Extraction of crambe seed oil using subcritical propane: Kinetics, characterization and modeling

J. of Supercritical Fluids 104 (2015) 54–61

Contents lists available at ScienceDirect

The Journal of Supercritical Fluids journal homepage: www.elsevier.com/locate/supflu

Extraction of crambe seed oil using subcritical propane: Kinetics, characterization and modeling Kátia Andressa Santos a , Reinaldo Aparecido Bariccatti a , Lúcio Cardozo-Filho b , Ricardo Schneider c , Fernando Palú a , Camila da Silva d,∗ , Edson Antônio da Silva a a

Centro de Engenharias e Ciências Exatas, Universidade Estadual do Oeste do Paraná, Toledo, PR, Brazil Departamento de Engenharia Química, Universidade Estadual de Maringá (UEM), Maringá, PR, Brazil c Departamento de Processos Químicos, Universidade Tecnológica Federal do Paraná (UTFPR), Toledo, PR, Brazil d Departamento de Tecnologia, Universidade Estadual de Maringá (UEM), Umuarama, PR, Brazil b

a r t i c l e

i n f o

Article history: Received 26 January 2015 Received in revised form 24 May 2015 Accepted 24 May 2015 Available online 5 June 2015 Keywords: Crambe abyssinica oil Extraction Subcritical propane Oil characterization Mathematical modeling

a b s t r a c t In this study, the extraction of crambe seed oil Crambe abyssinica H. FMS Brilhante using subcritical propane as a solvent was investigated. The extraction yield and oil characteristics were compared with the oil extracted using n-hexane and dichloromethane. A factorial experimental design was used in order to evaluate the effects of temperature (313–353 K) and pressure (8–16 MPa) on the extracted yield using subcritical propane. It was observed that the temperature has the most significant effect on the extraction yield and the highest yield (32.8 wt%) was obtained at 353 K and 16 MPa. The experimental conditions showed no significant influence on fatty acids composition. Low levels of free fatty acids were found in the extracts (<2%), and the amounts of phytosterols and tocopherols were affected by the subcritical extraction conditions. A longer oxidative induction time was observed for the obtained oil using the subcritical method. The Sovová mathematical model represented satisfactorily the experimental data. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Crambe (Crambe abyssinica Hochst) is an oilseed crop of the Brassicaceae family related to canola and mustard [1,2]. Native to the Mediterranean region, it adapts well to cold and dry climates. The tree can reach 1–2 m in height. The spherical seeds range from 0.8 to 2.6 mm in diameter and have a high oil content of around 38% [3,4]. The seed oil is the main product, and it contains high levels of erucic acid (50–60%), a fatty monounsaturated acid with many applications in the pharmaceutical, cosmetic, lubricant and plastic industries, among others [5]. This oilseed was introduced in Brazil in 1995 by Fundac¸ão MS in Maracajú, Mato Grosso do Sul State, where it was identified as a promising crop for biodiesel production. In 2007, the FMS Brilhante variety was registered, with an initial production of between 1000 and 1500 kg ha−1 [5]. By 2010, production yields of up to 2300 kg ha−1 had been registered [6]. For oil extraction, the methods commonly used are press extraction and solvent extraction. Both methods involve a long extraction period followed by further steps required to separate the oil and

∗ Corresponding author. E-mail address: [email protected] (C.d. Silva). http://dx.doi.org/10.1016/j.supflu.2015.05.026 0896-8446/© 2015 Elsevier B.V. All rights reserved.

solvent. In this regard, the use of pressurized, supercritical or subcritical fluids is an option for replacing the conventional methods. This technology is considered to be clean because, at the end of the extraction, the solvent is completely removed by system depressurization and can be recovered. Although carbon dioxide is the most commonly used fluid in this kind of extraction, subcritical propane allows high extraction rates when used in processes with vegetable oil [7–11] due to the greater solubility of the triglycerides in this solvent [12]. In addition, it can be used at lower pressures, which is an important advantage in the oil extraction industry. Few studies on subcritical and supercritical extraction technology for obtaining oil from crambe seeds have been carried out. Onorevoli et al. [13] studied the free fatty acids composition and the yield of crambe oil obtained after 30 min of total extraction using propane at 313 K and 15 MPa. However, no information could be found in the literature regarding the extraction kinetics using subcritical propane and the effects of the temperature and pressure conditions on the chemical composition and extraction yield. Thus, studies need to be carried out on the oxidative stability and antioxidant content of extracts obtained by this method. In this context, the aim of this study was to analyze subcritical propane as an extraction fluid for obtaining crambe seed oil. The effects of the temperature and pressure on the extraction yield were investigated, and the Sovová model was applied in the mathe-

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

matical modeling of the extraction kinetics. In addition, oil samples obtained by the subcritical method and by the Soxhlet extraction using n-hexane and dichloromethane as solvents were analyzed and compared in terms of their chemical compositions (fatty acids profile, free glycerol compounds and tocopherols) and oxidative stability.

55

and its mass was determined at time intervals of 5 (up to 30 min), 10 (30–60 min) and 20 (60–80 min of extraction). The yields were calculated as the ratio of the extracted oil mass to the initial crambe seed mass. StatisticaTM, version 7, software (Statsoft) was used to analyze the experimental data at the 95% confidence level. 2.4. Mathematical modeling

2. Materials and methods 2.1. Sample preparation Crambe seeds (Crambe abyssinica H.) of FMS Brilhante variety (Fundac¸ão MS, Mato Grosso do Sul State, Brazil) were used in this study. The peeled seeds were crushed and classified using a system of Tyler series (12–32 mesh) sieves. The particles retained on the 14 mesh sieve were used in the experiments. Seed moisture (4.06 ± 0.02 wt%) was obtained by the gravimetric method drying the sample at 378 K until mass was stabilized. The density (1.120 g cm−3 ) was determined by pycnometry using helium gas (Micromeritics, AccuPyc model 1330). 2.2. Reagents and standards For the conventional extraction method, dichloromethane 99.5% (Vetec) and n-hexane 99% (F. Maia) were used. Propane P.A. 95% (Linde Gás) was used in the subcritical extractions. Potassium hydroxide P.A. (Biotec), methanol PA (Vetec) and heptane 99.6% (Merck) were used in the oil derivatization. To determine free glycerol compounds, content N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA), trimethylchlorosilane (TMCS), the internal standards of 5␣-cholestane and methyl heptadecanoate (obtained from Sigma–Aldrich) were used in the oil derivatization. For the determination of the tocopherol levels in the extracts, ␣, ␥ and ␦tocopherol standards (Sigma–Aldrich), methanol (J.T. Baker, grau HPLC) and ultrapure water (Milli-Q) were used. Nitrogen 99.9% (White Martins) and oxygen 99.9% (Linde Gás) were used in DSC analysis. 2.3. Oil extraction 2.3.1. Soxhlet extractions The extractions were performed with a Soxhlet extractor (Laborgas) to determine the oil content of the seeds and to compare the characteristics of the oil obtained by this method with those of the oil obtained with subcritical propane. Approximately 10 g of seeds were used in the exhaustive extraction (480 min) with dichloromethane and n-hexane at their respective boiling points carried out according to the method described in AOAC 920.39 [14]. The extractions were carried out in triplicate, and the results are reported as mean value ± standard deviation. 2.3.2. Subcritical propane extraction The experiments were performed in a laboratory scale unit (Fig. 1). The equipment consists of a solvent, a solvent reservoir, a syringe pump (Isco, 500D model) and two thermostatic baths – one (Julabo, F25-ME model) used to cool the fluid in the syringe pump and the other (Quimis, Q214M2 model) to maintain the extractor at the temperature set point, and a stainless steel extractor with 58 cm3 of capacity (1.95 cm of diameter and 19.4 cm of height). A 22 full factorial experimental design with center points was used to analyze the influence of the independent variables, temperature and pressure on the extraction yield. In each extraction, the vessel was loaded with approximately 30 g of seeds. The experiments were performed in the temperature and pressure ranges of 313–353 K and 8–16 MPa, respectively, with a mass flow rate of 1.6 × 10−3 kg min−1 . The oil was collected in an amber glass vessel,

The Sovová model [15] was used to describe the oil extraction kinetic curves with subcritical propane. The analytical solution of Sovová’s model is described by Eqs. (1)–(3). Fort < tCER : ˙ F YS t [1 − exp (−Z)] m(t) = m

(1)

For tCER ≤ t ≤ tFER :

 ˙ F YS m(t) = m



ZYS ln WX0

t − tCER exp



1 1−r



 exp

˙F Wm ms



 (tCER − t) − r

 −Z

(2) For t > tFER :



m(t) = ms X0 −

Ys ln W





1+

 exp

WX0 YS



 −1

 exp

˙F Wm ms



 (tCER − t) r

(3)

where: Z=

kF ams F ˙ F S m

W=

(4)

ms kS a ˙ F (1 − ) m

tCER =

(5)

(1 − r) ms X0 ˙F YS Z m

tFER = tCER +

(6)



ms ln r + (1 − r)exp ˙F Wm

 WX  0

YS

(7)

˙ F is the solvent mass flow rate (g min−1 ), YS is the where m extract solubility in the solvent (goil gpropane −1 ), t is the extraction time (min), X0 is the initial oil concentration in the solid matrix (goil gsolid −1 ), mS is the solid mass on an oil-free basis (g), r is the easily accessible oil fraction (XP /X0 ), tCER is the end of the first extraction period (min), tFER is the end of the second extraction period (min), kF a is the solvent phase mass transfer coefficient (min−1 ),  is the extraction bed porosity, kS a is the solid phase mass transfer coefficient (min−1 ), ␳F is the fluid density (g cm−3 ) and ␳S is the solid density (g cm−3 ). Z and W are dimensionless model parameters. As described by Ribas et al. [16], the parameter r is a constant associated with the crushing and sieving process and is the same for the entire sample. It was adjusted by the “golden-search” method using the following objective function:



n exp N

F=

mCalc oili,j

Exp − moil i,j

2 (8)

i=1 j=1

The parameters Z and W were calculated using the downhill simplex method [17] minimizing the following objective function: F=

N  

Exp

mCalc − moil oil j

j

2 (9)

j=1

where mCalc is the calculated mass of the oil extracted using the oil,j Exp

Sovová model, moil,j is the mass of oil obtained experimentally, n exp is the number of the extraction experiments and N is the number of experimental data points on the kinetic curve.

56

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

Fig. 1. Experimental module of extraction with subcritical propane: (A) gas cylinder; (B) syringe pump; (C and D) thermostatic baths; (E) temperature controller in the micrometric valve; (F) extractor; (V1), (V2) and (V3) needle valves; (V4) micrometric valve of flow rate controller.

2.5. Fatty acids composition The fatty acids composition was determined as described by Garcia et al. [18] using a gas chromatograph (Agilent, model 7890) coupled with a mass spectrum (MS 8990), equipped with a capillary column (ZBWAX, 30 m × 0.25 mm × 0.25 ␮m). The fatty acids were derivatized according to the AOCS Ce 2-66 methodology [19], and the identification was performed by comparing the spectral data with the Wiley library spectra.

membrane, Millipore) and quantified in a liquid chromatograph (LC-20AT, coupled to a SPD-20A UV–VIS, Shimadzu) equipped with a C-18 column (Shim-pack CLC-ODS M) with 5 ␮m particle diameter (4.6 mm × 25 cm) at 298 K with a loop of 20 ␮L. The quantification of the identified compounds was performed using the external standard method. Calibration curves for ␣, ␥ and ␦tocopherols were prepared at concentrations of 0.5–5 mg L−1 with their respective standards, which showed regressions of R2 > 0.99. The analyses were carried out in duplicate, and the results are presented as mean values ± standard deviation.

2.6. Determination of free glycerol compounds The free glycerol compounds were derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) and trimethylchlorosilane (TMCS) according to the methodology described by Freitas et al. [11], with few modifications. Around 20 mg of oil were derivatized with 20 ␮L of BSTFA/TMCS and left to stand at 333 K for 30 min. A volume of 80 ␮L of the internal standard was added for the quantification of phytosterols (5␣cholestane, 3470 mg L−1 ), and 50 ␮L of the internal standard was used for the free fatty acids quantification (methyl heptadecanoate, 5570 mg L−1 ). The solution was mixed, and the total volume made up to 1 mL with heptane. The analysis was performed using a gas chromatograph coupled to a mass spectrometer (ThermoFinnigan) equipped with an Agilent HP-5MS capillary column (30m × 0.25 mm × 0.25 ␮m). It was carried out at an initial temperature of 373 K for 6 min, heating to 503 K at 5 K min−1 and then to 553 K at 15 K min−1 ; this temperature was being held for 15 min. The carrier gas flow was set at 1 mL min−1 . The analysis was performed with 553 K as both the injector and detector temperature and an injection volume of 0.4 ␮L in split mode 1:10. The Xcalibur® (Thermo Electron) software was used to identify the compounds. The analyses were carried out in duplicate, and the results are presented as mean values ± standard deviation. 2.7. Tocopherol composition The ␣, ␥ and ␦-tocopherols present in crambe oil were determined as reported by Freitas et al. [20]. Around 20 mg of the oil were dissolved in 1 mL of 2-propanol which was filtered (0.45 ␮m nylon

2.8. Oxidative stability The oxidative stability of the extracted crambe oil samples using subcritical propane at 353 K and 16 MPa with n-hexane and dichloromethane was determined using differential scanning calorimetry (DSC Shimadzu-60). According to the method described by Tan et al. [21], 2.0 ± 0.5 mg of the oil samples were placed in an aluminum pan and then in the sample compartment of the instrument. Each sample was submitted to four different temperatures (383, 393, 403 and 413 K) with an oxygen flow rate of 50 mL min−1 .

3. Results and discussion 3.1. Extraction yield Table 1 shows the experimental conditions and the yields for the extraction of the crambe oil using subcritical propane, n-hexane and dichloromethane. The oil solubility (YS ) values were calculated from the linear part of the extraction curves. The total oil content in the crambe seeds was determined by the exhaustive Soxhlet extraction, and the obtained values were 51.0 wt% and 47.5 wt% using dichloromethane and n-hexane, respectively. This difference occurs due to the low polarity of dichloromethane, which means that this solvent extracts both neutral and polar lipids [22]. With subcritical propane, after 80 min of total extraction, 51.6–64.3% of the total oil is removed. The highest yields of 32.8 and 29.7 wt% were obtained in experiments 3 and

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

57

Table 1 Experimental conditions and results for crambe oil extraction yields using subcritical propane, n-hexane and dichloromethane. Run

T (K)

P (MPa)

1 2 3 4 5 6 7

313 313 353 353 333 n-Hexanedichloromethane

8 16 8 16 12 480 480

a b c

Time (min)

␳F a (g cm−3 )

YS b (goil gpropane −1 )

Yieldc (wt%)

80 80 80 80 80

0.485 0.509 0.424 0.459 0.474

0.170 0.149 0.128 0.139 0.169 47.5 ± 0.45 51.0 ± 0.33

26.3 26.3 29.7 32.8 29.6 ± 0.06

Propane density. Oil solubility in the solvent. Average value ± standard deviation of three replicates.

4, respectively, and performed at the highest temperature used (353 K). The solubility values presented in Table 1, 0.128–0.170 goil gpropane −1 , can be considered low when compared with other studies that used propane at subcritical conditions for the oil extraction, and that explains the yields obtained at the end of 80 min of extraction. Silva et al. [23] observed solubility values between 0.40 and 0.48 goil gpropane −1 and in 80 min removed from 84.1 to 86.8% of the total oil content in Perilla seeds. Pessoa et al. [24] found a solubility of 0.428 goil gpropane −1 and removed from 74.6 to 82.8% of the total oil content from the pequi pulp in 80 min. Nimet et al. [8] obtained between 87.8 and 100% of the oil in sunflower seeds with 40 min of extraction and the solubility values varying from 0.88 to 1.50 goil gpropane −1 . For the temperature and pressure conditions investigated in the experimental design, the temperature was the only variable that had a significant effect (p < 0.05). With an increase in temperature, the extraction yield increased. Although the pressure did not have a statistically significant influence on the extraction yield at the highest temperature used (experiments 3 and 4), the results did suggest a positive effect for this variable. A similar effect of pressure has been reported for the extraction of oil from sunflower seed [8], canola seed [9], sesame seed [10] and pequi pulp [24]. As reported by Jesus et al. [25], an increase in pressure under conditions where the temperature is close to the critical temperature of propane has a more pronounced effect on the solvent density, which results in an improvement in the solvating power and a consequent improvement in the extraction yield. 3.2. Mathematical modeling The crambe oil extraction kinetics using subcritical propane was represented by the Sovová’s model [15]. The parameters used for the mathematical modeling were: initial oil concentration of 0.488 goil gsolid −1 , solid density of 1.120 g cm−3 , bed density of 0.343 g cm−3 , bed porosity of 0.684, solid mass on an oil-free basis of 19.9 g and solvent mass flow rate of 1.6 × 10−3 kg min−1 . The solvent density and oil solubility in the solvent values are given in Table 1. Table 2 shows the adjustable parameters of the Sovová’s model. The external mass transfer coefficients, kFa , were obtained using the Z values and Eq. (4) as parameters, while for the internal mass transfer coefficient, kSa , W and Eq. (5) were used as parameters. According to the data in Table 2, the extraction can be divided into three stages. The first period (CER) is short due to the high extraction rate (kFa between 0.111 and 0.789 min−1 ) of the oil fraction directly exposed to the solvent (0.670). In the last period (DCR), there is no easily accessible oil remaining and, therefore, the oil extraction is dependent on the efficiency of the solvent in terms of accessing the internal part of the seeds where the cells are still intact. Given the difficultly associated with extracting this part of the oil, this stage is considered a limiting step in the extraction, and,

Fig. 2. Experimental extraction kinetics curves for crambe oil extraction using propane fitted using Sovová model: (353 K, 16 MPa);  (353 K, 8 MPa);  (313 K, 16 MPa);  (313 K, 8 MPa); 䊉 (333 K, 12 MPa).

thus, the yield is a function of the mass transfer rate, which is low (kSa between 6.298 × 10−3 and 1.541 × 10−2 min−1 ). The highest value for kSa was obtained applying the highest temperature and pressure (353 K and 16 MPa) demonstrating the highest yield obtained in this experiment. The relation between the yield and the effects of temperature and pressure are discussed in Section 3.1, the yield being higher when higher temperatures are used. The experimental extraction kinetics curves and the fitting Sovová model are shown in Fig. 2. The model shows a good fit for the experimental data (R2 > 0.99) under all conditions reported with a maximum mean absolute error of 5.7%. 3.3. Oil characterization Figs. 3 and 4 show the unsaturated and saturated fatty acids distribution for the crambe oils obtained under different temperature and pressure conditions, extracted with subcritical propane and with conventional extraction using n-hexane (HEX) and dichloromethane (DCM). Significant differences (at a significance level of 5%) were not observed indicating that neither the solvent used nor the experimental conditions had any influence on the fatty acids composition, as also observed by Onorevoli et al. [13]. The unsaturated fatty acids are predominant in the crambe oil composition [26]. In this study, it was observed that these accounted for 94% of the total fatty acids, with erucic acid (C22:1n9) being the main contributor representing 59.4% followed by oleic acid (C18:1n-9) contributing 20.17%, and linoleic (C18:2n-6) and linolenic (C18:3n-3) with values of 7.52 and 5.71%, respectively.

58

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

Table 2 Fitting parameters of Sovová Model for extractions using subcritical propane. T (K)

P (MPa)

Z

W

r

tCER (min)

tFER (min)

kFa (min−1 )

kSa (min−1 )

313 313 353 353 333

8 16 8 16 12

2.146 4.312 11.875 8.769 9.303

0.297 0.246 0.401 0.583 0.365

0.670 0.670 0.670 0.670 0.670

12.166 6.276 2.620 3.206 2.567

41.534 36.550 39.948 38.200 30.012

0.111 0.235 0.788 0.549 0.546

6.924 × 10−3 6.298 × 10−3 1.040 × 10−2 1.541 × 10−2 9.336 × 10−3

Fig. 3. Unsaturated fatty acids distribution in crambe oil samples extracted with subcritical propane - (a) 313 K, 8 MPa; (b) 313 K, 16 MPa; (c) 353 K, 8 MPa; (d) 353 K, 16 MPa; (e) 333 K, 12 MPa and with organic solvents – (f) n-hexane; (g) dichloromethane. Fatty acids: eicosadienoic;

oleic;

linoleic;

linolenic;

gadoleic;

erucic.

Fig. 4. Saturated fatty acids distribution in crambe oil samples extracted with subcritical propane – (a) 313 K, 8 MPa; (b) 313 K, 16 MPa; (c) 353 K, 8 MPa; (d) 353 K, 16 MPa; (e) 333 K, 12 MPa and with organic solvents – (f) n-hexane; (g) dichloromethane. Fatty acids:

palmitic;

estearic;

araquidic;

behenic;

lignoceric.

Similar results were obtained in other studies [13,26,27]. The saturated fatty acids together represent 6% of the total, and, of these, behenic acid (C22:0) is predominant accounting for 2% of the total composition.

In the free glycerol compounds fraction, the free fatty acids (FFA), phytosterols (PHY) and tocopherols (TOC) were considered. The free fatty acids are produced by hydrolytic reactions, where the bond between the ester and triacylglycerol is broken, and this accelerates the oil oxidation process. The free fatty acids quantification (Table 3) revealed significant differences between the oil samples extracted under different conditions, although all samples had low contents (<2%) indicating that the oils can be stored for long periods without the deterioration of lipids. For the extractions using subcritical propane, higher temperatures resulted in higher amounts of FFA due to the action of heat in the hydrolytic reactions, as reported by Osawa et al. [28]. In addition, the phytosterols, a group of sterols found in plants, can be used for plant identification; for instance, brassicasterol is found in species of the Brassicaceae family [29]. In crambe oil, the following phytosterols were identified: brassicasterol, campesterol and ␤-sitosterol. In their quantification (Table 3), the obtained values were 180.79 and 186.44 mg of PHY for each 100 g of oil in the samples extracted with n-hexane and dichloromethane, respectively. The oil extracted with subcritical propane presented corresponding values of 180.00–201.05 mg × (100 goil )−1 , the latter referring to the sample obtained at 353 K and 16 MPa. The temperature and pressure conditions used in the propane extractions affected the phytosterol content. The experiments performed at the highest pressure (16 MPa) showed that the factor with the greatest effect on the extraction of these compounds was the solute vapor pressure rather than the solvent density, as reported by Xu et al. [30] for the extraction of phytosterols from lotus (Nelumbo nucifera Gaertn) bee pollen. Although the quantity of these sterols is dependent on several factors, such as variety, cultivation conditions, storage and extraction method [31], the individual content of each phytosterol in the crambe oil samples (average of 14% for brassicasterol, 29.7% for campesterol and 56.3% for ␤-sitosterol) is similar to the results reported by Lechner et al. [32] and Lalas et al. [26]. In relation to the antioxidants, the tocopherols ␣, ␥ and ␦ were identified and quantified, as shown in Table 4. The results revealed high amounts of tocopherols in the crambe oil (102.35–202.18 mg of tocopherols for 100 g of oil). The highest values were obtained for ␥-tocopherol (74.5–94.4%) followed by ␦-tocopherol (3.7–23.4%) and ␣-tocopherol (1.5–3.4%). El-Beltagi and Mohamed [33] studied the tocopherols composition of five different crops of rapeseed and reported values of 73.02–138.3 mg × (100 goil )−1 . The contents of tocopherols (␣, ␥ and ␦) reported by Ali et al. [34] for five different crops of canola varied from 35.96 to 68.56 mg × (100 goil )−1 . For the samples extracted with organic solvents, dichloromethane provided a lower quantity of tocopherols than n-hexane and propane, which had similar values. These results can be explained by the solvent polarity. As discussed previously, n-hexane is a nonpolar solvent, and it has a high affinity for nonpolar lipids, such as tocopherols [35,36], in contrast with dichloromethane which is a polar solvent. For the oil extracted with propane, it was observed that the total tocopherols composition is influenced by the used temperature and pressure conditions. The temperature has a stronger positive effect, while the pressure shows a weaker negative effect. Thus, for the

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

59

Table 3 Free fatty acids (FFA) and phytosterols (PHY) in the crambe oil extracted with subcritical propane, n-hexane (HEX) and dichloromethane (DCM) as solvents. Compound

Extraction conditions Propane

HEX

313 K, 8 MPa FFA (%) PHY (mg × (100 goil )−1 ) Brassicasterol Campesterol ␤-Sitosterol

1.26 180.00 28.64 51.40 99.96

± ± ± ± ±

313 K, 16 MPa

0.01−2b 0.23d 0.21 0.14 0.31

1.16 184.86 25.97 52.73 106.15

± ± ± ± ±

353 K, 8 MPa

353 K, 16 MPa

1.62 ± 0.04−2a 181.14 ± 0.78d 25.82 ± 0.37 53.82 ± 0.57 101.50 ± 0.98

1.70 201.05 26.83 62.15 112.06

0.04−2b 0.34c 0.28 0.76 0.70

± ± ± ± ±

333 K, 12 MPa

0.05−2a 0.45a 0.45 0.21 0.21

1.24 192.76 29.79 58.24 104.73

± ± ± ± ±

DCM

Boiling point

0.03−2b 0.05b 0.89 0.62 0.22

1.19 180.79 23.42 54.09 103.28

± ± ± ± ±

0.01b 0.41d 0.56 0.54 0.43

1.71 186.44 23.59 55.05 107.80

± ± ± ± ±

0.01a 0.32c 0.32 0.55 0.55

Same letter in the same row indicates no significant difference at 5% confidence level.

Table 4 Tocopherol contents (␣, ␥ and ␦) in the crambe oil. Tocopherol concentration (mg × (100 goil )−1 )

Extraction conditions Solvent

T (K)

Propane

313 313 353 353 333 Boiling point

n-Hexane Dichloromethane

␣-Tocopherol

P (MPa) 8 16 8 16 12

3.04 2.69 3.09 2.95 3.43 2.63 3.48

–

± ± ± ± ± ± ±

␥-Tocopherol

0.16 0.021 0.02 0.05 0.01 0.01 0.05

118.00 132.32 171.57 146.54 155.04 138.59 95.00

± ± ± ± ± ± ±

␦-Tocopherol

0.36 1.63 0.34 1.20 0.88 1.58 0.17

37.38 5.22 27.52 25.28 11.06 22.19 3.97

± ± ± ± ± ± ±

0.06 0.004 0.25 0.31 0.05 0.11 0.007

Total 158.41 140.23 202.18 174.81 169.53 163.41 102.35

± ± ± ± ± ± ±

0.58e 1.65f 0.11a 1.56b 0.82c 1.67d 0.21g

Same letter in the same column indicates no significant difference at 5% confidence level.

experiments performed at the same pressure, an increase in the temperature led to an increase in the vitamin E levels. However, an increase in the pressure led to lower levels of vitamin E, but the effect of this parameter was not as strong as that of the temperature. Thus, it can be concluded that in the extraction procedure higher temperatures and lower pressures lead to higher levels of tocopherols in the oil. A similar finding was reported by Rebolleda et al. [37] and Bong and Loh [38] for the extraction of lipids from corn germ oil and marine microalgae, respectively, both using supercritical CO2 . According to Mendiolla et al. [39], in the case of other seeds, low temperatures and high pressures increase the solubility of pigments, FFA and glycerides and decrease the amount of tocopherols extracted. 3.4. Oxidative stability The thermal stability of crambe oil was determined as a function of the oxidation induction period using a differential scanning calorimeter. The oxidation induction time (t0 ) was obtained from the oxidation curve considering the intersection of the baseline and the tangent line at the edge of the isotherm (Fig. 5, curve A). The experimental extraction conditions, the oxidation induction times (t0 ) for each temperature used in the analyses, and the fitting equations that describe the relationship between t0 at the temperatures of the analyses and t0 presented by Tan et al. [21], along with their determination coefficients R2 , can be observed in Table 5. A straight line was observed for the crambe oil sample (extracted with n-hexane) running under nitrogen flow at 50 mL min−1 and

Fig. 5. Isothermal curves for crambe oil extracted with n-hexane subjected to oxygen flow (A) and nitrogen flow (B) at 413 K.

413 K (Fig. 5, curve B), while only exothermic oxidation curves were observed when the samples were subjected to oxygen flow. The results show that the extraction conditions and the used solvent influence the oxidative stability of crambe oil, since the analyzed samples showed significant differences regarding the oxidative induction time. The longest induction times were observed for the sample extracted with propane at 353 K and

Table 5 Crambe oil oxidation induction time of oils extracted with propane, n-hexane and dichloromethane. Ta DSC analysis (K)

Extraction conditions

Solvent

T (K)

Propane n-Hexane Dichloromethane

353 Boiling point

a b

Temperature in DSC analysis (K). Oxidative induction time.

P (MPa) 16 –

383 t0 b (min)

393

403

420.2 225.3 90.2

265.2 152.95 35.0

155.2 90.4 22.9

Regression equation

R2

T = 217.86−40.70 × log10 t0 T = 222.23−47.32 × log10 t0 T = 172.06−32.14 × log10 t0

0.987 0.993 0.907

413 77.5 53.5 –

60

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61

16 MPa. The sample extracted with dichloromethane showed faster oxidation for all applied temperatures, and, when subjected to 413 K, the sample was already oxidized, and the analysis could not be carried out in this case. It can be concluded that the oxidative stability of the oil is related to the quantity of tocopherols present in the oil, these being the most important antioxidants in vegetable oils [37,40], since they act as oxidation retardants [41,42]. Silva et al. [23] also verified that the oil extracted with subcritical propane presented a longer oxidative stability when compared with the oil extracted in Soxhlet with n-hexane, which can be justified by the greater concentration of tocopherols in the first sample. Other authors have shown that vegetable oils obtained using subcritical propane or supercritical CO2 are less susceptible to oxidation when compared with those extracted using the conventional method with n-hexane [8,9,43]. 4. Conclusions The extraction of crambe oil with subcritical propane provides satisfactory extraction yields within 80 min. The best results were obtained applying the highest pressure and temperature (353 K and 16 MPa) values investigated in this study with a yield of 32.8 wt%. For the studied experimental conditions, the temperature had most influence on the extraction yield. The oil obtained applying different processes and conditions revealed no significant differences in relation to the fatty acids composition, and erucic acid was predominant followed by oleic acid. However, the reduced temperature and pressure may make the extraction of crambe oil with subcritical propane more economically attractive. The free fatty acids content was lower than 2%, and ␤-sitosterol was the predominant phytosterol in the crambe oil. The oil showed high tocopherols concentration, and, in the oxidative stability analysis, the oil extracted with subcritical propane had the longest oxidation induction times. Regarding the extraction kinetics, the Sovová model provided a good fit with the experimental data. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.supflu.2015.05. 026 References [1] B.B. Desai, Seeds Handbook: Biology, Production Processing and Storage, 2nd ed., Marcel Dekker, New York, 2004, pp. 199–232. [2] L. Lazerri, O. Leoni, L. Conte, S. Palmieri, Some technological characteristics and potential uses of Crambe abyssinica products, Ind. Crops Prod. 3 (1994) 103–112. [3] A.S. Carlsson, Plant oils as feedstock alternatives to petroleum—a short survey of potential oil crop platforms, Biochimie 91 (2009) 665–670. [4] A.E. Atabani, A.S. Silitongaa, H.C. Onga, T.M.I. Mahiac, H.H. Masjuki, I.A. Badruddina, H. Fayaz, Non-edible vegetable oils: a critical evaluation of oil extraction, fatty acid compositions, biodiesel production, characteristics, engine performance and emissions production, Renew. Sustain. Energy Rev. 18 (2013) 211–245. [5] S.L. Falasca, N. Flores, M.C. Lamas, S.M. Carballo, A. Anschau, Crambe abyssinica: an almost unknown crop with a promissory future to produce biodiesel in Argentina, Int. J. Hydrogen Energy 35 (2010) 5808–5812. [6] C. Pitol, D.L. Broch, R. Roscoe, Tecnologia e produc¸ão: crambe, Maracaju: Fundac¸ão MS, 2010, pp. 05–06. [7] B. Ahangari, J. Sargolzaei, Extraction of pomegranate seed oil using subcritical propane and supercritical carbon dioxide, Theor. Found. Chem. Eng. 46 (2012) 258–265. [8] G. Nimet, E.A. Silva, F. Palú, C. Dariva, L.S. Freitas, A.M. Neto, L. Cardozo-Filho, Extraction of sunflower (Heliantus annuus L.) oil with supercritical CO2 and subcritical propane: experimental and modeling, Chem. Eng. J. 168 (2011) 262–268, 1996. [9] M.M. Pederssetti, F. Palú, E.A. Silva, J.H. Rohling, L. Cardozo-Filho, C. Dariva, Extraction of canola seed (Brassica napus) oil using compressed propane and supercritical carbon dioxide, J. Food Eng. 102 (2011) 189–196.

[10] M.P. Corso, M.R. Fagundes-Klen, E.A. Silva, L. Cardozo-Filho, J.N. Santos, L.S. Freitas, C. Dariva, Extraction of sesame seed (Sesamun indicum L.) oil using com-pressed propane and supercritical carbon dioxide, J. Supercrit. Fluids 52 (2010) 56–61. [11] L.S. Freitas, J.V. Oliveira, C. Dariva, R.A. Jacques, E.B. Caramão, Extraction of grape seed oil using compressed carbon dioxide and propane: extraction yields and characterization of free glycerol compounds, J. Agric. Food Chem. 56 (2008) 2558–2564. [12] P.M. Ndiaye, M. Lanza, F.W. Tavares, C. Dariva, D. Oliveira, J. Vladimir Oliveira, Phase behavior of olive and soybean oils in compressed propane and n-butane, Braz. J. Chem.Eng. 23 (2006) 405–415. [13] B. Onorevoli, M.E. Machado, C. Dariva, E. Franceschi, L.C. Krause, R.A. Jacques, E.B. Caramão, A one-dimensional and comprehensive two-dimensional gas chromatography study of the oil and the bio-oil of the residual cakes from the seeds of crambe abyssinica, Ind. Crops Prod. 52 (2014) 8–16. [14] A.O.A.C. International, Official Methods of Analysis, vol. 2, 16a ed., 3a rev. AOAC International, Gaitherburg, 1997, pp. 1–43 (Chapter 32). [15] H. Sovová, Rate of the vegetable oil extraction with supercritical CO2 –I modeling of extraction curves, Chem. Eng. Sci. 49 (1994) 409–414. [16] M.C. Ribas, D. Mantovani, J.A. Awadallak, R.L. Canevesi, N.M. Tazinafo, L. Cardozo-Filho, F. Palú, E.A. Silva, Study of candeia oil extraction using pressurized fluids and purification by adsorption process, J. Supercrit. Fluids 92 (2014) 177–182. [17] J.A. Nelder, R. Mead, A simplex method for function minimization, Comput. J. 7 (1965) 308–313. [18] V.A.S. Garcia, V.F. Cabral, E.F. Zanoelo, C. Silva, L. Cardozo-Filho, Extraction of mucuna seed oil using supercritical carbon dioxide to increase the concentration of l-dopa in the defatted meal, J. Supercrit. Fluids 69 (2012) 75–81. [19] Official Methods and Recommended Practices of the American Oil Chemists’ Society (Method AOCS Ce 2-66) (4th), in: R.E. Walker (Ed.), American Oil Chemists Society, Champaign, 1990. [20] L.S. Freitas, R.A. Jacques, M.F. Richter, A.L. Silva, E.B. Caramão, Pressurized liquid extraction of vitamin E from Brazilian grape seed oil, J. Chromatogr. A 1200 (2008) 80–83. [21] C.P. Tan, Y.B. Che Man, J. Selamat, M.S.A. Yusoff, Comparative studies of oxidative stability of edible oils by differential scanning calorimetry and oxidative stability index methods, Food Chem. 76 (2002) 385–389. [22] R. Tir, P.C. Duttas, A.Y. Badjah-Hadj-Ahmed, Effect of the extraction solvent polarity on the sesame seeds oil composition, Eur. J. Lipid Sci. Technol. 114 (2012) 1427–1438. [23] C.M. Silva, A.B. Zanqui, A.K. Gohara, A.H.P. Souza, L. Cardozo-Filho, J.V. Visentainer, L.U.R. Chiavelli, P.R.S. Bittencourt, E.A. Silva, M. Matsushita, Compressed n-propane extraction of lipids and bioactive compounds from perilla (Perilla frutescens), J. Supercrit. Fluids 102 (2015) 1–8. [24] A.S. Pessoa, R. Podestá, J.M. Block, E. Franceschi, C. Dariva, M. Lanza, Extraction of pequi (Caryocar coriaceum) pulp oil using subcritical propane: determination of process yield and fatty acid profile, J. Supercrit. Fluids 101 (2015) 95–103. [25] A.A. Jesus, L.C. Almeida, E.A. Silva, S. Egues, E. Franceschi, M. Fortuny, J. Santos, E.M.B.D. Souza, C. Dariva, Extraction of palm oil using propane, ethanol and its mixtures as compressed solvent, J. Supercrit. Fluids 81 (2013) 245–253. [26] S. Lalas, O. Gortzi, Athanasiadis, Full characterisation of Crambe abyssinica Hochst. seed oil, J. Am. Oil Chem. Soc. 89 (2012) 2253–2258. [27] S.P. Singh, D. Singh, Biodiesel production through de use of different sources and characterization of oils and their esters as the substitute of diesel: a review, Renew. Sustain. Energy Rev. 14 (2010) 200–216. [28] C.C. Osawa, L.A.G. Gonc¸alves, S. Ragazzi, Titulac¸ão potenciométrica aplicada na determinac¸ão de ácidos graxos livres de óleos e gorduras comestíveis, Química. Nova 29 (2006) 593–599. [29] M.K. Gül, S. Amar, Sterols and the phytosterol content in oilseed rape (Brassica napus L.), J. Cell Mol. Biol. 5 (2006) 71–79. [30] X. Xu, J. Dong, X. Mu, L. Sun, Supercritical CO2 extraction of oil, carotenoids, squalene and sterols from lotus (Nelumbo nucifera Gaertn) bee pollen, Food Bioprod. Process. 89 (2011) 47–52. [31] K.M. Phillips, D.M. Ruggio, J.I. Toivo, M.A. Swank, A.H. Simpkins, Free and esterified sterol composition of edible oils and fats, J. Food Compos. Anal. 15 (2002) 123–142. [32] M. Lechner, B. Reitter, E. Lorbeer, Determination of free and esterified sterols in potential new oil seed crops by coupled on-line liquid chromatography-gas-chromatography, Eur. J. Lipid Sci. Technol. 101 (1999) 171–177. [33] H.E.S. El-Beltagi, A.A. Mohamed, Variations in fatty acid composition, glucosinolate profile and some phytochemical contents in selected oil seed rape (Brassica napus L.) cultivars, Grasas Aceites 61 (2010) 143–150. [34] S. Ali, F. Anwar, S. Ashraf, F.N. Talpur, M. Ashraf, Evaluation of canola seeds of different cultivars with special emphasis on the quantification of erucic acid and glucosinolates, Grasas Aceites 60 (2009) 89–95. [35] P. Mehrshahi, C. Johnny, D. Dellapenna, Redefining the metabolic continuity of chloroplasts and er, Trends Plant Sci. 19 (2014) 501–507. [36] T.U. Nwabueze, K.S. Okocha, Extraction performances of polar and non-polar solvents on the physical and chemical indices of African breadfruit (Treculia africana) seed oil, Afr. J. Food Sci. 2 (2008) 119–125. [37] S. Rebolleda, N.R. Rubio, S. Beltrán, M.T. Sanz, M.L. González-Sanjosé, Supercritical fluid extraction of corn germ oil: study of the influence of

K.A. Santos et al. / J. of Supercritical Fluids 104 (2015) 54–61 process parameters on the extraction yield and oil quality, J. Supercrit. Fluids 72 (2012) 270–277. [38] S.C. Bong, S.P. Loh, A study of fatty acid composition and tocopherol content of lipid extracted from marine microalgae Nannochloropsis oculata and Tetraselmis suecica, using solvent extraction and supercritical fluid extraction, Food Research International 20 (2013) 721–729. [39] J.A. Mendiola, D. Garcia-Martinez, F.J. Ruperez, P.J. Martin-Alvarez, G. Reglero, A. Cifuentes, C. Barbas, E. Ibanez, F.J. Senorans, Enrichment of vitamin E from Spirulina platensis microalga by SFE, J. Supercrit. Fluids 43 (2008) 484–489. ´ Potential application of oilseeds as sources of [40] S. Schmidt, J. Polorny, antioxidants for food lipids – a review, Czech J. Food Sci. 23 (2005) 93–102.

61

[41] A. Kamal-Eldin, Effect of fatty acids and tocopherols on the oxidative stability of vegetable oils, Eur. J. Lipid Sci. Technol. 108 (2006) 1051–1061. [42] M. Yang, C. Zheng, Q. Zhou, F. Huang, C. Liu, H. Wang, Minor components and oxidative stability of cold-pressed oil from rapeseed cultivars in China, J. Food Compos. Anal. 29 (2013) 1–9. [43] A.B. Zanqui, D.R. Morais, C.M. Silva, J.M. Santos, L.U.R. Chiavelli, P.R.S. Bittencourt, M.N. Eberlin, J.V. Visentainer, L. Cardozo-Filho, M. Matsushita, Subcritical extraction of Salvia hispanica L. oil with n-propane: composition purity and oxidation stability as compared to the oils obtained by conventional solvent extraction methods, J. Braz. Chem. Soc. 25 (2014) 1–8.