −1 . A small amount (3-5 μl) of the extracted oil sample was deposited between the two well-polished KBr pellet and the Pasteur pipette was used to create a thin film.

| Identification of volatile compounds
Volatile compounds were determined using a GC (Agilent 7890B, Heilbronn, Germany) coupled to a mass spectrometer (Agilent 5974, Heilbronn, Germany) and a Headspace sampler (Agilent 7697A, Heilbronn, Germany). 10.00 g of camellia oil was introduced into 20-ml headspace vials of headspace analyzer. The vials were sealed air-tight with a silicone/polytetrafluoroethylene (PTFE) septum.
Samples were subjected to dynamic headspace for 30 min at 200°C.
Chromatographic separation was performed on an HP-5 column (30 m × 0.32 mm × 0.25 μm; Supelco, USA). Helium (purity 99.99%) was used as a carrier gas at a constant pressure of 16 psi. Samples were injected in a splitless mode. The temperature program was as follows: 3 min at 40°C, first ramp 5°C/min to 250°C, and total analysis run was 60 min. The mass spectrometer was operated in electron impact (EI) ionization mode at 70 eV using full-scan mode from m/z 25 to 550. Source and quadrupole temperatures were 230 and 150°C, respectively.

| Statistical analyses
All experimental measurements were conducted at least in triplicate and data are expressed as mean ± standard deviation, where feasible. The data obtained in this study were analyzed by one-way analysis of variance (ANOVA) using SPSS 16 (SPSS Inc., Chicago, IL).
Statistical significance was considered at the 5% level (p < 0.05).

| Effect of steam explosion conditions on oil extraction
The effect of steam explosion at 30 s for different pressure on camellia seed oil extraction is shown in Figure 1a. Steam explosion treatment had the highest oil yield of 86.56% at 1.6 MPa from the yield of 68.06% for untreated sample. Simultaneously, free oil content had gradually improved with the increase of pressure from 0 to 1.6 MPa, this is due to the fact that in some pressure range, steam explosion could destroy the cell thoroughly and enhance the porosity, which released the oil surrounded in the cell and decreased the oil in the seed cake. However, after 1.  of steam explosion which has a greater impact on the oil yield. In Figure 1b, the free oil yield peaks was showed at 30 s. However, there is no significant difference between 30 and 60 s, but free oil yield at other resident time was significantly lower than 30 s. That is because higher pressure and resident time may urge protein denaturation and aggregation . Oil droplets were wrapped in aggregated protein and are difficult to release. However, higher pressure and resident time could improve the content of tea saponin-one of the emulsifiers, which would increase emulsification and prevent the oil from the emulsion after centrifugation (Fang et al., 2016). In addition, the color of oil from steam-exploded camellia seed was slightly darker than that of the untreated camellia seed oil. The

| Physicochemical properties
To understand the effect of steam explosion on the physicochemical properties of camellia oil, the acid, peroxide, iodine, and saponification value were analyzed, which are shown in Figure 2a As shown in Figure 2a, acidity decreased with different steam pressure pretreatment, and all treated sample's acidity was lower than raw materials (1.86 ± 0.08 mg KOH/g). The changes of peroxide value also show a similar trend, and all steam-exploded sample's peroxide value was lower than raw materials (4.81 ± 0.14 meq/kg).
As for different resident time (Figure 2b), both acid and peroxide value showed a peak at 1.6 MPa 30 s, 0.92 ± 0.09 mg KOH/g and 2.15 ± 0.17 meq/kg, respectively, still lower than untreated seed oil. In theory, oil extracted from pretreated oilseeds at high temperatures is prone to hydrolysis, which will lead to higher acid value than that of raw sample. The lower acidity of the camellia oil in the steam-exploded sample may be due to the short retention time and evaporation of free fatty acids in the oil at the moment of explosion. Acid value of steam-exploded sumac fruit also decreased with corresponding pressure 1.3-1.5 MPa (Chen & Chen, 2011b). And, Timilsena, Vongsvivut, Adhikari, and Adhikari (2017) also reported that the low acid and peroxide value indicated that chia seed oil embodied lower quantities of oxidation by-products and free fatty acids including hydroperoxides and aldehydes. Oil sample of low acid value and peroxide value can be regarded as good quality.
The effect of different steam pressure on iodine value of camellia is presented in Figure 2a. The iodine value of steam-exploded samples increased with the increase of pressure from 0 MPa to 1.6 MPa, after which, it decreased to 2.3 MPa. Also, as for different resident time which showed in Figure 2b, the iodine value showed a peak at 30 s (95.97 ± 1.37 g I 2 /100 g) higher than other resident time. All steam explosion treated samples extracted an oil higher in iodine value compared with untreated samples (82.13 ± 0.57 g I 2 /100 g), indicating a higher degree of unsaturation fatty acids (Jiao et al., 2014), which agreed with a high oleic acid content (Zhang & Jin, 2011).
The saponification values indicated short-chain fatty acids levels in vegetable oils (Timilsena et al., 2017). As illustrated in Figure 2a,b, the saponification values of the camellia seed oils ranged from 176.32 ± 16.62 to 213.29 ± 21.01 mg KOH/g oil, however, had no significant difference was found between the saponification value and steam pressure or resident time, which showed that steam explosion has no significant influence on the levels of short-chain fatty acids. all of them were present in the highest amount (13.96 ± 0.08 μg GAE/g oil, 507.85 ± 17.62 mg/kg, 188.34 ± 11.46 mg/100 g and 253.52 ± 19.02 mg/100 g) at 1.6 MPa 30 s. The amount of these four bioactive compounds in seed oil treated with other steam explosion pressures are less than 1.6 MPa 30 s, but higher than that of without steam explosion pretreatment extraction (5.00 ± 1.02 μg GAE/g oil, 319.08 ± 21.59 mg/kg, 162.38 ± 9.67 mg/kg and 186.69 ± 40.10 mg/100 g). As shown in Figure 3b, the content of total phenolics, α-tocopherol, and squalene at 30 s was higher than that of other resident time treatments. Although the amount of sterols showed a peak at 15 s, there is no significant (p < 0.5) difference between 15 s and 30 s for all the content of them higher than untreated seed. These results may explain the fact that steam explosion pretreatment destroyed the intact cellular structure of oilseed, increased the release of α-tocopherols, total phenolics, squalene, and sterols in short time, thereby enhancing the bioactive components in the extracted oil as shown in Figure 3a,b.

| Total tocopherols, total phenolics, squalene, and sterol contents
However, if the resident time of steam explosion is too long or the steam pressure is too high, steam explosion pretreatment will reduce the α-tocopherols, total phenolics, squalene, and sterol contents of extracted oils. This may be due to the fact that bioactive compounds are easily decomposed to some extent by either relatively long time or exposure too extreme high pressure saturated steam and temperature.

| Fatty acid composition of oils
The percentages of the fatty acids of oil from camellia seed were presented in Table 1. It can be observed that the tested oils con- eleven fatty acids were analyzed by GC/MS, and the oleic acid content was 75.05 ± 0.15%, which were lower than those in untreated and steam-exploded seed. Wang et al. (2017) also found that compared with Oleifera and Camellia oils, there was a significant difference in the content of oleic acid, linoleic acid, and palmitic acid in thea oil. These dissimilarities could be ascribed to different causes, such as genotype, growing condition, method of extraction (Huang, Wang, & Liang, 2015). In addition, the refining conditions for camellia oil, such as different temperatures, can also have a great influence on the type and content of camellia seed fatty acids (Wei et al., 2015).

| Antioxidant activity
Antioxidants could scavenge the free radicals of oil and inhibit the chain reaction and thus prevent the oxidation of lipid (Samaram et al., 2015). The antioxidant activities of camellia seed oils pretreatment with steam explosion or not were assessed using DPPH radical-scavenging assay. Oil with steam explosion pretreatment (1.6 MPa 30 s) exhibited significantly (p < 0.05) superior efficacy in scavenging the DPPH radicals (24.65 ± 1.23 mg/ml) compared with untreated oils (29.74 ± 1.09 mg/ml). A variety of bioactive components of camellia seed oil have the major effect on its oxidative stability behavior, such as tocopherols, total phenolics, squalene, and sterols (Ma et al., 2011). And other study has shown that the composition of fatty acids plays a vital role in the antioxidant activity of oil, especially the high level of unsaturated fatty acids . The antioxidant activities of camellia seed oils were improved mainly because steam explosion pretreatment enhances the content of tocopherols, total phenolics, squalene, sterols, and unsaturated fatty acid, resulting in a much higher availability of such bioactive components into oils. These results were in agreement that there was a positive correlation between the content of these bioactive components and antioxidant activity of oils.
Methanol extract of tea seed oil showed higher antioxidant capacity by improving the content of total phenolics, but the transparency, odor, and flavor of solvent extracted oil were not satisfactory (Long et al., 2012).

| Analysis of microscopic changes
To gain further insight into the effect of the steam explosion pretreatment on the oil extraction from camellia seed and to understand the extraction mechanism, the camellia seed powder was examined by SEM to elucidate the morphological changes of steamexploded and raw camellia seed. The obvious effect of steam explosion pretreatment on the morphology of camellia seed could have been observed in Figure 4. Untreated camellia seed still possesses relatively complete structures, regular or compact shapes, and smooth surfaces, which is not conducive to the release of oil from the seed (Figure 4a). On contrary, steam-exploded seed surface became rough and fissured and the morphology of samples was obviously broken down ( Figure 4b). During steam explosion process, high temperature and pressure could cause physicochemical modifications structural, resulting in fissures and cavities, which in agreement with the previous study (Chen & Chen, 2011a,b). As a result of the reduced particle size, mass transfer resistance, and the generation of micropores, surface area increases giving the high oil yield. This result was consistent with the above-mentioned high oil yields in the extraction process of oil from steam-exploded camellia seed also performed steam explosion pretreatment which helped increase sesame seed oil yield at equilibrium. The SEM results indicated that the steam explosion technique efficiently promoted the release of oil by breaking down the cell structure of camellia seed.
Until now, no study on the cell structures of steam explosion pretreatment camellia oil seed samples has been reported.

| 1 H NMR spectroscopy
The 1 H NMR spectra of the extracted oil from steam explosion and untreated camellia seed are presented in Figure 5. The majority of the characteristic peaks were observed in the range of 3.0-0.5 ppm.
The region between 0.5 and 5.5 ppm in both oils contains all the -(CH 2 )n-(acyl chains) were identified at 1.22-1.30 ppm. The peaks at 1.61 ppm were due to β-carbonyl methylene protons, while those found at 2.29-2.34 ppm were related to α-methylene protons (Sherahi, Shahidi, Yazdi, & Hashemi, 2017). A triplet centered around 2.25-2.79 ppm due to methylene protons in the carbonyl.
The peaks at 4.12-4.32 ppm were related to the hydrogen atoms on 1 and 3 carbon atoms of the glyceryl methylenes. The multiples at 5.27-5.37 ppm were attributed to olefinic hydrogen atoms of the different acyl groups. Otherwise, the peak at 7.26 ppm was attributed to CDCl 3 .
As it can be observed from Figure

| FTIR spectroscopy
FTIR spectroscopy can be used to monitor structure changes in oils and fats and reveal the characteristic peak that specifically represents unsaturated fatty acids (UFAS) in the observed spectrum. To investigate whether steam pressure pretreatment produces structural changes to camellia seed oil, FTIR analysis was performed. The representative FTIR spectra of steam explosion and untreated camellia seed oil are presented in Figure 6. In general, the small band at 3,503 cm −1 is due to the overtone of the glyceride ester carbonyl absorption.

| Volatile compounds
Flavor is an important quality criterion for camellia oil. The untreated camellia oil imparted a slight odor with oily and botanical flavor notes, while the steam-exploded camellia oil presented an attractive nutty and caramel-like odor. Therefore, identification of flavor compounds is a demand for quality control of camellia oil production. To reveal the difference in flavor notes, HS GC-MS was conducted to identify the volatile compounds in untreated and steam-exploded camellia oil. Volatile compounds are considered to be major contributors to the overall flavor characteristics (Zhang, Wang, Yuan, Yang, & Liu, 2016). About 40 volatile compounds were identified and characterized, including hydrocarbons, esters, aldehydes, acids, alcohol, naphthalene, benzene derivatives, and furans ( Table 2). The profiles of volatile compounds were changed in steam explosion process.
Compared with untreated camellia oil, some new minor volatile compounds were generated, such as p-xylene, 1,3-dimethyl-benzene, esters, and furans. However, the main volatile compounds were still hydrocarbons.
Furan derivatives were a new class of volatile compounds in camellia seed oil after steam explosion. During thermal processing of foods, furans are generated by the Maillard, lipid oxidation/degradation, and caramelization reactions, which contributes to the fruity, sweet, and nutty characteristics of heated foods (Zou, Gao, He, & Yang, 2017).
Three furans were qualified as 2-pentyl-furan, 2-furanmethanol, and 3-methyl-furan in this study. 2-furanmethanol (with an earthy, mild sweet and oily odor) was the unique volatile compounds in treated oil, which is mainly formed from pentoses in thermal reactions . The presence of 2-furanmethanol in camellia oil after steam explosion implies that Maillard reaction occurred during steam explosion.

CO N FLI C T O F I NTE R E S T
There are no conflicts of interest to declare. TA B L E 2 (Continued)