LWT - Food Science and Technology 111 (2019) 548–554

Contents lists available at ScienceDirect

LWT - Food Science and Technology

journal homepage: www.elsevier.com/locate/lwt

Comparative study on the extraction of Xanthoceras sorbifolia Bunge (yellow T

horn) seed oil using subcritical n-butane, supercritical CO2, and the Soxhlet
method
Ling-Biao Gua, Guang-Jie Zhanga, Lei Dua, Juan Dua, Kun Qia,d, Xin-Liang Zhua,d,
Xiao-Ying Zhanga,b,c,∗, Zhi-Hui Jianga,∗∗
a
Henan Joint International Research Laboratory of Veterinary Biologics Research and Application, Anyang Institute of Technology, 455000 Anyang, China
b
College of Veterinary Medicine, Northwest A & F University, 712100 Yangling, China
c
Centre of Molecular and Environmental Biology, Department of Biology, University of Minho, Campus de Gualtar, 4704-553 Braga, Portugal
d
Key Laboratory of Subcritical Extraction of Anyang City, Henan Subcritical Biotechnology Co. LTD, 455000 Anyang, China

A R T I C LE I N FO A B S T R A C T

Keywords: Xanthoceras sorbifolia Bunge seed oils are suitable for broad nutritional and bio-functional purposes. In this study,
Xanthoceras sorbifolia Bunge (yellow horn) seed the oil yield, physicochemical properties, fatty acid profile, Fourier transform infrared spectra, tocopherol
oil content, antioxidant activity, oxidative and thermal stability, and rheological properties of seed oils, obtained by
Subcritical n-butane extraction subcritical n-butane, supercritical carbon dioxide (CO2), and Soxhlet method, were investigated and compared.
Supercritical CO2 extraction
The results indicated that subcritical n-butane had higher yield (58.79%) than supercritical CO2 (56.47%),
Soxhlet extraction
corresponding to 97.29% efficiency of the Soxhlet method. All the extracted oils were rich in linoleic acid
Rheological behavior
(379.61–385.86 mg/g oil), and oleic acid (276.58–285.77 mg/g oil), with high thermal stabilities and Newtonian
flow behavior. The subcritical n-butane-extracted oil showed the highest tocopherol content, antioxidant ac-
tivity, and oxidative stability. Thus, subcritical n-butane extraction was the best method to extract yellow horn
seed oil and potentially other edible oils.

1. Introduction temperature process (Czaikoski et al., 2015).

Supercritical extraction has been introduced as a significant alter-
Xanthoceras sorbifolia Bunge (yellow horn) is native to northwestern native to conventional oil extraction techniques. Commonly, CO2 is
China and highly tolerant of cold, drought, salt, and starvation (Li et al., used as an extraction solvent under supercritical conditions because of
2013). Its edible seeds contain 55–65% oils that are good sources of its broad availability, non-toxicity, and environmental friendliness
unsaturated fatty acids (85–93%), especially linoleic and oleic acids. compared to organic solvents. However, supercritical CO2 is less ef-
High linoleic acid content is favorable for nutritional and medicinal fective than organic solvents in extracting lipids because of the low
applications, because of its myocardial preservation, antimicrobial, and solubility of triacylglycerides. Additionally, it has shortcomings such as
anti-diabetic activities (Szentmihályi, Vinkler, Lakatos, Illés, & Then, low processing capacity, and high equipment investment and produc-
2002). tion costs (Rai, Mohanty, & Bhargava, 2015). In contrast, n-butane
Edible plant oils are conventionally extracted using mechanical (critical pressure and temperature: 425.05 K and 3.79 MPa) dissolves
pressing or solvent extraction (Gu et al., 2016). However, mechanical triacylglycerides at low pressures and provides higher yields than su-
pressing may reduce the shelf-life of active compounds due to oxygen percritical CO2 (Ndiaye et al., 2006), leading to shorter extraction times
and light exposure, and has low extraction rates (Herrero, Cifuentes, & and higher oil yields using less solvent. Subcritical n-butane is more
Ibañez, 2006). Solvent extraction typically utilizes organic solvents, efficient in extracting active compounds such as tocopherols, phytos-
which leads to high costs due to high energy and water consumption, terols, and carotenoids than supercritical CO2 and conventional
and possible damage to bioactive ingredients due to the high- methods (Xu, Han, Zhou, Wu, & Ding, 2016). Additionally, it is much

∗
Corresponding author. College of Veterinary Medicine, Northwest A & F University, 712100 Yangling, China.
∗∗
Corresponding author. Henan Joint International Research Laboratory of Veterinary Biologics Research and Application, Anyang Institute of Technology, 455000
Anyang, China.
E-mail addresses: [email protected], [email protected] (X.-Y. Zhang), [email protected] (Z.-H. Jiang).

https://doi.org/10.1016/j.lwt.2019.05.078
Received 20 November 2018; Received in revised form 26 March 2019; Accepted 17 May 2019
Available online 17 May 2019
0023-6438/ © 2019 Elsevier Ltd. All rights reserved.
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

Abbreviation FTIR Fourier transform infrared

HPLC high-performance liquid chromatography
ANOVA analysis of the variance IT induction time;
FLD fluorescence detection MUFA monounsaturated fatty acid
DPPH 2,2-diphenyl-1-picrylhydrazyl PUFA polyunsaturated fatty acid
DSC differential scanning calorimetry SFA saturated fatty acid
DTF difference thermogravimetric SFE supercritical fluid extraction
FAME fatty acid methyl esters TG thermogravimetric
FID flame ionization detector UFA unsaturated fatty acid

safer due to the relatively lower pressure, lower equipment investment weight.
and lower operating cost than supercritical CO2 (Sun et al., 2018).
Notably, the application and usability of yellow horn seeds are still 2.3. Analytical methods
not well explored. In this respect, this study aimed at providing a
comprehensive and comparable analysis of the effects of subcritical n- 2.3.1. Determination of physicochemical properties
butane, supercritical CO2, and Soxhlet extractions on yellow horn seed The refractive index and specific gravity were measured at room
oil. Therefore, a quality assessment of the oils obtained from the three temperature (25 °C) using an Abbe 60 refractometer (Bellingham and
methods was also performed. Stanley Ltd, Tunbridge Wells, Kent, UK) and a 10 mL specific gravity
bottle (Shanghai Bomei Glass Instrument Co. LTD, Shanghai, China),
2. Materials and methods respectively. Standard American Oil Chemists’ Society methods (AOCS,
2009) Cd 3–25 (03), Cd 1d-1992 (97), Ca 6b-53 (01), Cd 3d-63 (03),
2.1. Sample preparation and Cd 8b- 90 (03) were used to determine the saponification value,
iodine value, unsaponifiable matter, acid value, and peroxide value,
Yellow horn seeds were harvested from Xi'an, Shaanxi Province, respectively.
China. The seeds were dried at 60 °C for 24 h in a hot air oven (FD115,
Binder GmbH, Tuttlingen, Germany). The shells were removed, and the 2.3.2. Determination of the fatty acid composition
kernels were ground and sieved for 20 min using a vertical vibratory The fatty acid compositions of the oil samples were analyzed using a
sieve shaker (Labortechnik GmbH, Ilmenau, Germany). The average gas chromatography system coupled to a flame ionization detector
particle size of the obtained mixture was 250–1000 μm. (FID) (7890A, Agilent Co., Santa Clara, CA, USA) and a HP-88 capillary
column (100 m × 0.25 mm × 0.20 μm, Agilent Co., Santa Clara, CA,
2.2. Oil extraction methods USA). Fatty acid methyl esters (FAME) were prepared by methylating
lipids using a previously reported procedure (Gu et al., 2016). The in-
2.2.1. Subcritical n-butane extraction jector and detector temperatures were set at 250 and 280 °C, respec-
Subcritical n-butane extraction was carried out using a laboratory tively. The temperature program was as follows: 1 min at 120 °C, in-
scale unit (Henan Subcritical Biological Technology Co., LTD, Anyang, creasing at 10 °C/min up to 175 °C, 10 min at 175 °C, increasing at 5 °C/
China) standardized in our previous study (Gu et al., 2016). For each min up to 210 °C, and finally 27 min at 210 °C. The peak areas were
extraction, powdered seed mixture (100 g) was transferred into the determined by the MSD ChemStation software. For qualitative analysis,
extractor, and 1000 mL of subcritical n-butane was added; the ratio of the retention times of the fatty acids were compared with those of
plant sample to n-butane was 1:10 (g/mL). The extraction was per- standard methyl esters. Quantitative analysis (mg/g) was performed
formed at 35 °C for 30 min over 3 cycles. At the end of extraction, according to Zanqui et al. (2016) using tricosanoic acid methyl ester as
subcritical n-butane containing the lipid was transferred to a separator, an internal standard.
and after the solvent had been removed by reduction vaporization, the
extracted oil was collected. 2.3.3. Fourier transform infrared (FTIR) analysis
The functional groups of the oils were analyzed using a Tensor II
2.2.2. Supercritical fluid extraction FTIR spectrometer (Bruker Optics, Billerica, USA). A droplet of oil was
Oil was extracted using a supercritical fluid extraction (SFE) device placed on a KBr tablet. Spectra were recorded at 4000–400 cm−1 at a
(HA121-50-01, Hua'an Research Instrument Co. Ltd., Nantong, China). resolution of 4 cm−1 using 64 co-added scans.
The extraction pressure (30 MPa), temperature (50 °C), and CO2 flow
rate (10 L/h) were controlled using regulating valves. 2.3.4. Determination of the tocopherol content
An aliquot of the oil extracted with methanol/tetrahydrofuran was
2.2.3. Soxhlet extraction evaporated under a nitrogen stream and re-dissolved in n-hexane/me-
Soxhlet extraction was performed using an SER 148/6 (Velp thyl tert-butyl ether (98 + 2, v/m). High-performance liquid chroma-
Scientifica, Usmate, Italy) solvent extractor. For each extraction, tography (HPLC) analysis was performed with an Agilent 1260 Series
ground dried seeds (10 g) were packed in a thimble and extracted with liquid chromatograph (Agilent Technologies, Santa Clara, USA)
n-hexane (60 mL). The immersion, washing, and recovery steps were equipped with a fluorescence detection (FLD) system and Kinetex-C18
performed at 70 °C and lasted for 1, 3, and 1 h, respectively. column (250 mm × 4 mm i.d., 7 μm particle size, CA, USA). The
The extraction conditions of the three methods were derived from column temperature was maintained at 35 ± 1 °C. The individual
preliminary experiments based on single factor experiments and re- compounds, i.e., α-, γ-, and δ-tocopherols, were identified and quanti-
sponse surface analysis performed in our laboratory. All the extracted fied by comparing the peak areas with those of external standards of
oils were collected and the residual solvent was removed utilizing a known concentrations (Tasioula-Margari & Okogeri, 2001).
rotary vacuum evaporator. Then the oils were weighed using an ana-
lytical balance (ME204, Mettler-Toledo GmbH, Greifensee, 2.3.5. Determination of the oxidative stability
Switzerland), and stored at 4 °C for further analysis. The yields were The oxidative stability was measured using an OXITEST apparatus
calculated as the ratio of extracted oil mass to initial plant sample (Velp Scientifica, Usmate, Italy), according to the AOCS standard

549
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

method Cd 12b-92. Each oil sample (10 g) was loaded into the chamber. independent measurements ± standard deviation (SD).
The working temperature was 90 °C, the initial O2 pressure was
0.6 MPa, and the induction time (IT) was recorded.
3. Results and discussion
2.3.6. Determination of the antioxidant activity
The antioxidant activity of obtained oils were evaluated using a 2,2- 3.1. Extraction yield and physicochemical properties of yellow horn seed
diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assay. 4 mL of oils
DPPH in ethanol (0.004%) was added to 1 mL of the oil in ethanol
(10.0 mg/mL). The absorbance was measured at 517 nm after 60 min In this work, the Soxhlet extraction of yellow horn seed oils pro-
incubation in the dark at room temperature. A DPPH solution without duced the highest yield (60.43%), which was comparable to yields in
the sample was used as a control. The radical-scavenging activities of previous studies (Li et al., 2013; Qu et al., 2013; Zhang et al., 2010).
the samples, expressed as percentage inhibition of DPPH, were de- This indicates that the Soxhlet extraction is the most efficient technique
termined using the following equation: for the extraction of plant oil (Teixeira, Ghazani, Corazza, Marangoni, &
Ribani, 2018). The subcritical n-butane and supercritical CO2 extrac-
Aa − Ab
DPPH scavenging activity (%) = × 100 tions achieved 97.29% and 93.45% of the yield of Soxhlet extraction
Ab (1)
(Table 1), but Soxhlet extraction required a longer processing period
where Ab and Aa are the absorbance values of the blank and tested (300 min) than subcritical n-butane extraction (90 min). Furthermore,
samples, respectively. the high cost of high pressure operation (∼30 MPa) in supercritical CO2
extraction hinders its application in the food industry. In contrast,
2.3.7. Determination of the thermal stability subcritical n-butane is an appropriate solvent for low pressure oil pre-
Thermogravimetric (TG) analysis was conducted using a thermo- paration (∼0.35 MPa) (Xu et al., 2016).
gravimetric analyzer (SDT Q600, TA Instruments, New Castle, DE, USA) Longer fatty acid chain lengths lower the specific gravity, while high
under a nitrogen atmosphere (30 mL/min) and aluminum crucibles. degrees of unsaturation increase the specific gravity. In this work, the
The mass of each sample was ∼15 mg. The analysis was conducted at specific gravity did not vary significantly (p > 0.05) among the oils
50–650 °C with a constant heating rate of 10 °C/min. Simultaneously, a (0.914–0.915 g/mL at 25 °C). The refractive indices were 1.471, 1.467,
thermal analyzer was used to identify the TG/difference TG (DTG) data and 1.475 (at 25 °C) for subcritical n-butane, supercritical CO2, and
and differential scanning calorimetry (DSC) information. Soxhlet-extracted oils, respectively. The acid value of the subcritical n-
butane extracted oil (0.39 mg KOH/g) was significantly lower than that
2.3.8. Rheological measurements of the supercritical CO2 extracted oil (0.52 mg KOH/g). Furthermore,
Rheological measurements were performed with an MCR 102 rhe- the peroxide value of the subcritical n-butane extracted oil (0.58 meq
ometer (Anton Paar, Graz, Austria) with a concentric cylinder of inner O2/kg) was significantly lower than that of both the other two kinds.
diameter 27 mm (CC24). Rheoplus software version 3.21 (Anton Paar, This was due to the mild temperature and pressure used in the sub-
Graz, Austria) was used to collect the rheological data. Flow curves critical n-butane extraction process, which reduced the decomposition
were plotted over the shear rate range 1–1000 s−1 at 25 °C. To in- and oxidation of the triacylglycerol, and subsequently decreased the
vestigate the effect of the temperature on the oils, a temperature sweep contents of peroxides, hydroperoxides, and free fatty acids (Sun et al.,
test was conducted at 5–100 °C; the obtained data was fitted to the 2018).
Arrhenius equation (Eq. (2)): The highest saponification value was found in the subcritical n-bu-
tane extracted oil (187.43 mg KOH/g), followed by the Soxhlet
μ = Ae (−Ea /RT ) (2) (184.67 mg KOH/g) and supercritical CO2 (180.59 mg KOH/g) ex-
where μ is the viscosity (Pa·s), A is the Arrhenius constant, Ea is the tracted oils, indicating the higher solubility of low-molecular-weight
activation energy (kJ/mol), R is the universal gas constant (8.314 J/ triacylglycerols in subcritical n-butane (Nehdi et al., 2012). Ad-
mol/K), and T is the temperature (K). ditionally, the unsaponifiable matter of the subcritical n-butane-ex-
tracted oil was higher than that of the supercritical CO2- and Soxhlet-
2.4. Statistical analyses extracted oils. This confirmed that subcritical n-butane extraction in-
creased unsaponifiable matter, such as sterols, carotenoids, and toco-
The collected data was analyzed using SPSS Statistics software pherols (Gu et al., 2016). In this study, the iodine value (137.60 g I/
version 19.0 (SPSS Inc., Chicago, IL, USA). To identify significant dif- 100 g) of the Soxhlet extracted oil was significantly higher than that of
ferences among the oils, analysis of the variance (ANOVA) followed by the other two kinds, indicating the presence of more unsaturated fatty
Duncan's multiple range procedure was applied at a 95% significance acid bonds in the Soxhlet extracted oil (Ruan et al., 2017).
level (a = 0.05). All results were expressed as arithmetic means of three

Table 1
Extraction yield and major physicochemical properties of yellow horn seed oils extracted through different methods.
Property Oils

Subcritical-extracted Supercritical-extracted Soxhlet-extracted

b c
Extraction yield (%) 58.79 ± 1.06 56.47 ± 1.31 60.43 ± 2.14a
Specific gravity (g/mL, 25 °C) 0.914 ± 0.02a 0.914 ± 0.04a 0.915 ± 0.03a
Refractive index (25 °C) 1.471 ± 0.03a 1.467 ± 0.01b 1.475 ± 0.04a
Acid value (mg KOH/g) 0.39 ± 0.02b 0.52 ± 0.03a 0.37 ± 0.02b
Peroxide value (meq O2/kg) 0.58 ± 0.02b 0.76 ± 0.03a 0.83 ± 0.03a
Saponification value (mg KOH/g) 187.43 ± 2.10a 180.59 ± 1.83b 184.67 ± 1.47a
Unsaponifiable matter (%) 0.52 ± 0.01a 0.44 ± 0.01b 0.46 ± 0.02b
Iodine value (g I/100 g) 131.42 ± 1.31b 129.77 ± 1.82b 137.60 ± 1.07a

Results are mean values ± standard deviation of duplicates. The same letters indicate no significant difference (p > 0.05).

550
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

3.2. Fatty acid composition of yellow horn seed oils

In this study, eleven fatty acids were identified (Table 2). Un-
saturated fatty acids (UFAs) predominated (852.79–858.28 mg/g oil) in
all samples. Monounsaturated fatty acids (MUFA) constituted
400.85–405.10 mg/g oil and polyunsaturated fatty acids (PUFAs) con-
stituted 447.69–456.13 mg/g oil. Similar results have been reported by
Qu et al. (2013) and Zhang et al. (2010). This similarity indicated
subcritical n-butane successfully extracted the same FAs as those ex-
tracted by the supercritical CO2 and Soxhlet methodologies.
Fatty acid composition is a major determinant of oil quality; good
quality oils contain high percentages of UFAs. Statistical analysis in-
dicated significant differences in the fatty acid levels for different
methods (Table 2). Linoleic acid (ω-6) (379.61–385.86 mg/g oil) and Fig. 1. FTIR spectra of yellow horn seed oils extracted through different
linolenic acid (ω-3) (68.08–70.27 mg/g oil) were the main PUFAs in the methods. (For interpretation of the references to colour in this figure legend,
oils, regardless of extraction technique. These compounds are essential the reader is referred to the Web version of this article.)
fatty acids that play important beneficial roles in human health (Gu
et al., 2016; Sodeifian, Ardestani, Sajadian, & Moghadamian, 2018). 3470 cm−1 corresponding to the stretching of the O–H bonds present in
Extraction using subcritical n-butane resulted in a higher level of fatty acids, carbohydrates, and proteins (Sodeifian et al., 2018;
linoleic acid and linolenic acid, but a lower quantity of oleic acid Sodeifian, Sajadian, & Ardestani, 2016).
compared to conventional extraction methods. This could be explained The FTIR spectra in the low-wavenumber region consisted of a
by the polarity of the solvents, which have more affinity with certain prominent peak at 1742 cm−1 attributable to the C=O carbonyl
fatty acids. The major MUFA in the oils was oleic acid stretching of lipid and fatty acid ester groups. This peak corresponded
(276.58–285.77 mg/g oil), followed by erucyl alcohol to the total lipids in the oil. The weak peak observed at 1654 cm−1 was
(86.37–87.40 mg/g oil), and nervonic acid (32.21–37.42 mg/g oil). attributed to UFA C=C stretching; its intensity was used to determine
Palmitoleic acid was identified in an amount less than 0.74 mg/g oil. the total unsaturation. The overlapped peaks at 1460–1372 cm−1 cor-
The MUFA oleic acid is important for its stability and health-promoting responded to the combination of the methyl and methylene group de-
properties, which include treatment and prevention of cardiovascular formation modes. The strong peaks at 1160 and 1099 cm−1 were at-
diseases (Teixeira et al., 2018). tributed to diacyl glycerol ethers in the oils. They represented the
Saturated fatty acids (SFAs) constituted 73.15–75.98 mg/g oil. The combined asymmetric stretching of C–C(=O)–O and O–C–C bonds and
main SFAs were palmitic acid (47.07–48.76 mg/g oil) and stearic acid C–O–C symmetric stretching of triglycerides and cholesterol esters, re-
(19.46–19.74 mg/g oil). Behenic acid was identified in smaller spectively. Additionally, all spectra exhibited a peak associated with the
amounts, while only traces of eicosanoic acid were present. Thus, based cis C=C out-of-plane bending at 720 cm−1 (Jing et al., 2014).
on the PUFA/SFA ratio, subcritical n-butane extraction can to extract
better quality oil compared to conventional techniques.
3.4. Tocopherol content, antioxidant activity, and oxidative stability of
3.3. FTIR analysis yellow horn seed oils

The peak in the FTIR spectra of the oils at 3010 cm−1 corresponded Tocopherols are natural antioxidants present in most plant oils that
to cis-olefinic –HC=CH- bond C–H stretching and was used to detect have a significant effect on oil quality. The antioxidant activity of to-
UFAs (Fig. 1). Its position depended strongly on the number of con- copherols is mainly attributed to their ability to donate hydrogen atoms
jugated C=C bonds in the fatty acids, and was used to determine the to free radicals, thereby inhibiting lipid oxidation (Grajzer et al., 2015).
degree of unsaturation. Fig. 1 shows the strong triplet peaks at All extracted oils contained α-, γ-, and δ-tocopherols (Table 3). The
2980–2800 cm−1 corresponding to the C–H stretching of the methyl major tocopherol was γ-tocopherol, at a concentration of 71.25, 71.83,
and methylene backbones of the oils (Timilsena, Vongsvivut, Adhikari, and 58.39 mg/100 g oil for the subcritical n-butane, supercritical CO2,
& Adhikari, 2017). In addition, all the FTIR spectra displayed a peak at and Soxhlet-extracted oils, respectively. The subcritical n-butane-

Table 2
Fatty acid (FA) composition (mg/g oil) of the yellow horn seed oils obtained through different extraction methods.
Fatty Acid Oils

Subcritical-extracted Supercritical-extracted Soxhlet-extracted

Myristic acid (C14:0) 0.28 ± 0.01a 0.32 ± 0.00a 0.31 ± 0.00a

Palmitic acid (C16:0) 47.07 ± 0.16c 47.74 ± 0.09b 48.76 ± 0.17a
Palmitoleic acid (C16:1) 0.74 ± 0.03a 0.65 ± 0.02b 0.74 ± 0.05a
Stearic acid (C18:0) 19.47 ± 0.12a 19.74 ± 0.07a 19.46 ± 0.21a
Oleic acid (C18:1) 276.58 ± 2.6b 276.97 ± 3.2b 285.77 ± 1.8a
Linoleic acid (C18:2) 385.86 ± 3.1a 382.36 ± 5.6b 379.61 ± 4.8c
Linolenic acid (C18:3) 70.27 ± 0.38a 70.15 ± 1.02a 68.08 ± 1.30b
Eicosanoic acid (C20:0) 0.84 ± 0.02b 0.93 ± 0.00a 0.93 ± 0.11a
Behenic acid (C22:0) 5.49 ± 0.25b 6.31 ± 0.09a 6.52 ± 0.24a
Erucyl alcohol (C22:1) 87.40 ± 1.61a 87.15 ± 0.85a 86.37 ± 0.73b
Nervonic acid (C24:1) 37.42 ± 0.82a 36.08 ± 1.45b 32.21 ± 1.27c
Total SFA 73.15 ± 3.51b 75.04 ± 2.64a 75.98 ± 2.82a
Total MUFA 402.15 ± 1.96b 400.85 ± 5.38c 405.10 ± 7.21a
Total PUFA 456.13 ± 5.36a 452.52 ± 2.87b 447.69 ± 7.83c

Results are mean values ± standard deviation of duplicates. The same letters indicate no significant difference (p > 0.05).

551
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

Table 3
Tocopherol content (α, γ, and δ), oxidative stability, and anti-oxidant activity (DPPH) of yellow horn seed oils obtained through different extraction methods.
Oils Tocopherol concentration (mg/100 g) Oxidative stability (IT, h) Anti-oxidant activity (%)

α γ δ Total

a a a
Subcritical-extracted 17.62 ± 0.71 71.25 ± 2.06 17.40 ± 0.42 106.27 ± 2.46a 11.36 ± 0.42a 84.64 ± 1.07a
Supercritical-extracted 13.81 ± 0.40b 71.83 ± 1.97a 12.48 ± 0.31b 98.12 ± 2.58b 8.42 ± 0.63b 75.56 ± 0.81b
Soxhlet-extracted 13.46 ± 0.53b 58.39 ± 1.24b 11.43 ± 0.37c 83.28 ± 1.75c 7.63 ± 0.47c 72.40 ± 0.94c

Results are mean values ± standard deviation of duplicates. The same letters indicate no significant difference (p > 0.05).

Fig. 2. TG and DSC curves of yellow horn seed oils obtained through different
extraction methods. (For interpretation of the references to colour in this figure
legend, the reader is referred to the Web version of this article.)
Fig. 4. Effect of the temperature on the viscosity of yellow horn seed oils ex-
tracted through different methods. (For interpretation of the references to
colour in this figure legend, the reader is referred to the Web version of this
article.)

Fig. 3. Flow behaviors of yellow horn seed oils extracted through different
methods. Inset (A) is expanded view of the spectral features. (For interpretation
of the references to colour in this figure legend, the reader is referred to the
Web version of this article.)
Fig. 5. Arrhenius plots of ln viscosity versus 1/T.

extracted oil had the highest total tocopherol content (106.27 mg/100 g
Table 4
oil), indicating the selectivity of the solvents for tocopherol. Compar-
Arrhenius model parameters of yellow horn seed oils.
ison of these results with previous reports shows that subcritical short-
chain alkanes (n-butane) have a better solubility for tocopherol than Oils Ea (KJ/mol) A (Pa s) R2
supercritical CO2 and n-hexane (Nimet et al., 2011; Sun et al., 2018).
Subcritical extracted 21.06 1.70E-5 0.9925
These results confirmed that yellow horn seed oil can serve as a dietary Supercritical extracted 21.08 1.71E-5 0.9926
source of tocopherols. Soxhlet extracted 20.69 1.60E-5 0.9931
Furthermore, the DPPH scavenging activities of the oils were de-
termined to be 72.40–84.64% (Table 3). The subcritical n-butane-ex- A is the Arrhenius constant, Ea is the activation energy, and R2 is the coefficient
tracted oil showed significantly higher antioxidant activity than the of determination.
other oils. This was consistent with the tocopherol results, as tocopherol
content is positively correlated with antioxidant activity (Delfan- potential antioxidant properties.
Hosseini, Nayebzadeh, Mirmoghtadaie, Kavosi, & Hosseini, 2017). Finally, the oxidative stability of the oils was evaluated by mea-
Moreover, fatty acids and some trace components affect the antioxidant suring their ITs with the Oxitest method. The IT values for subcritical-,
activity of oils (Zheng, Shi, Zhao, Jin, & Wang, 2017). However, it is supercritical-, and Soxhlet-extracted oils were 11.36, 8.42, and 7.63 h,
difficult to attribute the antioxidant effect to specific active constituents respectively (Table 3). The oxidative stability of the subcritical n-bu-
because a mixture of different chemical constituents is present. tane-extracted oil was significantly higher than those of the other two
Nevertheless, this study suggests that yellow horn seed oil shows samples (p < 0.05). The lower oxidative stability of the Soxhlet-

552
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

extracted oil was probably due to the elevated temperature and pro- health, such as UFAs and tocopherol. The subcritical n-butane extracted
longed duration of the process, which can accelerate oil oxidation. The oil had higher tocopherol content, oxidation stability, and in vitro an-
higher oxidative stability of the subcritical n-butane-extracted oil was tioxidant activity than the oils obtained using supercritical CO2 and
ascribed to its higher tocopherol content and antioxidant activity. Soxhlet extraction. These findings demonstrated that the use of n-bu-
tane as a solvent for oil extraction did not negatively affect the oil
3.5. Thermal property of yellow horn seed oils composition; thus, this efficient and eco-friendly method could be op-
timized in order to obtain yellow horn seed oil with high quality and
The TG curves indicated progressive degradation of the extracted industrial applications.
oils with three main stages (Fig. 2). In the first stage, the onset of mass
loss (about 1%) was observed at 180 °C, followed by a second stage up Acknowledgments
to 390 °C, with 48% of mass loss, and by a third and more severe mass
loss stage (> 98%) up to 480 °C. The three stages likely represent PUFA, The work was supported by the Scientific Research Fund for
MUFA, and SFA degradation, followed by volatilization of the poly- Doctoral of Anyang Institute of Technology (BSJ2018011), the Science
merization products and pyrolysis (Gu et al., 2016). and Technology Tackling Key Project of Henan Province
The DSC curves confirmed the oil degradation stages; the end of the (192102110102), and the Program for Science and Technology
first stage was characterized by exothermic peaks at 180 °C and was Innovation Talents in Universities of Henan Province (18HASTIT035).
associated with the formation of peroxides, which decomposed to other We thank Elsevier Webshop for its linguistic assistance during the
products at 390 °C. The thermal stability of the yellow horn oil was preparation of this manuscript. Declarations of interest: none.
similar to that of other mainly-UFA (> 80%) edible oils (Zanqui et al.,
2016). In a general way, the study indicates that yellow seed oil can be References
used as a frying oil and processed at temperatures lower than 180 °C.
Blayo, A., Gandini, A., & Le, N. J. (2001). Chemical and rheological characterizations of
3.6. Rheological behaviors of yellow horn seed oils some vegetable oils derivatives commonly used in printing inks. Industrial Crops and
Products, 14(2), 155–167. https://doi.org/10.1016/S0926-6690(01)00079-6.
Czaikoski, K., Mesomo, M. C., Scheer, A. D. P., Santa, O. R. D., Queiroga, C. L., & Corazza,
All extracted oils had similar flow patterns at 25 °C, and their shear M. L. (2015). Kinetics, composition and biological activity of Eupatorium intermedium
stress was directly proportional to the shear rate (Fig. 3), demonstrating flower extracts obtained from scCO2 and compressed propane. The Journal of
Supercritical Fluids, 97, 145–153. https://doi.org/10.1016/j.supflu.2014.10.008.
Newtonian flow characteristics due to their long chain molecules Delfan-Hosseini, S., Nayebzadeh, K., Mirmoghtadaie, L., Kavosi, M., & Hosseini, S. M.
(Blayo, Gandini, & Le, 2001; Kim, Kim, Lee, Yoo, & Lee, 2010). Fur- (2017). Effect of extraction process on composition, oxidative stability and rheolo-
thermore, the oil viscosities given by the slopes of the flow curves were gical properties of purslane seed oil. Food Chemistry, 222, 61–66. https://doi.org/10.
1016/j.foodchem.2016.11.150.
58.98, 58.36, and 56.16 mPa s for the supercritical CO2-, subcritical n- Grajzer, M., Prescha, A., Korzonek, K., Wojakowska, A., Dziadas, M., Kulma, A., et al.
butane-, and Soxhlet-extracted oils, respectively (Fig. 3A). The viscosity (2015). Characteristics of rose hip (Rosa canina L.) cold-pressed oil and its oxidative
decreased as the content of fatty acids with two or more double bonds stability studied by the differential scanning calorimetry method. Food Chemistry,
188, 459–466. https://doi.org/10.1016/j.foodchem.2015.05.034.
(linoleic and linolenic acids) in the chain increased (Delfan-Hosseini
Gu, L.-B., Pang, H.-L., Lu, K.-K., Liu, H.-M., Wang, X.-D., & Qin, G.-Y. (2016). Process
et al., 2017). Since the total MUFA content in the Soxhlet-extracted oil optimization and characterization of fragrant oil from red pepper (Capsicum annuum
was higher than that of the other two oils, its lower viscosity was jus- L.) seed extracted by subcritical butane extraction. Journal of the Science of Food and
tified. In addition, decreased free fatty acid content can decrease oil Agriculture, 97(6), 1894–1903. https://doi.org/10.1002/jsfa.7992.
Herrero, M., Cifuentes, A., & Ibañez, E. (2006). Sub- and supercritical fluid extraction of
viscosity, perhaps because less shear force is required for flow (Liu, functional ingredients from different natural sources: Plants, food-by-products, algae
Yang, & Huang, 2012). The characteristic acid value of the free fatty and microalgae: A review. Food Chemistry, 98(1), 136–148. https://doi.org/10.1016/
acids was significantly lower for the Soxhlet-extracted oil than for the j.foodchem.2005.05.058.
Jing, L., Pan, C., Jing, H., Deng, L., Wang, L. Y., Lei, J., et al. (2014). Extraction of oil from
subcritical n-butane- and supercritical CO2-extracted oils (Table 1). Jatropha curcas seeds by subcritical fluid extraction. Industrial Crops and Products,
The flow behaviors of the extracted oils were studied at tempera- 62(62), 235–241. https://doi.org/10.1016/j.indcrop.2014.08.039.
tures 5–85 °C. All oils showed a similar pattern; their viscosity de- Kim, J., Kim, D. N., Lee, S. H., Yoo, S.-H., & Lee, S. (2010). Correlation of fatty acid
composition of vegetable oils with rheological behaviour and oil uptake. Food
creased with increasing temperature (Fig. 4). The temperature effect on Chemistry, 118(2), 398–402. https://doi.org/10.1016/j.foodchem.2009.05.011.
oil viscosity has been attributed to greater molecular movement due to Liu, C., Yang, M., & Huang, F. (2012). Influence of extraction processing on rheological
reduced intermolecular interactions and the destruction of plant oil properties of rapeseed oils. Journal of the American Oil Chemists Society, 89(1), 73–78.
https://doi.org/10.1007/s11746-011-1892-y.
aggregates. The Soxhlet-extracted oil had the lowest viscosity over the Li, J., Zu, Y.-G., Luo, M., Gu, C.-B., Zhao, C.-J., Efferth, T., et al. (2013). Aqueous enzy-
entire temperature range. This result was consistent with the steady matic process assisted by microwave extraction of oil from yellow horn (Xanthoceras
shear measurements. sorbifolia Bunge.) seed kernels and its quality evaluation. Food Chemistry, 138(4),
2152–2158. https://doi.org/10.1016/j.foodchem.2012.12.011.
The Arrhenius curve of ln(viscosity) versus inverse temperature
Ndiaye, P. M., Lanza, M., Tavares, F. W., Dariva, C., Oliveira, D., & Oliveira, J. V. (2006).
(Fig. 5) was employed to determine the thermodynamic parameters of Phase behavior of olive and soybean oils in compressed propane and n-butane.
the oils (Table 4). The R2 values of all oils were higher than 0.99, in- Brazilian Journal of Chemical Engineering, 23(3), 405–415. https://doi.org/10.1590/
dicating their good Arrhenius behavior. The activation energy (Ea) of S0104-66322006000300014.
Nehdi, I. A., Sbihi, H., Tan, C. P., Zarrouk, H., Khalil, M. I., & Al-Resayes, S. I. (2012).
the oil samples ranged from 20.69 to 21.08 kJ/mol−1. The supercritical Characteristics, composition and thermal stability of Acacia senegal (L.) Willd. seed
CO2-extracted oil had the highest Ea value, suggesting that it is more oil. Industrial Crops and Products, 36(1), 54–58. https://doi.org/10.1016/j.indcrop.
sensitive to temperature. Oils with more double bonds have lower 2011.08.005.
Nimet, G., da Silva, E. A., Palú, F., Dariva, C., Freitas, L. d. S., Neto, A. M., et al. (2011).
viscosities and lower activation energies due to their loosely-packed Extraction of sunflower (Heliantus annuus L.) oil with supercritical CO2 and sub-
structure (Delfan-Hosseini et al., 2017). The present study revealed that critical propane: Experimental and modeling. Chemical Engineering Journal, 168(1),
the Soxhlet-extracted oil with the most MUFAs had the lowest viscosity 262–268. https://doi.org/10.1016/j.cej.2010.12.088.
Qu, X.-J., Fu, Y.-J., Luo, M., Zhao, C.-J., Zu, Y.-G., Li, C.-Y., et al. (2013). Acidic pH based
and activation energy, and confirmed that the extraction method af- microwave-assisted aqueous extraction of seed oil from yellow horn (Xanthoceras
fected the rheological behavior of the plant oils. sorbifolia Bunge.). Industrial Crops and Products, 43(5), 420–426. https://doi.org/10.
1016/j.indcrop.2012.07.055.
Rai, A., Mohanty, B., & Bhargava, R. (2015). Modeling and response surface analysis of
4. Conclusion supercritical extraction of watermelon seed oil using carbon dioxide. Separation and
Purification Technology, 141, 354–365. https://doi.org/10.1016/j.seppur.2014.12.
The results of this study showed that the yellow horn seed is a good 016.
Ruan, C.-J., Yan, R., Wang, B.-X., Mopper, S., Guan, W.-K., & Zhang, J. (2017). The im-
source of high-quality edible oil with superior nutritional value and
portance of yellow horn (Xanthoceras sorbifolia) for restoration of arid habitats and
high contents of valuable compounds with positive effects on human

553
L.-B. Gu, et al. LWT - Food Science and Technology 111 (2019) 548–554

production of bioactive seed oils. Ecological Engineering, 99, 504–512. https://doi. from sapucaia (Lecythis pisonis) nuts. The Journal of Supercritical Fluids, 133(1),
org/10.1016/j.ecoleng.2016.11.073. 122–132. https://doi.org/10.1016/j.supflu.2017.10.003.
Sodeifian, G., Ardestani, N. S., Sajadian, S. A., & Moghadamian, K. (2018). Properties of Timilsena, Y. P., Vongsvivut, J., Adhikari, R., & Adhikari, B. (2017). Physicochemical and
Portulaca oleracea seed oil via supercritical fluid extraction: Experimental and opti- thermal characteristics of Australian chia seed oil. Food Chemistry, 228, 394–402.
mization. The Journal of Supercritical Fluids, 135, 34–44. https://doi.org/10.1016/j. https://doi.org/10.1016/j.foodchem.2017.02.021.
supflu.2017.12.026. Xu, B., Han, J., Zhou, S., Wu, Q., & Ding, F. (2016). Quality characteristics of wheat germ
Sodeifian, G., Sajadian, S. A., & Ardestani, N. S. (2016). Supercritical fluid extraction of oil obtained by innovative subcritical butane experimental equipment. Journal of
omega-3 from Dracocephalum kotschyi seed oil: Process optimization and oil proper- Food Process Engineering, 39(1), 79–87. https://doi.org/10.1111/jfpe.12201.
ties. The Journal of Supercritical Fluids, 119, 139–149. https://doi.org/10.1016/j. Zanqui, A. B., da Silva, C. M., de Morais, D. R., Santos, J. M., Ribeiro, S. A. O., Eberlin, M.
supflu.2016.08.019. N., et al. (2016). Sacha inchi (Plukenetia volubilis L.) oil composition varies with
Sun, D., Cao, C., Li, B., Chen, H., Li, J., Cao, P., et al. (2018). Antarctic krill lipid extracted changes in temperature and pressure in subcritical extraction with n-propane.
by subcritical n-butane and comparison with supercritical CO2 and conventional Industrial Crops and Products, 87, 64–70. https://doi.org/10.1016/j.indcrop.2016.04.
solvent extraction. LWT-Food Science and Technology, 94, 1–7. https://doi.org/10. 029.
1016/j.lwt.2018.04.024. Zhang, S., Zu, Y.-G., Fu, Y.-J., Luo, M., Liu, W., Li, J., et al. (2010). Supercritical carbon
Szentmihályi, K., Vinkler, P., Lakatos, B., Illés, V., & Then, M. (2002). Rose hip (Rosa dioxide extraction of seed oil from yellow horn (Xanthoceras sorbifolia Bunge.) and its
canina L.) oil obtained from waste hip seeds by different extraction methods. anti-oxidant activity. Bioresource Technology, 101(7), 2537–2544. https://doi.org/10.
Bioresource Technology, 82(2), 195–201. https://doi.org/10.1016/S0960-8524(01) 1016/j.biortech.2009.11.082.
00161-4. Zheng, L., Shi, L.-K., Zhao, C.-W., Jin, Q.-Z., & Wang, X.-G. (2017). Fatty acid, phyto-
Tasioula-Margari, M., & Okogeri, O. (2001). Simultaneous determination of phenolic chemical, oxidative stability and in vitro antioxidant property of sea buckthorn
compounds and tocopherols in virgin olive oil using HPLC and UV detection. Food (Hippophaë rhamnoides L.) oils extracted by supercritical and subcritical technologies.
Chemistry, 74(3), 377–383. https://doi.org/10.1016/S0308-8146(01)00176-5. LWT-Food Science and Technology, 86, 507–513. https://doi.org/10.1016/j.lwt.2017.
Teixeira, G. L., Ghazani, S. M., Corazza, M. L., Marangoni, A. G., & Ribani, R. H. (2018). 08.042.
Assessment of subcritical propane, supercritical CO2 and Soxhlet extraction of oil

554