n-hexane sequential extraction method

Industrial Crops and Products 97 (2017) 308–315

Contents lists available at ScienceDirect

Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop

Nontoxic oil preparation from Jatropha curcas L. seeds by an optimized methanol/n-hexane sequential extraction method Yunyun He a,b,1 , Tong Peng a,b,1 , Yanfang Guo a,b , Shushu Li a,b , Yiran Guo c , Lin Tang a,b , Fang Chen a,b,∗ a

Key Laboratory of Bio-resources and Eco-environment of Ministry of Education, College of Life Sciences, Sichuan University, Chengdu, PR China National and Local Joint Engineering Laboratory for Energy Plant Bio-oil Production and Application, Chengdu, PR China c The School of Biological Science and Medical Engineering, Beihang University, Beijing, PR China b

a r t i c l e

i n f o

Article history: Received 1 October 2016 Received in revised form 7 December 2016 Accepted 20 December 2016 Keywords: Jatropha UPLC Phorbol esters By-products PE-free oil

a b s t r a c t Ultrasound-assisted extraction (UAE) of phorbol esters (PEs) from Jatropha curcas L. (JC) seed oil using methanol as the extraction solvent was optimized by response surface methodology (RSM). Jatropha curcas L. seed oil is an excellent feedstock for biodiesel production, in which the presence of major toxic factor (phorbol esters) has limited its industrial applications. The aim of this work is to prepare the nontoxic oil from Jatropha curcas L. whole seeds with the first phorbol esters removal step and the second oil extraction step. The Box-Behnken design was applied to optimize three key parameters in the first phorbol esters removal step including temperature (X1 , ◦ C), extraction time (X2 , min) and liquid-to-solid ratio (X3 , mL:g) for achieving a high extraction efficiency and minimal energy consumption. Ultimately, the optimized conditions were as follows: extraction temperature 36◦ C, extraction time 52 min and liquid-tosolid ratio 5.78:1. Among these three parameters, temperature had the most significant effect on phorbol esters extraction efficiency followed by extraction time and liquid-to-solid ratio. Maximum content of phorbol esters obtained were 0.33 mg/g Jatropha curcas L. seed in the first step and 89.20 ± 2.96% of seed oil could be recovered by n-hexane in the second step. A sensitive ultra-performance liquid chromatography method was also established to determine the phorbol esters content within 5 min, with a limit of detection (LOD) of 0.04 ␮g/g. Since the concentrations of phorbol esters were all below the limit of detection, most of the virtually phorbol ester-free (PE-free) oil was produced to extend the Jatropha biodiesel production chain for household chemicals, cosmetic or jet fuel preparation. Meanwhile, the phorbol esters enriched fraction (PEEF, 71.9 ± 0.3 mg/g) obtained from the first step can be utilized as value-added by-products in pharmaceutical and agro-pharmaceutical fields. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Jatropha curcas L. (JC), a worldwide cultivated energy crop belonging to the Euphorbiaceae family (Makkar and Becker, 2009; Valdes-Rodriguez et al., 2013), is currently gaining lots of importance for its multipurpose use (Kumar and Sharma, 2008). Jatropha plants have excellent drought tolerance and adaption capacity to a large variety of soil conditions (Achten et al., 2008). The average oil content of dry seed on the mass basis is 34.4%, with the seed kernel containing about 60% oil that can be converted into

∗ Corresponding author at: College of life science, Sichuan University, No.24 South Section 1, Yihuan Road, Chengdu, PR China. E-mail address: [email protected] (F. Chen). 1 These authors contributed equally to this paper and should be considered cofirst authors. http://dx.doi.org/10.1016/j.indcrop.2016.12.034 0926-6690/© 2016 Elsevier B.V. All rights reserved.

biodiesel (Goel et al., 2007; Achten et al., 2008). For extraction of the JC oil, two primary methods are identified as mechanical extraction and solvent extraction (Achten et al., 2008). Out of 27-40% oil available in the seed, 60-80% and 70-90% can be recovered respectively by these aforesaid methods (Achten et al., 2007). Moreover, solvent extraction is more economical in a large-scale production, commonly using n-hexane to obtain oil from Jatropha seed cake (Achten et al., 2008). The seed cake remaining after oil extraction is a suitable feedstock for by-products of fertilizer, biogas or low-cost industrial enzymes (Staubmann et al., 1997; Gübitz et al., 1999; Mahanta et al., 2008). However, the abundant oil as well as defatted meal of Jatropha plant is not edible because of the toxic phorbol esters (PEs) (Makkar et al., 1998; Baldini et al., 2014). PEs, a group of diterpenes having tigliane skeleton, are the main toxic principle present in JC seed (Makkar and Becker, 2009; Devappa et al., 2010a; Duran-Pena et al., 2014). Li et al. (2010) reported that PEs had a LD50

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

of 28 mg/kg body weight in mice. The toxic Jatropha oil exhibited severe histological alterations and inflammatory responses on the reconstructed human epidermis and corneal epithelium (Devappa et al., 2013b). PEs can also act as a potent protein kinase C (PKC) activator to amplify the efficacy of carcinogens (Goel et al., 2007; Devappa et al., 2010a). Seeds from non-toxic Mexican genotype of Jatropha have a low level of PEs concentrations of 0.11 mg/g (Makkar and Becker, 1997b) and normally, the non-toxic oil is edible (Makkar and Becker, 2009). As for toxic genotypes, there is a variety of physical, chemical and biological methods established to extract (Devappa et al., 2010b; Devappa et al., 2013a) as well as detoxify PEs (Kumar et al., 2010; Bu et al., 2012; Sadubthummarak et al., 2013; Wang et al., 2013; Guedes et al., 2014). Besides, PEs are used as a starting material to synthesis the important HIV treatment adjuvant candidate medicine prostratin and DPP (Wender et al., 2008; Devappa et al., 2013d). Devappa et al. (2010b) established a method to successfully extract 78-80% of PEs present in the crude Jatropha oil at 60◦ C for 5 min, but to obtain virtually PE-free oil (PEs 23 ␮g/g oil) it cost 60 min. A central composite rotatable design (CCRD) methodology was applied to optimize the extraction conditions and reduced 97.3% of PEs content in Jatropha seed cake after 8 h (Guedes et al., 2014). In terms of industrial applications, high temperature, long extraction time and large amount of solvent consumption are relatively uneconomic for practical considerations. In addition, PEs are currently determined by qualitative methods such as thin layer chromatography (TLC), gas-liquid chromatography (GLC), mass spectrometry (MS), high performance liquid chromatography (HPLC) (Makkar and Becker, 1997a; Haas et al., 2002; Wang et al., 2012) and ultra-performance liquid chromatography (UPLC) (Hua et al., 2015). Since PEs concentration of TPA equivalent is overestimated compared with that of Jatropha factor C1 (JC1) equivalent (Roach et al., 2012; Wang et al., 2012; Devappa et al., 2013a; Hua et al., 2015), in this work JC1 is used as an external standard for determination of the total PEs content. The objective of this study was to establish a reliable method to obtain virtually PE-free oil, wherein RSM was utilized to optimize the conditions of PEs extraction step. Meanwhile, a rapid and sensitive detection method using UPLC was also built for determination of JC1 as well as the total PEs content in a relatively short time.

309

Table 1 Independent variables and the actual experimental values applied for optimization. Independent variable

Symbol

Temperature Time Liquid-to-solid ratio

X1 X2 X3

Coded values −1

0

+1

30 5 10

45 32.5 30

60 60 50

formic acid; solvent B, ultrapure grade water containing 25 mM ammonium formate and 0.1% formic acid. A program of 88% A and 12% B was performed from 0 to 5 min at a flow rate of 0.5 mL/min. The sample injection volume was 2 ␮L with a detection wavelength set at 280 nm (␭max of the UV absorbance of PEs). 2.3. Single factor experiments UAE experiments were conducted on an ultrasound processor (SB-300DTY, Scientz, Ningbo, China). The effect of temperature on PEs extraction efficiency was determined at 30, 40, 50 and 60◦ C with an extraction time of 20 min, a liquid-to-solid ratio of 30:1 and a frequency of 59 kHz. The effect of time on PEs extraction efficiency was determined at 5, 10, 20, 30 and 50 min with a temperature of 40◦ C, a liquid-to-solid ratio of 30:1 and a frequency of 59 kHz. Finally, the effect of liquid-to-solid ratio on PEs extraction efficiency was determined at 5:1, 10:1, 20:1, 40:1 and 60:1 with a temperature of 40◦ C, an extraction time of 50 min and a frequency of 59 kHz. All the samples of these three parameters (temperature, time and liquid-to-solid ratio) were done in triplicate using methanol as the solvent. PEs extraction efficiency was calculated by the relative content of JC1 as follows: PEs extraction efficiency =

m0 Peakarea JC1 × m1 105

(1)

where m0 was the theoretical weight of JC seeds (g) and m1 was the actual weight of JC seeds (g). The 105 was a correction coefficient in this equation to normalize the data of PEs relative contents. 2.4. RSM design and statistical analysis

2. Materials and methods 2.1. Materials and chemicals Jatropha curcas L. (JC) seeds of toxic genotype were collected from cultivated trees (mature, approx. age 8 years) existing in places around Yongren County, Yunnan, China in 2015. Jatropha factor C1 (JC1, 98.3% purity) was prepared and purified by our laboratory (Hua et al., 2015). Water (ultrapure grade) was prepared by an ultrapure water system (WSD, Chengdu, China). Acetonitrile and formic acid were HPLC grade reagents (Tedia, Anhui, China). Methanol, n-hexane and ammonium formate were analytical grade reagents (Kelong, Chengdu, China) 2.2. UPLC method and analysis A UPLC method was built to determine the concentrations of PEs on an ACQUITY UPLC H-class system (Waters, Miford MA, USA) equipped with a QSM quaternary solvent system, a FTN auto sampler manager and an eLambda photodiode array detector. PEs were separated using a CORTECS UPLC C18+ column (1.6 ␮m, 2.1 mm × 150 mm, Waters, USA) at 40◦ C. The conditions of binary solvent system were as follows: solvent A, acetonitrilewater (90/10, v/v) containing 25 mM ammonium formate and 0.1%

UAE parameters for PEs extraction were optimized using RSM (Bezerra et al., 2008) on Design-Expert v8.0.5b software according to the single factor experiments. Box-Behnken design used a three-level-three-factor method with seventeen experimental runs including five replicates at the center point. The extraction variables were temperature (X1 , ◦ C), time (X2 , min) and liquid-to-solid ratio (X3 , mL:g). Their coded values are given in Table 1. The relative content of JC1 was chosen as a response variable to measure the PEs extraction efficiency. Table 2 shows the run order, variable conditions, and the actual experimental values, which is fitted with a second-order polynomial model as follows: Y = ˇ0 +ˇ1 X1 +ˇ2 X2 +ˇ3 X3 +ˇ11 X12 +ˇ22 X22 +ˇ33 X32 +ˇ12 X1 X2 +ˇ13 X1 X3 +ˇ23 X2 X3

(2)

where ˇ0 , ˇ1 , ˇ2 , ˇ3 , ˇ11 , ˇ12 . . .are the regression coefficients. Y is utilized to represent the response variable (relative content of JC1) while X1 , X2 and X3 are the non-coded values for extraction temperature, time and liquid-to-solid ratio, respectively. To determine the interactive effects of variables on PEs extraction efficiency, the experimental data were analyzed by ANOVA with the lack of fit for the quadratic model. A 3D response surface plot was finally generated by Design-Expert software.

310

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

Table 2 Box-Behnken design with the observed responses for PEs extraction efficiency using UAE method. Run

Temperature X1 (◦ C)

Time X2 (min)

Liquid-to-solid ratio X3 (mL:g)

PEs extraction efficiency Experimental values

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

45 60 45 30 30 45 60 45 30 45 45 45 60 45 30 45 60

32.5 60 32.5 32.5 5 32.5 32.5 32.5 32.5 60 32.5 60 32.5 5 60 5 5

30 30 30 50 30 30 10 30 10 50 30 10 50 10 30 50 30

258.44 88.05 258.04 251.14 211.64 284.14 87.99 258.73 288.9 261.25 238.52 274.9 89.01 234.74 266.89 206.41 74.68

2.5. Methanol/n-hexane sequential JC oil extraction 2.5.1. Step one for PEs extraction In this step, methanol was utilized as the only solvent to get PEs under RSM optimized conditions. 2 g JC seed powder was extracted three times in 50 mL sloping bottom centrifuge tubes by two different methods respectively, UAE and heating-assisted extraction (HAE). Moreover, UAE method was performed at an extraction frequency of 59 kHz while HAE method was done without ultrasonic treatment. After each process for PEs extraction, the mixture was filtered and the filtrate was concentrated to get PEEF. The residue obtained in the first UAE process was collected to extract oil in the second step as well.

2.5.2. Step two for oil extraction In this step, the aforementioned residue was extracted by nhexane using soxhlet extraction method according to Liu et al. (2011). The determination of unsaponifiables was based on Adebowale and Adedire (2006). After the oil extraction process, hexane was removed by vacuum evaporation to get JC seed oil, which was mixed with methanol again and extracted under RSM optimized conditions. Thereafter, the suspension was centrifuged at 12000 rpm (Centrifuge 5418, Eppendorf, Hamburg, Germany) for 10 min and the upper methanolic layer was collected for preparation of UPLC detection.

2.5.3. Comparison of PEs extraction solvent and frequency 2 g JC seeds were powdered into 50 mL sloping bottom centrifuge tubes and extracted under RSM optimized conditions by methanol (MeOH), ethanol (EtOH), dichloromethane (DCM) and ethyl acetate (EtAc) at 25 kHz, respectively. Each solvent extracted the seed powder twice. Meanwhile, another series of seed powder was mixed with methanol (1:5.78, w/v) and also extracted using the same method, but at different ultrasonic frequency (25 kHz, 28 kHz, 40 kHz and 59 kHz).

2.5.4. Determination of PEs content in JC seed The total PEs content in JC seed was expressed as JC1 equivalent. All the experiments were done in triplicate and each sample was filtered through a 0.22 ␮m PTFE filter (Jinteng, Tianjing, China) for UPLC preparation. The UPLC method was modified on our previous PEs analysis (Hua et al., 2015; Li et al., 2015). According to self-control method of principle component (Wang et al., 2012;

Devappa et al., 2013a), the final concentrations of PEs can be calculated by the equation as follows:

 n 

cPEs = − 0.387 +



Peakarea (JC1 + JC2 + · · · + Jcn)

i=1

1528

(3)

3. Result and discussion 3.1. Optimization of UAE for PEs extraction by RSM 3.1.1. Single factor analysis Firstly, a series of single factor experiments were performed to directly extract PEs from JC seeds. As shown in Fig. 1a, extraction efficiency increased obviously with increasing UAE temperature up to 40◦ C and then declined rapidly. The temperature of 60◦ C was very close to the boiling point of methanol, so in this experiment, the maximum temperature set at 60◦ C could clearly reflect the effect of high temperature closed to solvent boiling point on PEs extraction efficiency. Compared with the previous study (Devappa et al., 2010b), UAE temperature below 40◦ C was a better choice for lower consumption of energy and solvent. PEs extraction efficiency increased from 5 to 20 min, reaching a maximum at 20 min (Fig. 1b), but with prolonged time, the efficiency gently reduced and tended to be relatively constant. Therefore, optimum extraction time was more than 20 min. A slight reduction in PEs extraction efficiency also appeared with increasing liquid-to-solid ratio at less than 10:1 (Fig. 1c), suggesting that a better liquid-to-solid ratio could fall within the range of less than 10:1. In consideration of the full homogenization between the sample matrix and solvent, liquid-to-solid ratio was designed to start from 5:1. As shown in Fig. 1c, a slight reduction in PEs extraction efficiency appeared with increasing liquid-to-solid ratio at less than 10:1, suggesting that a better liquid-to-solid ratio could fall within the range of less than 10:1. 3.1.2. Fitting the model The actual experimental values are shown in Table 2 and based on the data, the analysis of variance (ANOVA) results for the quadratic polynomial model is summarized in Table 3. The model was significant for its F value of 50.93, with only a 0.01% chance occurred due to an error. Moreover, the determination coefficient (R2 ) was 0.9850, indicating that the model fitted the actual data well (Lee et al., 2010; Majd et al., 2014). The lack of fit value was not significant (P > 0.05), which suggested that the number of experiments was sufficient to determine the effects of independent variables on

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

311

Fig. 1. The efficiency of UAE temperature (a), time (b) and liquid-to-solid ratio (c) for PEs extraction.

Table 3 Analysis of variance (ANOVA) of response surface quadratic model for the prediction of PEs extraction efficiency from JC seeds. Source

Sum of squares

df

Mean Square

F Value

P-value Prob > F

Model X1 X2 X3 X1 X2 X1 X3 X2 X3 X1 2 X2 2 X3 2 Residual Lack of Fit Pure Error Cor Total R2 Adj R2

93010.29 57602.97 3346.44 774.6 438.48 375.97 53.88 28423.55 1230.77 14.38 1420.48 369.37 1051.11 94430.78 0.985 0.9656

9 1 1 1 1 1 1 1 1 1 7 3 4 16

10334.48 57602.97 3346.44 774.6 438.48 375.97 53.88 28423.55 1230.77 14.38 202.93 123.12 262.78

50.93 283.86 16.49 3.82 2.16 1.85 0.27 140.07 6.07 0.071

<0.0001a <0.0001a 0.0048a 0.0917c 0.1850c 0.2157c 0.6222c <0.0001a 0.0433b 0.7978c

0.47

0.7202c

a b c

P < 0.01 highly significant. 0.01 < P < 0.05 significant. P > 0.05 not significant.

PEs extraction efficiency. Since the P value (< 0.0001) was very low, the model terms were statistically significant with a suitable mathematical model given by the second-order polynomial equation as follows:

Y = 259.57-84.85X1 +20.45X2 -9.84X3 -82.16X12 -17.10X22 +1.85X32 − 10.47X1 X2 +9.70X1 X2 +3.67X2 X3

(4)

According to the ANOVA results, X1 , X2 , X1 2 and X2 2 had significantly effect on PEs extraction efficiency (P < 0.05) and among these terms, X1 and X1 2 were highly significant (P < 0.01), suggesting that temperature was the mainly essential factor to influence the dependent variable. In fact, rising temperature could greatly weaken the strong interactions between solute and matrix caused by the Van der Waals force, hydrogen bonds and dipole attraction of solute molecules with sample matrix active positions, thus accelerating the dissecting dynamics process of solute molecules to lower its activation energy and meanwhile reducing the viscosity of the solvent. Consequently, temperature increased the proliferation of solvent into the sample matrix (Mou and Liu, 2001). However, high temperature could also cause more solvent loss and energy consumption. For that matter, choices of extraction temperature between 35◦ C to 40◦ C were better. X3 , X3 2 , X1 X2 , X1 X3 and X2 X3 had no significant effects (P > 0.05) on PEs extraction efficiency. As a result, the final order of factors affecting the dependent variable was temperature, time and liquid-to-solid ratio.

3.1.3. RSM analysis Three-dimensional (3D) response surface graphs were utilized to illustrate the interaction effects of extraction temperature, time and liquid-to-solid ratio (Fig. 2-3). As shown in Fig. 2, the effect of temperature and time at a constant liquid-to-solid ratio (5.78:1) was mainly dominated by the temperature factor. When it was below 35.7◦ C, PEs extraction efficiency increased and then reached a plateau, suggesting that a temperature below 40◦ C could be an optimal condition. Extraction time, however, had little positive effect at less than 51.67 min. The interaction between these two variables had a non-significant effect on PEs extraction efficiency (P > 0.05), but their quadratic variable values were significant (P < 0.05), which might be explained by the highly significant value of extraction temperature. Fig. 3a showed the effect of temperature and liquid-to-solid ratio at a constant time of 52.18 min. The interaction between these two variables was not significant and this could be also due to the highly significant value of extraction temperature. Moreover, the effect of extraction time and liquid-to-solid ratio at a constant temperature (36.08◦ C) was not significant, either (Fig. 3b). Although liquid-to-solid ratio had little interaction with extraction time or temperature, it was still necessary to choose an appropriate value for that lower liquid-to-solid ratio did have better extraction efficiency and consume less solvent. As time prolonged close to 52 min, PEs extraction efficiency increased slightly until reaching a plateau, after which excessive time was not useful for PEs extraction any more. According to the Fick’s second law of diffusion (Majd et al., 2014), a final equilibrium of the solute concentrations between the solid matrix and the bulk solution could cost a certain time. There-

312

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

Fig. 2. Response surface (a) and the corresponding contour plot (b) showing the combined effect of temperature and time at a constant liquid-to-solid ratio of 5.78:1.

Fig. 3. Response surface plots showing the combined effect of temperature and liquid-to-solid ratio at a constant time of 52.18 min (a); and the combined effect of liquidto-solid ratio and time at a constant temperature of 36.08◦ C (b).

fore, this result implied that different containers used might cause some minor changes on PEs extract efficiency for influencing the system equilibrium, but the general trend would not be altered. Since temperature mainly affected the extraction efficiency, optimal conditions fell in a relatively wide range when the predicted efficiency more than 95% with temperature ranged from 32 to 40◦ C and extraction time ranged from 25 to 60 min (Fig. 2b). The liquid-to-solid ratio, as discussed above, could be selected within a range from 5:1 to 10:1. This result revealed that the parameter ranges of the optimal conditions could have a wide operation window when actually utilized for industrial applications. To obtain higher extraction efficiency (> 97%), however, there was a slight reduction in the range of optimal conditions. Furthermore, the total energy consumption was approximately evaluated for each optimum design, in which the heating process consumed the most energy. After overall consideration of the RSM results, the optimized conditions were finally selected for improving PEs extraction process.

3.1.4. Optimization of the extraction conditions To validate the predictive capacity of the given models, optimal conditions were established based on the maximum desirability for higher PEs extraction efficiency and lower energy consumption in the whole extraction process. Therefore, the final optimized conditions were selected including a temperature of 35.7◦ C, an extraction time of 51.67 min and a liquid-to-solid ratio of 5.78:1. Then the UAE method was conducted under optimal conditions with a slightly modified for a temperature of 36◦ C and an extraction time of 52 min. As shown in Table 4, PEs extraction conditions optimized by RSM were reliable and practical for the experimental value of 309.13 ± 0.51 (P > 0.05).

3.1.5. The comparison of extraction methods, solvent and frequency Yields of PEs (as JC1 equivalent) using different extraction methods, solvent and frequency were compared in Fig. 4. All the

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

313

Table 4 Comparison of predicted value and experimental value for the response variables. Independent variables

Dependent variable

Temperature (◦ C)

Time (min)

Liquid-to-solid ratio (mL:g)

Predicted values

Experimental values

36

52

5.78

309.879

309.13 ± 0.51a

a

P > 0.05 (n = 3).

Fig. 4. The comparison of extraction methods (a), solvent (b) and frequency (c).

extraction conditions were carried out in accordance with RSM optimum results. The results of t-test for UAE and HAE showed that there was a significant difference between these two methods. From the first to the third extraction process, when using UAE method, obtained PEs content in JC seed were 319.54 ± 16.52, 11.25 ± 2.12 and 0.92 ± 0.18 ␮g/g, respectively. For HAE method, PEs content were 197.96 ± 9.81, 96.26 ± 9.02 and 26.89 ± 3.90 ␮g/g, respectively. The total PEs extracted by these two methods were almost equal, however, in the first process, UAE method could reach a PEs extraction rate of 96.33% compared with that of 59.68% by HAE method. This result indicated that UAE method was much more efficient than HAE method. In addition, several studies had proved the necessity of UAE method for improving the extraction efficiency of organic compounds (Luque-Garcia and Luque de Castro, 2003; Majd et al., 2014). UAE method with high extraction efficiency might be attributed to the capacity of ultrasound wave to increase the mass transfer. Therefore, in this study UAE method was necessary for PEs extraction. As shown in Fig. 4b, methanol (MeOH), ethanol (EtOH), dichloromethane (DCM) and ethyl acetate (EtAc) were utilized to compare their PEs extraction efficiency. Under the optimal conditions, each solvent extracted the JC seed twice. From the first to the second extraction process, PEs content obtained by methanol were 226.68 ± 8.08 and 28.00 ± 1.26 ␮g/g, respectively. PEs content obtained by ethanol were 207.93 ± 10.94 and 34.77 ± 3.53 ␮g/g, respectively. PEs content obtained by dichloromethane were 111.85 ± 0.27 and 106.60 ± 1.87 ␮g/g, respectively. Finally, PEs content obtained by ethyl acetate were 151.78 ± 7.19 and 50.45 ± 2.04 ␮g/g, respectively. Methanol system obtained more PEs in the first process with an extraction rate of 89.01%, followed by ethanol (85.67%), ethyl acetate (75.05%) and dichloromethane (51.44%) system. The result of t-test revealed that there was a significant difference between the total PEs content obtained by methanol and dichloromethane (P < 0.05). In this experiment, however, the extraction frequency was set at 25 kHz, distinct from the previous frequency of 59 kHz. The PEs extraction rate of the first extraction process was higher at 59 kHz (96.33%) than that was at 25 kHz (89.01%). Thereafter, the effect of extraction frequency was analyzed in Fig. 4c. Without concentration, JC seed was extracted once using methanol at different frequency (59,

40, 28 and 25 kHz). PEs content obtained at 59 kHz was the most (331.71 ± 14.26 ␮g/g), followed by 40 kHz (324.79 ± 5.22 ␮g/g), 28 kHz (321.90 ± 11.94 ␮g/g) and 25 kHz (290.34 ± 22.97 ␮g/g). As a result, the PEs extraction process should chose methanol (Devappa et al., 2010b) as the extraction solvent at an ultrasound frequency of 59 kHz. 3.2. Methanol/n-hexane sequential JC oil preparation In the first step, JC seed powder was extracted under RSM optimized conditions to obtain PEs up to the maximum. Then in the second step, n-hexane was utilized to extract oil from the residue. As shown in Table 5, oil content in JC seed powder was about 38.28%, consistent with what was reported previously of 36.8 ± 0.05% (Subroto et al., 2015). In industry, n-hexane was currently utilized for solvent extraction of Jatropha oil with a relatively high recovery rate. Methanol could also dissolve and extract a small amount of oil (Qian et al., 2010), thus indicating that the liquid-tosolid ratio in this work needed to be appropriate to sustain a lower loss of JC oil in methanol. Therefore, the value of 5.78:1 was selected. After the first treatment process with methanol, about 89.20% seed oil could be recovered by n-hexane. The chromatogram (Fig. 5) of PEs in oil extracted by n-hexane (A) showed eight compounds corresponding to the eight JC peaks, which were demonstrated to be homologues of PEs. Previous study (Hua et al., 2015) had reported that JC8 was an isomer of JC1 and their UV absorbance spectrogram (a1) were almost the same. The unknown peak (a2) was not a homologue of PEs for its ␭max at 197.3 nm. Moreover, Guedes et al. (2014) used 0.11 mg/g of PEs as PMA equivalent in the raw material to be non-toxic. Devappa et al. (2010b) also proved that PEs Table 5 Properties of JC seed and its extract. JCa samples

Wtb (mg g−1 JCa seed)

PEs (␮g g−1 JCa seed)

Total crude oil JCa seeds cake PEEFc from 1st step Oil from 2nd step

382.8 ± 0.6 635.8 ± 8.1 71.9 ± 0.3 341.5 ± 11.3

215.37 ± 2.88 12.16 ± 4.30 319.54 ± 16.52 < LODd

a b c d

JC − Jatropha curcas L. Wt − weight. PEEF − phorbol esters enriched fraction. Limit of detection = 0.04 ␮g/g (n = 3).

314

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315

Fig. 5. UPLC chromatogram of PEs in total oil (A) and oil from MeOH extraction residue (B) using a detection wavelength set at 280 nm.

content less than 23 ␮g/g in Jatropha oil was virtually PE-free (as PMA equivalent). The JC1 equivalent concentrations of PEs were approximately 41.3-fold lower than that of PMA equivalent concentrations (Roach et al., 2012; Devappa et al., 2013a), thus indicating that the actual PEs content less than 2.66 ␮g/g and 0.56 ␮g/g could be considered as non-toxic and virtually PE-free oil (to JC1 equivalent), respectively. In this study, the PEs content remained in JC oil (B) was below the limit of detection (LOD = 0.04 ␮g/g), which suggested that the final oil extracted by n-hexane was actually PE-free. Some properties of JC seed and its extract were also presented in Table 5. Total crude oil contained PEs of 215.37 ± 2.88 ␮g/g. Previous study indicated that on average, 70% of all PEs were retained in the oil (Makkar and Becker, 2009; Devappa et al., 2013c). In this work, about 65% PEs could be obtained from the total JC oil, however, after the first extraction step with methanol, the JC oil extracted by n-hexane in the second step was virtually PE-free. As a result, nontoxic JC oil was prepared by two sequential steps using two solvents, methanol 1st and n-hexane 2nd . In addition, the unsaponifiable matter in total crude oil and oil from the second step was 0.315 ± 0.068% and 0.055 ± 0.009%, respectively. According to Adebowale and Adedire (2006), the unsaponifiable matter content mainly composed of phytate, polar lipids and PEs in Jatropha oil was high, however, the corresponding peaks in the chromatogram of the final oil were substantially reduced, which suggested that the unsaponifiable matter might be removed in the first step as well. To measure the trace of PEs content in nontoxic Jatropha oil, a novel UPLC-UV method was subsequently established with a satisfactory resolution, wherein the total test time for each sample was only 5 min. The sub-2 ␮m core-shell particles column showed not only excellent column efficiency (selectivity, peak shape and signal-to-noise ratio) but also lower backpressure than sub-2 ␮m conventional silica column (comparison of the kinetic performance and retentivity of sub-2 ␮m core-shell, hybrid and conventional bare silica phases in hydrophilic interaction) (Heaton and Mccalley, 2014). In this study, a 1.6 ␮m core-shell C18+ was applied to achieve a better column efficiency, of which more than 50% of the assay time compared with our previous study (Hua et al., 2015) was reduced. Meanwhile, estimated for PEs in the analyzed matrices, the limit of detection (LOD) defined as signal to noise ratio of 3 (S/N = 3) was 0.04 ␮g/g and the limit of quantitation (LOQ) defined as signal to noise of 10 (S/N = 10) was 0.13 ␮g/g. The use of UPLC resulted in a low detection limit (LOD = 0.04 ␮g/g) blow the tandem mass spectrometry method for which a LOD of 0.07 ␮g/g had been reported

(Baldini et al., 2014). These results also suggested that the UPLC-UV method built in this study was rapid and sensitive for determination of PEs content. 4. Conclusion A kind of nontoxic Jatropha oil was prepared from Jatropha curcas L. whole seed using a sequential extraction method (methanol 1st and n-hexane 2nd ). The first step for PEs extraction was optimized by RSM to obtain a maximum content of PEs (319.54 ± 16.52 ␮g/g). Meanwhile, nontoxic JC oil extracted by nhexane from the second step was virtually PE-free with a recovery rate of 89.20 ± 2.96% in the total oil. And most of the unsaponifiable matter seemed to have been removed from the first step as well. A rapid and sensitive detection of PEs content determination method using UPLC was built for PEs high throughput screening in Jatropha products. Nontoxic Jatropha oil extends the Jatropha biodiesel production chain for household chemicals, cosmetic or jet fuel preparation, which also has the potential to be made into edible oil. As a value-added by-product, the PEEF (71.9 ± 0.3 mg/g) can be utilized in pharmaceutical and agro-pharmaceutical fields (Liu et al., 2013). Acknowledgements This work was supported by science and technology support project of Sichuan province (No. 2011JZ0002). And the authors also appreciate Wan Hua (M.S., College of Life Science, Sichuan University) and Zhanguo Wang (Ph.D., College of pharmacy, Sichuan University) for proving the JC1 fraction. References Achten, W.M., Mathijs, E., Verchot, L., Singh, V.P., Aerts, R., Muys, B., 2007. Jatropha biodiesel fueling sustainability? Biofuels Bioprod. Biorefin. 1, 283–291. Achten, W.M.J., Verchot, L., Franken, Y.J., Mathijs, E., Singh, V.P., Aerts, R., Muys, B., 2008. Jatropha bio-diesel production and use. Biomass Bioenergy 32, 1063–1084. Adebowale, K.O., Adedire, C.O., 2006. Chemical composition and insecticidal properties of the underutilized Jatropha curcas seed oil. Afr. J. Biotechnol. 5, 901–906. Baldini, M., Ferfuia, C., Bortolomeazzi, R., Verardo, G., Pascali, J., Piasentier, E., Franceschi, L., 2014. Determination of phorbol esters in seeds and leaves of Jatropha curcas and in animal tissue by high-performance liquid chromatography tandem mass spectrometry. Ind. Crops Prod. 59, 268–276.

Y. He et al. / Industrial Crops and Products 97 (2017) 308–315 Bezerra, M.A., Santelli, R.E., Oliveira, E.P., Villar, L.S., Escaleira, L.A., 2008. Response surface methodology (RSM) as a tool for optimization in analytical chemistry. Talanta 76, 965–977. Bu, Y., Ha, B.T.N., Eunice, Y., Chueng, L.L., Hong, Y., 2012. Light induced degradation of phorbol esters. Ecotoxicol. Environ. Saf. 84, 268–273. Devappa, R.K., Makkar, H.P.S., Becker, K., 2010a. Jatropha toxicity–a review. J. Toxicol. Environ. Health Part B Crit. Rev. 13, 476–507. Devappa, R.K., Makkar, H.P.S., Becker, K., 2010b. Optimization of conditions for the extraction of phorbol esters from Jatropha oil. Biomass Bioenergy 34, 1125–1133. Devappa, R.K., Bingham, J.-P., Khanal, S.K., 2013a. High performance liquid chromatography method for rapid quantification of phorbol esters in Jatropha curcas seed. Ind. Crops Prod. 49, 211–219. Devappa, R.K., Roach, J.S., Makkar, H.P.S., Becker, K., 2013b. Occular and dermal toxicity of Jatropha curcas phorbol esters. Ecotoxicol. Environ. Saf. 94, 172–178. Devappa, R.K., Makkar, H.P.S., Becker, K., 2013c. Shelf-life of isolated phorbol esters from Jatropha curcas oil. Ind. Crops Prod. 49, 454–461. Devappa, R.K., Malakar, C.C., Makkar, H.P.S., Becker, K., 2013d. Pharmaceutical potential of phorbol esters from Jatropha curcas oil. Nat. Prod. Res. 27, 1459–1462. Duran-Pena, M.J., Botubol Ares, J.M., Collado, I.G., Hernandez-Galan, R., 2014. Biologically active diterpenes containing a gem-dimethylcyclopropane subunit: an intriguing source of PKC modulators. Nat. Prod. Rep. 31, 940–952. Gübitz, G.M., Mittelbach, M., Trabi, M., 1999. Exploitation of the tropical oil seed plant Jatropha curcas L. Bioresour. Technol. 67, 73–82. Goel, G., Makkar, H.P.S., Francis, G., Becker, K., 2007. Phorbol esters: structure, biological activity, and toxicity in animals. Int. J. Toxicol. 26, 279–288. Guedes, R.E., Cruz, F.d.a., Lima, M.C.d., Sant’Ana, L.D.O., Castro, R.N., Mendes, M.F., 2014. Detoxification of Jatropha curcas seed cake using chemical treatment: analysis with a central composite rotatable design. Ind. Crops Prod. 52, 537–543. Haas, W., Sterk, H., Mittelbach, M., 2002. Novel 12-deoxy-16-hydroxyphorbol diesters isolated from the seed oil of Jatropha curcas. J. Nat. Prod. 65, 1434–1440. Heaton, J.C., Mccalley, D.V., 2014. Comparison of the kinetic performance and retentivity of sub-2 (m core-shell, hybrid and conventional bare silica phases in hydrophilic interaction chromatography. J. Chromatogr. A 1371, 106–116. Hua, W., Hu, H., Chen, F., Tang, L., Peng, T., Wang, Z., 2015. Rapid isolation and purification of phorbol esters from Jatropha curcas by high-speed countercurrent chromatography. J. Agric. Food Chem. 63, 2767–2772. Kumar, A., Sharma, S., 2008. An evaluation of multipurpose oil seed crop for industrial uses (Jatropha curcas L.): a review. Ind. Crops Prod. 28, 1–10. Kumar, V., Makkar, H.P.S., Becker, K., 2010. Dietary inclusion of detoxified Jatropha curcas kernel meal: effects on growth performance and metabolic efficiency in common carp Cyprinus carpio L. Fish Physiol. Biochem. 36, 1159–1170. Lee, M.H., Behera, S.K., Park, H.S., 2010. Optimization of operational parameters for ethanol production from Korean food waste leachate. Int. J. Environ. Sci. Technol. 7, 157–164. Li, C.Y., Devappa, R.K., Liu, J.X., Lv, J.M., Makkar, H.P.S., Becker, K., 2010. Toxicity of Jatropha curcas phorbol esters in mice. Food Chem. Toxicol. 48, 620–625. Li, Q.-I., Yan, F., Peng, T., Tang, L., 2015. Investigation of Jatropha curcas seed cake detoxification at low cost using UPLC. Res. Dev. Mark. 31, 1411–1414. Liu, S.Y., Zhang, H., Wang, H., Song, J., Yan, R.F., 2011. Extraction of oil from jatropha seeds. Adv. Mater. Res. 183–185, 1652–1656. Liu, X., Li, L., Li, W., Lu, D., Chen, F., Li, J., 2013. Quantitative determination of phorbol ester derivatives in Chinese Jatropha curcas seeds by

315

high-performance liquid chromatography/mass spectrometry. Ind. Crops Prod. 47, 29–32. Luque-Garcia, J.L., Luque de Castro, M.D., 2003. Ultrasound: a powerful tool for leaching. TrAC Trends Anal. Chem. 22, 41–47. Mahanta, N., Gupta, A., Khare, S.K., 2008. Production of protease and lipase by solvent tolerant Pseudomonas aeruginosa PseA in solid-state fermentation using Jatropha curcas seed cake as substrate. Bioresour. Technol. 99, 1729–1735. Majd, M.H., Rajaei, A., Bashi, D.S., Mortazavi, S.A., Bolourian, S., 2014. Optimization of ultrasonic-assisted extraction of phenolic compounds from bovine pennyroyal (Phlomidoschema parviflorum) leaves using response surface methodology. Ind. Crops Prod. 57, 195–202. Makkar, H.P.S., Becker, K., 1997a. Studies on nutritive potential and toxic constituents of different provenances of Jatropha curcas. J. Agric. Food Chem. 45, 3152–3157. Makkar, H.P.S., Becker, K., 1997b. In: Gubitz, G.M., Mittelbach, M., Trabi, M. (Eds.), Biofuels and Industrial Products from Jatropha Curcas. Dbv-Verlag für die Technische Universität Graz, Graz, pp. 190–205. Makkar, H.P.S., Becker, K., 2009. Jatropha curcas, a promising crop for the generation of biodiesel and value-added coproducts. Eur. J. Lipid Sci. Technol. 111, 773–787. Makkar, H.P.S., Becker, K., Schmook, B., 1998. Edible provenances of Jatropha curcas from Quintana Roo state of Mexico and effect of roasting on antinutrient and toxic factors in seeds. Plant Foods Hum. Nutr. 52, 31–36. Mou, S., Liu, Y., 2001. The principle and application of accelerated solvent extraction. Mod. Sci. Instrum. 20, 299–300. Qian, J., Shi, H., Zhi, Y., 2010. Preparation of biodiesel from Jatropha curcas L. oil produced by two-phase solvent extraction. Bioresour. Technol. 101, 7036–7042. Roach, J.S., Devappa, R.K., Makkar, H.P.S., Becker, K., 2012. Isolation, characterization, stability and bioactivity of Jatropha curcas phorbol esters. Fitoterpia 83, 586–592. Sadubthummarak, U., Parkpian, P., Ruchirawat, M., Kongchum, M., Delaune, R.D., 2013. Potential treatments to reduce phorbol esters levels in jatropha seed cake for improving the value added product. J. Environ. Sci. Health Part B 48, 974–982. Staubmann, R., Foidl, G., Foidl, N., Gübitz, G.M., Lafferty, R.M., Arbizu, V.M.V., Steiner, W., 1997. Biogas production from Jatropha curcas press-cake. Appl. Biochem. Biotechnol. 63–65, 457–467. Subroto, E., Manurung, R., Heeres, H.J., Broekhuis, A.A., 2015. Optimization of mechanical oil extraction from Jatropha curcas L. kernel using response surface method. Ind. Crops Prod. 63, 294–302. Valdes-Rodriguez, O.A., Sanchez-Sanchez, O., Perez-Vazquez, A., Caplan, J., 2013. The Mexican non-toxic Jatropha curcas L., food resource or biofuel? Ethnobot. Res. Appl. 11, 1–8. Wang, Z., Tang, L., Hu, H., Guo, Y., Peng, T., Yan, F., Chen, F., 2012. Metabolic profiling assisted quality control of phorbolesters in Jatropha curcas seed by high-Performance liquid chromatography using a fused-Core column. J. Agric. Food Chem. 60, 9567–9572. Wang, X.-H., Ou, L., Fu, L.-L., Zheng, S., Lou, J.-D., Gomes-Laranjo, J., Li, J., Zhang, C., 2013. Detoxification of Jatropha curcas kernel cake by a novel Streptomyces fimicarius strain. J. Hazard. Mater. 260, 238–246. Wender, P.A., Kee, J.M., Warrington, J.M., 2008. Practical synthesis of prostratin, DPP, and their analogs: adjuvant leads against latent HIV. Science 320, 649–652.