Optimized extraction and quality evaluation of Niger seed oil via microwave‐pulsed electric field pretreatments

Abstract In this study, oil extraction from Niger seeds was evaluated with different microwave irradiation times (0–200 s) and pulsed electric fields (PEF) intensities (0–5 kV/cm) as pretreatments. Then, oil extraction was completed with a screw press at different rotation speeds (11–57 rpm). Quality parameters including extraction efficiency, acidity and peroxide values (PVs), chlorophyll, and phenolic contents along with fatty acid profiles and tocopherol levels of the extracted oils were determined as responses. With enhancements in microwave time, PEF intensity and press rotation, the chlorophyll contents, acidity/PVs, and total phenolics of oils increased similar to oil extraction efficiency although it was reduced later. The optimized conditions selected by response surface methodology were determined as 156.23 s, 1.18 kV/cm, and 20 rpm for microwave time, PEF intensity and press speed, respectively. Fatty acid analysis revealed that linoleic acid was the most predominant fatty acid in the extracted oil. Application of the mentioned pretreatments may lead to a reduction in unsaturated fatty acids and escalation of saturated ones (p < .05). High‐performance liquid chromatography results indicated that α‐tocopherols are the most common tocopherols in Niger seed oil and microwave‐PEF pretreatments may lead to 2.79% increase in tocopherols content.

plant is similar to sunflower and safflower with a more quantity of linoleic acid (over than 85%). Niger seeds oil could be used as a replacement for olive and sesame seeds oils in pharmaceutical industries; moreover, it may be used in the production of soaps, paints, grease, and perfumes (Pradhan, Mishra, & Paikary, 1995).
Furthermore, this oil contains considerable quantities of antioxidants (Ramadan & Moersel, 2003). Also, the obtained Niger meal through oil extraction process contains about 35% protein and 23% raw fibers (Pradhan et al., 1995), that may be used in feeding of animals and birds as well as humans and also in the preparation of fertilizers (Adarsh, Kumari, & Devi, 2014).
Microwave irradiations as nonionized electromagnetic waves with the frequency of 300 MHz to 200 GHz result in heating in a selected rout without any heat loss to environment such as heating in enclosed systems. The heating mechanism may lead to a decrease in extraction time compared to other extraction approaches. The effect of this process is performed by two phenomena including ionic transmission and dipole rotation that in most cases happen simultaneously (Jafari, Mahdavee Khazaei, & Assadpour, 2019). In oil seeds, water as a dipolar compound is found in high amounts, along with other compounds such as salt and protein which may act as dielectric compounds (Sultana, Anwar, & Przybylski, 2007).
PEF processing includes short-term pulses (µs) with a high intensity in a strong electrical field on food materials. The application of PEF due to enhancements of cell permeation might increase oil extraction efficiency and there have been some reports on the positive effects of this process on the qualitative and sensory properties of the extracted oil (Bakhshabadi, Mirzaei, Ghodsvali, Jafari, & Ziaiifar, 2018;Boussetta, Soichi, Lanoiselle, & Vorobiev, 2014;Sarkis, Boussetta, Tessaro, Marczak, & Vorobiev, 2015). For instance, PEF has been applied for fat/oil production from microalgae (La et al., 2016). They used PEF with a low energy and expressed that it could replace the previous common fat extraction approaches.
Currently, regarding the population increase and growing demands for healthy and functional oils, it is necessary to work on nonconventional sources of edible oils. On the other hand, novel and emerging processes could have promising results on improving the extraction efficiency and quality of these oils. So, we proposed this study for the first time to optimize the oil extraction from Niger seeds, with the use of microwave-PEF approach as pretreatments and applying response surface methodology (RSM) as the optimization technique.

| MATERIAL S AND ME THODS
Niger seeds (containing 40% oil) were provided from Fars province (Iran). Chemical materials including sodium hydroxide, phenolphthalein, Folin-Ciocalteu reagent, methanol, ethanol, hexane, and acetonitrile were purchased from Merck and Sigma companies. Other applied chemicals were of analytical grade.

| Pretreatments of Niger seeds and oil extraction
For this purpose, the seeds were firstly treated by microwave with the power of 900 W in different times (0-200 s); then, they were treated by three PEF intensities (0-5 kV/cm). The applied power in PEF chamber was 30 pulse with the width of 20 micro seconds (Kittiphoom & Sutasinee, 2015). Afterward, the oil extraction was completed with a screw press at different rotation rates (11-57 rpm).

| Determination of oil extraction efficiency and its quality
Extraction efficiency was determined gravimetrically by the following equation: Refractive index of the extracted oils was detected at the temperature of 25°C according to AOCS (Society & Firestone, 1994).
Chlorophyll content was measured regarding the method described by Pokorny, Kalinova, and Dysseler (1995). To determine acidity, the method of AOCS 3-63, 1993 was applied. Firstly, 5 g oil was mixed with 20-30 ml ethanol and titrated with 0.1 N NaOH in the presence of phenolphthalein and acidity value was determined with the following equation: where N indicates normality of sodium hydroxide (NaOH), V presents the volume of consumed NaOH, W represents the weight of sample (g), and A expresses fatty acids content based on oleic acid in 100 g sample.
Peroxide value (PV) was analyzed according to AOCS Cd 8-53, 1993. Briefly, 5 g oil was moved into a beaker (250 ml), then 300 ml acetic-acid-chloroform with the proportion of 2:3 was (1) Oil extraction efficiency (%) = Extracted oil (g) added; after homogenization, 0.5 ml saturated iodide potassium was added and was left for 1 min in darkness. In the final solution, 30 ml distilled water was added and followed by titration using sodium thiosulfate 0.1 M till obtaining a yellow color. PV was determined as: where S is the consumed volume of sodium thiosulfate in mL, M indicates molarity of sodium thiosulfate, W represents the weight of sample (extracted oil) in g, and P demonstrates the PV based on milli equivalents of oxygen/kg of oil (meq. kg −1 oil).

| Determination of total phenolic content
Total phenolic content (TPC) content was measured with Folin-Ciocalteu reagent. In this regard, 1 g of each sample was mixed with 3 ml methanol: water (90:10), homogenized for 4 min and centrifuged (1,008 g) for 5 min; then, 20 µl of the supernatant solution was mixed with 8.2 ml water and 0.5 ml Folin-Ciocalteu reagent.
After 5 min, 1 ml sodium carbonate 10% was added to the solution and was left at darkness. After 1 hr, the absorbance was recorded using a spectrophotometer at 765 nm. To construct standard graph, gallic acid (0-1000 µg mL −1 ) was used, and the final TPC content was reported as mg GAE kg −1 Sample (Bail, Stuebiger, Krist, Unterweger, & Buchbauer, 2008).

| Analysis of fatty acids profile by gas chromatography
Firstly, methyl esters of fatty acids were prepared and the analysis of fatty acid profile was done by AOCS Ce 2-66, 1993. For this purpose, a gas chromatography (GC) equipped with silicon hairy column number 70 (length of 60 m and diameter of 0.25 µm) was used. The initial temperature was set on 80°C, and with a temperature enhancement rate of 15°C min −1 , the temperature reached to 200°C which was kept for 10 min; then the temperature increased up to 220°C and was kept for the next 5 min. The temperature of injection valve and detector temperature was set on 210°C and the flow rate of gas (helium) was adjusted on 1 ml min −1 . Finally, the obtained peak area by the GC was compared to the standard graph and the quantity of each fatty acid was detected and reported as percentage. (LiChrosorb) with the diameter of 250 × 4.5 ml and particle size of 5 µm was applied along with a florescence indicator. Mobile phase of acetonitrile in combination of distilled water with the proportion of 95-5 was selected and its rate was adjusted on 0.6 µl min −1 .

| Determination of tocopherols by highperformance liquid chromatography
Based on the retention time of tocopherols and the obtained chromatogram of the injected oil samples, amount of tocopherols was determined.

| Statistical analysis
Response surface methodology, with the use of Box-Behnken design was applied for the evaluation of the influence of independent variables including microwave irradiation power (X 1 ), PEF intensity (X 2 ), and press rotational speed (X 3 ) on different responses. To evaluate response surface behavior, a multivariate second-order equation was designed for each independent variable. Finally, to compare the fatty acid profiles and tocopherol levels, a factorial design by SAS software was applied. The comparison of obtained means was done by the Duncan test.

| The effect of microwave-PEF pretreatments on oil extraction efficiency
To improve oil extraction efficiency from Niger seeds, second-order multivariate model was applied as the optimized model (Table 1).
Based on the ANOVA results as presented in Table 2, the linear effects of microwave time, PEF intensity and press speed on oil extraction efficiency were detected significant (p < .05). The effects of second-order parameters of the studied variables on extraction efficiency were also significant, but the interactive effects between microwave time with PEF intensity were nonsignificant (p > .05).
Also, it was found that screw press rotational speed and quadratic effect of microwave irradiation time had the highest effects on oil extraction efficiency.
As shown in Figure 1, an increase in microwave irradiation time as well as PEF intensity firstly enhanced the oil extraction efficiency but with more enhancements of the above-mentioned parameters, extraction efficiency was decreased. Improvement in oil extraction efficiency by microwave irradiation could be related to creation of more fracture in the cells containing oil during microwave pretreatment (Bakhshabadi, Mirzaei, Ghodsvali, Jafari, Ziaiifar, et al., 2017;Uquiche, Jeréz, & Ortíz, 2008). Moreover, it has been reported that increase of oil extraction efficiency when applying microwave may be due to the destruction/denaturation of proteins (Mohamed & Awatif, 1998). The achieved results of this section were in agreement with other studies (Momeny, Rahmati, & Ramli, 2012;Nde, Boldor, & Astete, 2015;Terigar, Balasubramanian, Sabliov, Lima, & Boldor, 2011). In terms of PEF, the causes of oil extraction efficiency enhancements may be attributed to electrical destruction of the cells and more penetration possibility of solvent into them (Schroeder, Buckow, & Knoerzer, 2009) which is similar to the results of other researchers (Bakhshabadi et al., 2018;Guderjan, Töpfl, Angersbach, & Knorr, 2005). Further decrease in oil extraction efficiency at higher microwave irradiation time and PEF intensity could be explained by more destruction of internal structures of cells and blocking of oil extraction pores/routes. Bakhshabadi et al. (2018) demonstrated that application of high PEF intensity may lead to a decrease in oil extraction efficiency which were in agreement with the obtained results of the current study. Increase in screw press rotational TA B L E 2 Analysis of variance (ANOVA) for determined parameters in oil extraction by microwave-PEF pretreatment

| The effects of independent variables on the quality of extracted oil
According to the results in Table 1, all evaluated models on the refractive index (RI) of extracted oil were not significant (p > .05) and the effects of all variables including microwave time, PEF intensity, and screw press rotational speed on RI were almost not significant.
The value of RI for all of the studied samples was equal to 1.478.
RI is mostly used as a criterion of the oil purity determination. This parameter increases at higher chain lengths of fatty acids (while the relation is not linear). Also, RI has considerable impacts on controlling the catalytic hydrogenation and isomerization processes. Moreover, temperature and saturation rate are effective on RI as well (Uquiche et al., 2008). The findings of the present study were in agreement with the obtained results by Tale Masouleh, Asadollahi, and Eshaghi (2015).
The ANOVA results in Table 2  were also detected significant on chlorophyll content of the extracted oil. Based on the results of Table 1, the most optimized model explaining the effects of microwave irradiation on chlorophyll content was Quadratic. Figure 2 indicates the interactive effects of microwave irradiation time and PEF intensity on F I G U R E 1 3D graphs of (a) microwave irradiation time and pulsed electric field intensity (b) microwave time and screw press rotational speed (c) pulsed electric field intensity and screw press rotational time on oil extraction efficiency chlorophyll content of extracted oils from Niger seeds. It can be seen that at higher rates of microwave irradiation time and PEF intensity, chlorophyll content increases, which is more obvious in low irradiation times and PEF intensities. Moreover, with the increase in screw press rotational speed, chlorophyll content improves. The possible reason of chlorophyll increase with enhancements in microwave time, PEF intensity, and screw press speed may be the high solubility and penetration of the chlorophylls into the extracted oil sample as reported in the study of (Megahed, 2001 F I G U R E 2 3D graphs of (a) microwave irradiation time and pulsed electric field intensity (b) microwave time and screw press rotational speed (c) pulsed electric field intensity and screw press rotational speed on chlorophyll content of the extracted oil F I G U R E 3 3D graphs of (a) microwave irradiation time and pulsed electric field intensity (b) irradiation time and screw press rotational time (c) pulsed electric field intensity and screw press rotational speed on oil acidity Ghodsvali, Jafari, Ziaiefar, et al., 2017;Guderjan, Elez-Martínez, & Knorr, 2007).
The provided data in Table 2 as well as suggested model for the acidity value of the extracted oil in Table 3 revealed that the highest effect on acidity value for the extracted oil was in conditions when the combined pretreatment of microwave-PEF intensity was applied which affected the oil extraction significantly by linear and quadratic modes of screw press rotational speed. On the other hand, it was observed that an increase in microwave time, PEF intensity and press rotational speed may lead to an enhancement in acidity value of the extracted oil ( Figure 3). This could be attributed to higher activity of lipase enzyme leading to free fatty acid production which are considered as undesirable compounds in edible oils (Guderjan et al., 2007). Puértolas and de Marañón (2015), reported similar results to the present study. At higher press speeds, the oil acidity enhancement occurs which might be due to temperature increase during heating process (Puértolas & de Marañón, 2015) which is in agreement with previous studies (Amalia Kartika, Pontalier, & Rigal, 2005;Sriti et al., 2012). Finally, oil acidity increase at higher microwave irradiation times might be attributed to the chemical degradation of triglycerides and production of free fatty acids. Lipolytic enzymes like lipase are placed in the internal parts of cells; in normal cells, these enzymes are not able to attack fats/oils. But during extraction and at high temperatures, the membranes of cells are damaged physically, therefore lipase enzyme starts its activities (Ghavami, Gharachorloo, & Ezatpanah, 2003). The achieved results of this section were in agreement with the study of Kittiphoom and Sutasinee (2015) and Veldsink et al. (1999); but, opposite to the results of Uquiche et al. (2008) (Kittiphoom & Sutasinee, 2015;Uquiche et al., 2008;Veldsink et al., 1999).
Peroxide value is associated with hydroperoxides within extracted oil which is affected by temperature, time, and fatty acid profile (Zhang et al., 2010). Higher PVs show a developed production of secondary products by lipid oxidation such as carbonyls, aldehydes and conjugated di-en; therefore, PV determination is an essential factor for oxidation process analysis. Our results in Table 2 revealed that the linear effects of microwave time, PEF intensity, and screw speed on the PV were significant (p < .001). Moreover, the quadratic effects (except with screw press rotational speed) unlike the interactive effects were also significant. As shown in Figure 4, with an increase in PEF intensity, PV of the extracted oils increased which may be attributed to the high oxidation of fatty acids at higher temperatures when PEF intensity increased (Guderjan et al., 2007). Zeng et al., (2010) reported similar results. Also, higher microwave irradiation times, due to temperature enhancements, might lead to a higher oxidation rate and then, increased PVs; these results were in agreement with the some other studies (Hassanein, El-Shami, & El-Mallah, 2003;Valentova, Novotna, Svoboda, Schwarz, & Kas, 2000).
Finally, an increase in screw press rotational speed, resulted in higher PVs of the extracted oils. Table 3 presents the obtained models of PV for the oils extracted by combined PEF-microwave pretreatment demonstrating a high impact of PEF intensities on the PVs.
Phenolic compounds are an important group of secondary metabolites with antioxidant activities due to having hydroxyl groups in their structures. The application of phenolic compounds in food industry is increasing, because these compounds may retard lipid oxidation and improve qualitative and nutrition properties of food products (Assadpour, Jafari, & Esfanjani, 2017;Faridi Esfanjani, Assadpour, & Jafari, 2018). According to ANOVA results, the linear effects of different parameters on phenolic compounds were significant. As shown in Figure 5, at higher microwave irradiation times, PEF intensity, and screw speeds, total phenolic compounds of the extracted oils were increased. The heating effect of microwave energy is associated with The effects of independent variables on total phenolic compounds of oils F I G U R E 5 3D graphs of (a) microwave irradiation time and pulse electric field (b) microwave irradiation time and screw press rotational time (c) pulsed electric field intensity and screw rotational speed on total phenolic compounds content its direct effects on molecules with dipolar rotational mechanisms and ionic transmission. Polar molecules such as phenolic compounds and ionic solutions absorb microwave energy due to having dipolar torque leading to temperature enhancements and increase of their transfer into the extracted oils (Proestos & Komaitis, 2008) as shown in some other studies as well (Beejmohun et al., 2007;Jiao et al., 2014). Also, Rombaut et al. (2015) reported that at higher press rotational speeds, total phenolic compounds increased.

| Optimization of oil extraction process from Niger seeds
To find the optimized conditions of oil extraction from Niger seeds with the use of microwave-PEF in the range of applied independent variables, (microwave time of 0-200 s, PEF intensity at 0-5 kV/cm and press rotational speed ranged from 11-57 rpm), the targets were considered as the maximum oil extraction efficiency and total phenolic compounds content, and the minimum acidity and PVs. The desirability in the optimized conditions was determined as 0.789. It was found that microwave time of 156.34 s, PEF intensity of 1.18 kV/cm, and press speed of 20 rpm were the best conditions so that an oil with 35% extraction efficiency, RI = 1.478, acidity of 1.94%, and total phenolics of 410 ppm will be produced.

| Influence of microwave-PEF pretreatment on the fatty acid profile of oil samples
The profile of fatty acids in extracted oil of Niger seeds both at microwave-PEF conditions and standard sample has been shown in Table 4 and Figure 6. In both samples, linoleic acid was the most predominant fatty acid in the oil of Niger seeds. As mentioned, when combined microwave-PEF is used as a pretreatment in oil extraction process, the quantity of unsaturated fatty acids (oleic and linoleic acids) decreased and the amount of saturated fatty acids (palmitic and stearic acids) increased (p < .05). This might be attributed to the susceptibility of unsaturated fatty acids against high temperatures. It has been reported previously that the predominant fatty acid in Niger seeds is linoleic acid which may be varied based on the variety of seeds (Getinet & Teklewold, 1995;Pradhan et al., 1995). Also, the composition of fatty acids in oilseeds might be affected by variety, cultivation conditions, and ripening grade of plants up to their processing (Murkovic, Hillebrand, Draxl, Pfannhauser, & Winkler, 1997). The relevant effects of different pretreatments on fatty acid profile may be attributed to the slight alterations in fatty acids. For instance, our results in this section were in agreement with those obtained by Murkovic et al. (1997) who demonstrated that stearic acid is one of the most stable detected fatty acids and is not changed under different conditions. (Kim et al., 2002) also reported that application of microwave in oil extraction did not have considerable effects on fatty acid profile; on the other hand, Ariza-Ortega, Ramírez-Moreno, Ramos-Cassellis, and Díaz-Reyes (2014) revealed that by applying PEF, some alterations on unsaturated fatty acids is observed. The achieved results by Zeng et al. (2010) showed that application of PEF pretreatment in oil extraction of peanut increased stearic and palmitic acids contents but decreased oleic and linoleic acids, that were in agreement with the findings of the present study.

| The effects of microwave-PEF pretreatment on tocopherol content of the extracted oil
The comparison of tocopherol content determined in extracted oil samples revealed that α-tocopherol was the most predominant tocopherol in both the optimized and standard samples followed by γ-and Δ-tocopherols (Table 5 and Figure 7). As observed, application of combined microwave-PEF pretreatment may lead to 2.79% increase in tocopherol content of extracted oils from Niger seeds, particularly in α-and Δ-tocopherols which could be explained in lower reactions between these natural antioxidants and polysaccharides, proteins and different peptides within the seeds (Hamid Bakhshabadi, Mirzaei, Ghodsvali, Jafari, Ziaiifar, et al., 2017). Our results were in agreement with the obtained data of Wiktor et al. (2015) and Oomah, Liang, Godfrey, and Mazza (1998)

| CON CLUS ION
The results of present study demonstrated that with an increase in microwave irradiation time, PEF intensity and screw press rotational speed, chlorophylls content, acidity and PVs, and total phenolic compounds of the extracted oils from Niger seeds were increased, while extraction efficiency showed an enhancement at the beginning and then decreased. The RI of extracted oil was constant at 1.478 and was not affected by pretreatments. The optimized extraction conditions were microwave irradiation time of 156.34 s, PEF intensity of 1.18 kV/cm, and screw press rotational speed of 20 rpm. The characterization of fatty acid profile of the extracted Niger seed oil presented that linoleic acid was the predominant fatty acid, and the application of microwave-PEF pretreatment may lead to a decrease in unsaturated fatty acids and enhancements of saturated fatty acids content. The results of HPLC analysis showed that α-tocopherols had the highest quantity among the tocopherols in Niger seed oil and applying combined microwave-PEF pretreatment may lead to an increase in tocopherol content of the extracted oil. At the end, it may be suggested that application of microwave-PEF pretreatment might be useful as an appropriate pretreatment in oil extraction industries.

ACK N OWLED G EM ENTS
Islamic Azad University (Iran) should be acknowledged for its support.
F I G U R E 7 A typical HPLC chromatogram for determination of tocopherols in Niger seed oil. HPLC, highperformance liquid chromatography

CO N FLI C T O F I NTE R E S T
All authors declare that there is no conflict of interest.

E TH I C A L A PPROVA L
There was no human or animal testing in this study.