Rice bran oil extraction using alcoholic solvents: Physicochemical characterization of oil and protein fraction functionality

Industrial Crops & Products 104 (2017) 133–143

Contents lists available at ScienceDirect

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

Rice bran oil extraction using alcoholic solvents: Physicochemical characterization of oil and protein fraction functionality Maria C. Capellini, Vanessa Giacomini, Maitê S. Cuevas, Christianne E.C. Rodrigues

MARK

⁎

Separation Engineering Laboratory – (LES), Department of Food Engineering (ZEA-FZEA), University of Sao Paulo (USP), P.O. Box 23, Zip Code 13635-900, Pirassununga, SP, Brazil

A R T I C L E I N F O

A B S T R A C T

Keywords: Solid-liquid extraction Ethanol Isopropanol Nitrogen solubility index γ-Oryzanol

Rice bran, an underutilized rice processing by-product, is a promising source for food and biodiesel oil production and can also be used to produce protein for use in human food products. The main objective of this study was to assess the feasibility of replacing hexane, which is traditionally used to extract vegetable oils, with safer solvents, i.e., ethanol and isopropanol, in rice bran oil (RBO) extraction. Thus, the effects of the solvent type on the physicochemical characteristics of the oil and defatted bran products were studied. The results showed that the presence of water in the alcoholic solvents negatively affected the oil extraction; however, using absolute solvents in single-stage batch extractions at 80 °C resulted in oil yields of up to approximately 80%. The solvent water content and process temperature strongly impacted the properties of the protein fraction; the nitrogen solubility index (NSI) decreased from approximately 40% for the absolute solvents to 17 and 15% for the aqueous ethanol and isopropanol, respectively, when the extraction was performed at 80 °C. More of the minor nutraceutical compounds were transferred from the oleaginous matrix to the oil by aqueous ethanol than by hexane, yielding RBO with 1.53% γ-oryzanol and 769 mg/kg tocotrienols. On the other hand, absolute isopropanol exhibited a higher tocopherol extraction capacity; RBO with a tocopherol content of 98.1 mg/kg was obtained with this solvent. Based on these results, short-chain alcohols are promising alternatives to the conventional extraction solvent, because they enable high-quality protein fractions and oils to be obtained and add value to the rice production chain.

1. Introduction Rice is among the most important cereal crops, constituting about 25% of global production of cereal grains, and is consumed as a food staple by more than half the world's population (Adebiyi et al., 2008 Nesterenko et al., 2013). Rice bran, a byproduct of rice processing, constitutes about 8–10% of the grain composition (Nagendra Prasad et al., 2011; Orthoefer, 2005). Rice cultivation and its processing on a large scale, which ranges from 500 to 800 million tons per year (Gunstone, 2005), consequently generates a large amount of rice bran (on average 60 million tons). Due to the increasing worldwide need to provide a food supply and based on its composition, rice bran is considered an inexpensive high quality lipid and protein source for human consumption (Nesterenko et al., 2013). However, this material presents a very active enzyme system composed of lipoxygenase, peroxidase and, especially, lipase, which is endogenously present in the bran or produced as a result of microbial activity, and activated during the grinding process. Under normal conditions, when the bran is not subjected to any process of stabiliza⁎

Corresponding author. E-mail address: [email protected] (C.E.C. Rodrigues).

http://dx.doi.org/10.1016/j.indcrop.2017.04.017 Received 22 June 2016; Received in revised form 15 March 2017; Accepted 9 April 2017 0926-6690/ © 2017 Elsevier B.V. All rights reserved.

tion, it will degrade in approximately 6 h, making it unsuitable for human consumption. For this reason, until recently, rice bran has been underutilized and is only used as a source of protein in animal feed; the oil is used in formulations of soap and glycerin, or as a fertilizer and or boiler fuel (Nagendra Prasad et al., 2011; Orthoefer, 2005; Rajam et al., 2005). The rice bran contains in its composition about 20% lipids (Orthoefer, 2005) and, according to Rajam et al., 2005), among the vegetable oils, rice bran oil (RBO) is one of the most nutritious and healthy. Considered rich in minor components, RBO has become attractive for its unique nutraceutical characteristics and balanced fatty acid composition. In its unsaponifiable material, the more nutritionally important constituents are γ-oryzanol (about 2%) and tocopherols and tocotrienols (about 0.2%), of which about 70% are tocotrienols, a rare component in edible oils. These components have shown remarkable antioxidant activity, benefits to cardiovascular system health, the capacity to reduce serum cholesterol and anticancer properties (Lerma-García et al., 2009; Nagendra Prasad et al., 2011; Orthoefer, 2005 Orthoefer, 2005; Rajam et al., 2005).

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

deacidification. In fact, oil can be deacidified by liquid–liquid extraction with alcoholic solvents due to their partial miscibility, as shown by Rodrigues et al. (2014), who deacidified rice bran oil using ethanol as the solvent. To address safety concerns, Bessa et al. (2017) suggested that the safety measures for handling hexane must be much more rigorous than the corresponding safety measures for ethanol and isopropanol. The flash point of hexane is −22 °C, whereas those of ethanol and isopropanol range from +13 °C to +17 °C and from +14 °C to +18 °C, respectively, depending on the water content (absolute form or azeotropic mixture) (Astbury et al., 2004), meaning the fire hazard of hexane is greater than those of the alcohols. Finally, RBO extraction with alcoholic solvents could be coupled to biodiesel production, because biodiesel is usually produced by the transesterification of vegetable oils or animal fats in the presence of a short-chain alcohol (Zullaikah et al., 2005). In fact, the high free fatty acid (FFA) content due to the presence of active lipase in the bran and the absence of more economical stabilization methods renders most of the produced RBO unsuitable for food applications; therefore, it can be used in biodiesel production (Ju and Vali, 2005). Bessa et al. (2017) suggested that for biodiesel production, oil extraction with an alcohol could be advantageous, because the alcohol-containing extract could be directly used in the subsequent reaction step. This process integration is not possible for hexane-containing extracts; these extracts must undergo evaporation and distillation processes, which involve capital and energy expenses. Therefore, this study aims to replace fossil fuel-based solvents, especially hexane, with environmentally friendly solvents, namely ethanol and isopropanol, in the production of high-quality defatted rice meal and RBO. The effects of the extraction conditions, solvent type (ethanol or isopropanol with 0, 6 or 12 mass % water) and temperature (50 to 80 °C) on the lipid, protein and minor compound extraction yields were studied. Additionally, the physicochemical properties of the oil and protein fraction were evaluated and compared to those of RBO and defatted rice bran samples obtained by an industrial extraction process using hexane. The results of this study show the potential of alternative solvents in the extraction of RBO for use in the pharmaceutical, food and bioenergy industries.

Constituted of more protein than any other part of the rice grain (15% on average), bran is a source of protein suited for use in bakery products, morning cereals and others, in order to increase the nutritional value of these foods (Chandi and Sogi, 2007). According to Chandi and Sogi (2007), the quality of the rice bran protein fraction is only inferior to that of oats, surpassing that of wheat and corn. Moreover, the amount of lysine is higher than in the rice endosperm or other cereals, the digestibility is greater than 90% and, in addition, the amino acid profile is better than that of casein from milk and soy protein isolate regarding the compliance requirements for children aged 2 to 5 years. Also, because it is considered a hypoallergenic source of protein, rice bran may be used in food formulations for children on a restrictive diet (Chandi and Sogi, 2007; Tang et al., 2003; Wang et al., 1999; Zhang et al., 2012 Zhang et al., 2012). In fact, the nutritional quality of a protein ingredient alone does not guarantee its use and better implementation in the development of new products for food purposes. For this reason, it is essential to know the functional properties of the protein source. Among these properties, protein solubility is important because of its influence on other features such as emulsifying, gelling and foaming properties; thus, it is a good indicator of the potential applications of this ingredient (Tang et al., 2003 Vojdani, 1996). To enable the efficient use of bran and the subsequent application of protein in food formulations, the oil must be extracted using appropriate methods. Traditionally, hexane is used as the solvent in vegetable oil extraction. Despite its high stability, lower corrosiveness, and the possibility of obtaining meal with a low residual oil content and good sensory characteristics, hexane comes from a non-renewable source, is characterized by high flammability and toxicity, and also contributes to environmental pollution when not recovered properly (Rodrigues and Oliveira, 2010; Nagendra Prasad et al., 2011; Oliveira et al., 2012a; Tir et al., 2012). Because of the potential risks to human health and the environment associated with hexane use, many research efforts have been focused on finding alternative solvents. Of the alternatives, short-chain alcohols, especially ethanol and isopropanol, are particularly promising, because they have higher operational safety and low toxicity, can be produced from biorenewable sources, can be used to extract a high-quality oil, and improve the sensory and functional characteristics of the defatted meal (Johnson and Lusas, 1983; Chien et al., 1990; Abraham et al., 1993; Hron et al., 1994; Sineiro et al., 1998; Franco et al., 2007; Rodrigues and Oliveira, 2010; Rodrigues et al., 2010; Terigar et al., 2011; Oliveira et al., 2012a; Tir et al., 2012; Sawada et al., 2014; Baümler et al., 2015; Navarro et al., 2016). Due to their high polarity, alcoholic solvents can extract higher amounts of phospholipids and unsaponifiable material from solid matrices than hexane (Nagendra Prasad et al., 2011), thus increasing the nutritional value of the extracted oils. Thus, it is assumed that a more nutritious RBO that is richer in γ-oryzanol and tocols could be obtained by extraction with a short-chain alcohol. Furthermore, alcoholic solvents are considered to be safer than hexane by the US Food and Drug Administration (FDA, 2012). In addition to the possibility of obtaining an oil enriched in minor compounds by extraction with renewable alcoholic solvents, shortchain alcohols are also partially miscible with oils at room temperature, unlike hexane. According to Oliveira et al. (2012b), after hightemperature extraction with short-chain alcohols and subsequent extract cooling, two liquid phases, an alcohol phase and oil-rich phase, are formed, resulting in the partial desolventization of the solvent-oil mixture. Because this desolventization is achieved by simply decreasing the temperature without the need for evaporation or distillation processes, the energy demand of the entire process can be reduced by approximately 25% relative to that of the corresponding process utilizing hexane (Johnson and Lusas, 1983). This behavior of solventoil solutions is also particularly advantageous for subsequent refining processes, most notably free fatty acid removal, a process known as

2. Materials and methods 2.1. Materials Absolute ethanol (purity greater than 99.8%) and absolute isopropanol (purity greater than 99.5%), both purchased from Merck (Darmstadt, Germany), and aqueous solvents with water contents of (6.0 ± 0.3) and (12.0 ± 0.5) mass %, prepared by diluting absolute ethanol and absolute isopropanol, respectively, with deionized water (Millipore, Milli-Q, Bedford, MA, USA), were used as the alcoholic solvents. The water content was controlled by Karl Fischer titration using a KF Titrino apparatus (Metrohm, model 787 KF Titrino, Herisan, Switzerland). These solvents were coded and are hereafter referred to as Et0 (absolute ethanol), IPA0 (absolute isopropanol), Et6 and IPA12, aqueous solvents with approximately 6 and 12 mass % water. Rice bran was industrially stabilized and extruded to form rice bran pellets, which were kindly supplied by Irgovel/Nutracea (Pelotas, RS, Brazil). The rice bran pellets were stored at -20.0 °C to prevent enzymatic degradation until submitted to the extraction process and were used as received, without further pretreatment. Samples of rice bran in other stages of the industrial process (before extrusion, after solvent extraction using hexane and after meal desolventizing) and crude and degummed RBO, extracted with hexane, were also evaluated for comparison purposes. These samples were also kindly supplied by Irgovel/Nutracea.

134

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

2.3.2. Two-stage sequential extraction Two-stage sequential solid-liquid extraction experiments were also performed at 60 and 80 °C using Et0, Et6, IPA0 and IPA12. Specifically, two consecutive extractions were performed under the same experimental conditions (temperature, solid:solvent mass ratio and agitation speed) using the same meal. After the first extraction stage, which was performed according to the procedure described in Section 2.3.1, the solid was weighed and allowed to sit until the extractor equipment was completely cooled to room temperature. Then, the sample was weighed again and placed in the extractor, and the extraction procedure was repeated using fresh solvent. These extraction experiments were conducted at least in duplicate. The solvent was evaporated from the extract phases obtained after the first and second extraction stages in a rotary evaporator (Tecnal, model TE-211, Piracicaba, SP, Brazil) at an absolute pressure of 600 mm Hg and 50 °C. The fatty acid profile and the γ-oryzanol, steryl ferulate, tocopherol, tocotrienol, FFA and phospholipid contents of the resulting RBO samples were determined as described in Section 2.3.3. The raffinate phases obtained from the first and second alcoholic extraction stages performed at 60 and 80 °C were dried in a forced convection oven at 60 °C for 24 h. The protein contents, NSIs and thermal behavior (DSC characterization) of these samples and of the industrial extraction samples (before extrusion, after extraction with hexane and after meal desolventization) were evaluated according to the methodologies described in Section 2.2. In addition to this analysis, the residual oil contents of the raffinate phases obtained after the second extraction stage were also measured using a high-temperature solvent extraction system (ANKOM, model XT10, Macedon, NY, USA; method Am 5-04, AOCS, 1998) with hexane as the solvent at 90 °C for 1 h.

2.2. Raw material characterization Rice bran pellets were submitted to a proximate analysis in terms of the moisture (method Ac 2-41; AOCS, 1998), oil (method Am 2-93; AOCS, 1998), crude protein (method Ba 4e-93; AOCS, 1998) (Leco, model FP-528, St. Joseph, MI, USA), ash (AOAC, 2007), and fiber content (Van Soest et al., 1991). A nitrogen-to-protein conversion factor of 5.95 was used (AOAC, 2007) and the carbohydrate content was estimated by difference. Protein solubility was determined according to the method described by Morr et al. (1985) with minor modifications which were previously described by Sawada et al. (2014) for soybean collets. Rice bran pellets were dispersed in 0.1 M NaCl, then the pH of the dispersion was adjusted to a specific value (2.0, 4.0, 4.5, 5.0, 7.0, or 9.0) with 0.1 N HCl or 0.1 N NaOH, and kept under agitation for 2 h at 25 °C. After that, the dispersions were centrifuged and aliquots of the filtrate were taken for the determination of total nitrogen by a combustion method (method Ba-4f; AOCS, 1998). The nitrogen solubility index (NSI) was calculated according to Eq. (1).

NSI (%) =

NitrogenFiltrate (%) × NaClSolutionwt(g) 100 NitrogenSample (%) × Sample wt(g)

(1)

A TA 2010 Differential Scanning Calorimeter (DSC; TA Instruments, New Castle, DE, USA) was used to evaluate the thermal characteristics of the rice bran protein fraction, following the methodology described by Sawada et al. (2014) with slight modifications. After the sample conditioning period, the pans were hermetically sealed, then heated from 0 to 100 °C at a rate of 10 °C/min. The denaturation enthalpy (ΔH) and denaturation temperature (Td) were obtained and calculated from the thermograms using Universal Analyse V.3.9A software (TA Instruments). At least duplicate measurements were taken.

2.3.3. RBO compositional analysis The extract phases obtained at 60 and 80 °C were desolventized using a rotary evaporator under the experimental conditions described in Section 2.3.2. The fatty acid profiles and the γ-oryzanol, steryl ferulates, tocopherol, tocotrienol, FFA and phospholipid contents of the obtained crude RBO and the crude and degummed RBOs obtained by the industrial process (extracted with hexane) were determined as described below. The fatty acid compositions of the RBO samples were determined by FAME gas chromatography according to the official AOCS methods Ce 2-66 and Ce 1-62 (AOCS, 1998) under the experimental conditions used by Sawada et al. (2014). The iodine value was calculated according to the official AOCS method Cd 1c-85(97) (AOCS, 1998). The FFA content was determined by titration (method 2201, IUPAC, 1979) using an automatic buret (Metrohm, model Dosimat 775, Herisan, Switzerland). The phosphorus content was determined by inductively coupled plasma optical emission spectrometry (ICP-OES; Perkin-Elmer, model Optima 5300 DV, USA) according to the official AOCS method Ca 20-99 (AOCS, 1998). The amount of phosphorus was converted into phospholipid equivalents by multiplying the measured phosphorus content by a conversion factor of 30, according to the official AOCS method Ca 12-55 (AOCS, 1998). The contents of γ-oryzanol and its fractions (campesteryl, cycloartenyl, 24-methylene cycloartanyl, β-sitosteryl and cycloartanyl ferulates) were determined by the method of Cuevas et al. (2017) using an ultra-high-performance liquid chromatography (UPLC) system (Waters ACQUITY Ultra-performance, Milford, MA, USA) equipped with a precolumn and Waters BEH C8 column (2.1 mm × 100 mm, 1.7 μm particle size). The γ-oryzanol content was measured using a diode array detector (PDA) coupled to the UPLC system. Based on the UV spectra, the detection wavelength was set to 327 nm. The mass analysis was performed on a single quadrupole mass detector (SQD) equipped with an electrospray ionization source operating in negative mode (ESI−).

2.3. RBO extraction with alcoholic solvents and compositional analysis 2.3.1. Single-stage batch extraction Solid-liquid extraction was performed using rice bran pellets and various solvents, namely Et0, Et6, IPA0 and IPA12, at different temperatures (50, 60, 70, and 80 °C). The solvent:solid mass ratio was 3:1 (Sawada et al., 2014). The batch extractions were performed in a 500 mL stainless steel isothermal cylindrical reactor, as described by Oliveira et al. (2012a). The rice bran pellets and alcoholic solvent, which were weighed using an analytical balance with a readability of 0.0001 g (Adam, model PW 254, Milton Keynes, UK), were transferred to the extractor device and agitated (175 rpm) for 1 h at the desired temperature. The extraction experiments were performed at least in duplicate. After extraction, the defatted meal (raffinate phase) was weighed using a precision balance with a readability of 0.01 g (Adam, model PGW 1502i, Milton Keynes, UK) and dried in a forced convection oven (Nova Orgânica, model N035/3, Piracicaba, SP, Brazil) at 60 °C for 24 h. The protein content of this phase was determined by AOCS method Ba 4e-93 (AOCS, 1998) using a LECO FP-528 nitrogen determinator (St. Joseph, MI, USA). The extract phase profile was evaluated according to the following methods. The total solvent content was determined by drying at 60 °C for 24 h in a forced convection oven (Nova Orgânica, model N035/3, Piracicaba, SP, Brazil). The water content was determined by Karl Fischer titration (AOCS, 1998). The protein content was measured by AOCS method Ba 4e-93 (LECO, model FP-528, St. Joseph, MI, USA; AOCS, 1998) and is expressed as g protein/100 g dry rice bran. The oil content in the extracted phase was determined by differences and is expressed as the oil yield (g oil/100 g dry rice bran). All the measurements were performed at least in triplicate.

135

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

The total tocopherol and tocotrienol contents; δ-, (γ-+β-) and αtocopherol contents; and δ-, γ- and α-tocotrienol contents were determined according to the method of Bustamante-Rangel et al. (2007) using the UPLC system coupled to a mass spectrometer (Waters ACQUITY Ultra-performance, Milford, MA, USA) and SQD. The data were analyzed using the Mass Lynx software (version 4.1, Waters). 2.4. Validity of the experimental results In order to test the accuracy and repeatability of the obtained experimental data, the results were assessed according to the procedure previously used by Sawada et al. (2014). This procedure permits the mass calculation of the extract phase (MEP) from the mass of the raffinate phase (MRP) and the mass fractions of the system components in the extract phase (wiEP) by least-squares regression. Relative deviations (δ) between the sum (MRP + MEP) and the amount of the initial mixture in the overall composition (M°C) were calculated according to Eq. (2).

δ = (|(M RP + MEP ) − M OC |/ M OC ) × 100

(2) Fig. 1. Yield of extraction (g/100 g of dry bran) as a function of the process temperature. (a) RBO. (b) Protein fraction. (□) Et0; (■) Et6; (Δ) IPA0; (▴) IPA12.

2.5. Statistical analysis

independent of the solvent and hydration degree, there was an increase in the extraction yield of the lipid components. The same behavior was observed in previous studies of RBO extraction with ethanol with varying water contents (0-24 mass %) at different process temperatures (40–60 °C and 60–90 °C) (Rodrigues and Oliveira, 2010; Oliveira et al., 2012a); maximum extraction yields of approximately 74 and 99.9% were obtained with absolute ethanol at 50 °C and aqueous ethanol (6 mass % water) at 82.5 °C, respectively. Terigar et al. (2011) used ethanol to extract RBO under microwave irradiation with a 1:3 bran:solvent ratio in the temperature range of 50 to 73 °C. They also observed an increase in the oil extraction yield with increasing process temperature and achieved a maximum yield of approximately 20 g of oil per 100 g of dry bran. However, they cautioned that the oil yields at the higher temperatures might have been altered by the presence of large amounts of wax, which were also extracted. Previous studies also explored the effects of the alcohol type, solvent water content and extraction process temperature on the oil extraction yield from other oleaginous matrices. Sawada et al. (2014) obtained a maximum soybean oil yield of approximately 83% by extraction with Et0 at 80 °C. Gandhi et al. (2003) used n-propanol, isopropanol and ethanol (in absolute form and azeotropic mixtures) to extract soybean oil and reported yields ranging from 52 to 99.8%, depending on the solvent and extraction time. Baümler et al. (2015) obtained sunflower oil yields of approximately 97% by extraction with aqueous ethanol (5% water) at 50 and 60 °C. In another study, Navarro et al. (2016) extracted corn germ/bran oil at 80 °C and achieved yields of approximately 90% with IPA0, 85% with Et0 and IPA12 and approximately 70% with Et6, which exhibited the lowest performance. Zhang et al. (2002a) reported maximum cottonseed oil yields of 97.6, 98.1 and 98.5% for extraction with isopropanol containing 12, 7 and 3% water, respectively, at 75 °C. Franco et al. (2007) extracted the oil from rosehip seeds with ethanol containing 4 and 8% water at 50 °C and obtained yields of approximately 80 and 70%, respectively. In a later study, Franco et al. (2009) obtained maximum yields of approximately 44% for hazelnut oil extraction with absolute ethanol at 50 °C. In fact, according to Johnson and Lusas (1983), the solubility of vegetable oils in alcoholic solvents depends on both the temperature and solvent water content. Solubility data for cottonseed oil in ethanol and isopropanol (absolute form or azeotropic mixtures) revealed that increasing the temperature results in an increase in the oil solubility in a given solvent. On the other hand, an increase in the solvent water

The mean values of the results from the extraction experiments and analysis were assessed for significance using Duncan's multiple range test (DMRT) with SAS® software (v. 9.2, SAS Institute Inc., Cary, NC, USA). The significance level was established at p ≤ 0.05 (Duncan, 1955). 3. Results and discussion The chemical composition of the rice bran pellets used in the extraction experiments, on a dry basis, can be described as: moisture 7.8 ± 0.2 mass %, crude protein 15.0 ± 0.1 mass %, oil 20.4 ± 0.7 mass %, ash 10.0 ± 0.4 mass %, fiber 22.0 ± 0.7 mass % and nonfibrous carbohydrates 28 ± 2 mass %. In general, these results are in accordance with the findings of Khan et al. (2011), Oliveira et al. (2012a) and Orthoefer (2005). The γ-oryzanol content of the RBO obtained from the raw material using the cold extraction method (Bligh and Dyer, 1959) was 1.58 ± 0.01 mass %. This content is in accordance with the range of values observed by Lerma-García et al. (2009), from 0.9 to 2.9 mass % γ-oryzanol. Regarding the content of FFA, a high level of free acidity was observed, i.e. about 21.1 ± 0.9%, which indicates high enzymatic activity in the bran (Nagendra Prasad et al., 2011; Rajam et al., 2005). The average global mass balance deviations (δ) ranged from 1.60 to 2.90%. The low deviation values calculated, less than 10%, indicate that the experiments were performed with precision and the present data show good repeatability and quality. 3.1. Effects of alcoholic solvent type and temperature in yields of extraction The extraction yields of RBO and protein, for each experimental condition, are presented in Fig. 1 and expressed as the amount of the compound of interest extracted from 100 grams of dry rice bran. In Fig. 1a, it can be seen that the amount of water present in the alcoholic solvent negatively affected the performance of the extraction process by decreasing the extraction of lipid compounds with an increase in hydration, possibly due to a decrease in the solubility of these components. Thus, Et0 and IPA0 showed, on average, greater extraction ability with values of approximately 16 g of oil per 100 g of rice bran at 80 °C. In fact, for IPA12, with 12 mass % water, it was found that this solvent showed a similar oil extraction ability to the absolute solvents at 80 °C. Thus, it can be inferred that the presence of water mainly affects the performance of ethanol. Moreover, it was found that, with an increase in temperature, 136

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

content leads to a decrease in the oil solubility at a given temperature. These authors also reported that at low alcohol concentrations, i.e., high water contents, the oil solubility decreases considerably, and complete miscibility is not achieved, even at the boiling point. This same solubility behavior was also observed for unrefined vegetable oils (soybean, cottonseed, peanut, sesame, corn, linseed, tung, babassu, coconut, olive, palm, safflower and sunflower) in ethanol and isopropanol with different water contents (Rao et al., 1955; Rao and Arnold, 1956a, 1956b, 1957). In the RBO extraction experiments with alcoholic solvents, in a single stage, as well as monitoring the extraction of lipids, the influence of the process temperature and the type of solvent was investigated regarding the extraction of proteins. Fig. 1b presents the protein extraction yield in the extract phase. It was observed that, in general, when an azeotropic solvent was used, there was an increase in the affinity of the protein for the extract phase, i.e., the concentration of the protein in the extract phase increased with an increasing ethanol or isopropanol hydration level. However, Et6, with a lower water composition compared to IPA12, provided greater protein extraction capacity. This increased transfer of protein compounds was also observed when there was an increase in the process temperature, regardless of the solvent used. Sawada et al. (2014) observed the same behavior in the soybean collet protein fraction after extraction with Et6. However, when aqueous solvents were used, the protein concentration of the extract phase decreased with increasing temperature. The protein yields in the extract phase (g/100 g dry rice bran) obtained in this work are similar to those obtained by Sawada et al. (2014) (g/100 g soybean) when Et0 or Et6 was used as the solvent. However, if the percentage of protein extracted relative to the total amount of protein in the starting material is considered, the soybean protein extraction yields ranged from 2.4 to 4.0% for Et0 and from 5.5 to 11.0% for Et6, whereas the corresponding rice bran yields obtained in this work ranged from 8.9 to 18.6% for Et0 and from 15.5 to 20.4% for Et6. The corresponding protein yields obtained by extraction with IPA0 and IPA12 ranged from 5.7 to 9.3% and from 11.3 to 15.0%, respectively. According to Xia et al. (2012), the rice bran protein profile is complex. The major protein components are albumin (37%) and globulin (36%). For soybeans, globulins are the major component, accounting for 87% of the protein fraction (Kinsella, 1979). According to Sgarbieri (1996), albumins are readily soluble in water, whereas globulins are insoluble or very poorly soluble in water. Therefore, it can be inferred that the extraction of the rice bran proteins with aqueous solvents was more successful than that of the soybean proteins because the latter raw material is composed of mostly globulins.

Fig. 2. Nitrogen solubility index (NSI, %) for rice bran pellets as a function of pH.

defatted rice bran protein concentrates in water in the pH range of 3.0 to 11.0. The minimum solubility was observed at pH 4.0. The solubility increased gradually from pH 4.0 to 6.0 and then at slower rate above pH 9.0. According to Bera and Mukherjee (1989), phytates, which are compounds commonly found in rice bran, are soluble in water and can form complexes with cationic proteins that predominate in the acid pH range, leading to insolubilization. Thus, the NSI in this pH range was reduced. The low solubility of the protein fraction of rice bran may also be due to strong aggregation and/or disulfide bonds between proteins (Xia et al., 2012). According to Juliano (1985), the large amount of fiber (12%) present in bran may bind with proteins, making it very difficult to separate them from the other components. Additionally, heat stabilization of the bran often leads to protein denaturation and increases interactions among proteins and carbohydrates and other components, making protein less available and thus hindering further extraction, followed by a decrease in purity due to the precipitation of non-proteinaceous components (Xia et al., 2012). For these reasons, in this work, the solubility indices of the raffinate protein fractions were only measured at pH 9.0, i.e., the pH at which the highest protein solubilization and therefore the lowest experimental uncertainty were achieved. Table 1 presents the NSI values at pH 9.0 for the raffinate phases from the extraction process under all experimental conditions and for samples of rice bran at different stages of the industrial process. First, it is noted that the oil extraction process, independent of the solvent, hydration degree, temperature or the number of extractions, negatively affected the protein solubility of the raffinate phase, i.e., significant differences were observed among the solubility indices of the rice bran pellets and raffinate phases. With regard to the alcoholic solvents used, in general, it was observed that increasing the amount of water in both type of solvents, i.e. ethanol and isopropanol, resulted in a significant decrease in solubility. It can be inferred that the water contained in the solvent exerts negative influence on the NSI. Among the types of solvent used, IPA0 showed the highest protein solubility values for the raffinate phases. The process temperature, as well as the water content in the solvent, equally negatively affected the protein solubility, i.e., higher oil extraction temperatures led to lower levels of protein solubility in the raffinate phases. The declines in solubility were most significant for the raffinate phase samples from the extractions using Et6 and IPA12, due to the increase in the process temperature. Similar effects of the alcoholic solvent water content and extraction temperature on the solubilities of soybean collet and corn germ/bran defatted protein fractions were reported by Sawada et al. (2014) and

3.2. Protein solubility and thermal analysis of the raffinate phases In this work, solid-liquid extraction experiments were also conducted in order to assess the effect of process conditions on the solubility and thermal characteristics of the protein fraction present in the raffinate phases. First, a nitrogen solubility curve, in a solution of 0.1 M NaCl and in a pH range from 2.0 to 9.0, was determined for the rice bran pellets. The protein solubility curve as a function of pH is shown in Fig. 2. It was observed that the minimum solubility value was found at pH 2.0 and the maximum at pH 9.0. In the pH range between these values, it can be seen that until pH 5.0 there was a slight increase in NSI; after this value, the rise was steep. These observations are consistent with those of Bera and Mukherjee (1989), who reported NSI curves for rice bran protein concentrates prepared from full-fat and defatted raw rice brans in 0.1 M NaCl in the pH range of 2.0 to 10.5. The minimum and maximum solubilities were observed at pH 4.5 and in the pH range of 9.0 to 10.5, respectively. Zhang et al. (2012) evaluated the solubility of heat-stabilized 137

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

Table 1 NSI (%), at pH 9.0, for raffinate phases obtained under different process conditions and for samples from the industrial extraction process. Temperature (°C)

50 60 70 80

Stage 1 Stage 2 Stage 1 Stage 2

Rice bran Rice bran after pelleting Rice bran after extraction using hexane De-oiled rice bran after desolventizing

Rice bran pelletsa

Et0

Et6

IPA0

IPA12

38.7 ± 0.7a

38.7 ± 0.7A 28 ± 1B,c 24.9 ± 0.5C,c 25 ± 3C,b 24.6 ± 0.9C,c 20.2 ± 0.9D,c 21 ± 2D,c

38.7 ± 0.7A 25 ± 1B,d 23 ± 2C,c 25.0 ± 0.2B,b 16.2 ± 0.9D,d 17 ± 1D,d 17.2 ± 0.9D,d

38.7 ± 0.7A 32 ± 1B,b 31 ± 2B,b 27 ± 1C,b 28 ± 2C,b 23 ± 2D,b 23.8 ± 0.7D,b

38.7 ± 0.7A 29.1 ± 0.8B,c 22.3 ± 0.5C,c 19.2 ± 0.7D,c 18 ± 1E,d 14.6 ± 0.3F,e 15.3 ± 0.6F,e

56 ± 2A 38.7 ± 0.7B 23 ± 2C 15 ± 1D

These values are means ± standard deviation. Column values followed by different superscript capital letters are significantly different (p < 0.05) by DMRT. Line values followed by different superscript lower cases are significantly different (p < 0.05) by DMRT. a Non-treated pellets. Samples of raffinate phases were compared with dry rice bran pellets (60 °C, 24 h) and not submitted to the oil extraction process.

since there is no definite relation between the value of protein solubility and the amount of oil extracted under any specific condition. Indeed, it can be inferred that the degree of solvent hydration and the process temperature affect protein solubility most strongly. Thus, it can be concluded that the use of the alternative solvent IPA0, followed by Et0, Et6 and IPA12, increasingly affects the solubility of the protein fraction present in the raffinate phases. Samples of rice bran before pelletization, rice bran pellets and rice bran defatted before and after desolventization, from the industrial RBO extraction process which uses hexane as the solvent, were also studied in order to assess the influence of the pelletization step (extrusion), extraction with hexane and the desolventizing step on the quality of the rice bran protein fraction. According to Table 1, it was found that the bran treatment steps decreased the solubility value as they were performed. In fact, it was observed that the NSI decreased, on average, by 35% with the extrusion, extraction and desolventizing steps. Globally, the NSI decreased by 73%, from the full meal to the defatted and desolventized bran. Relating the results from the samples from the industrial process and the results obtained for the raffinate phases from the alcoholic extractions, it can be observed that the NSI for almost all raffinates, except for those extracted with Et6 and IPA12 at 80 °C, are in agreement with the values obtained for the defatted bran before passing through the industrial desolventizing step. In addition, it can be seen that the lowest value of the NSI was presented by the raffinate obtained with IPA12, at 80 °C. This value is consistent with the NSI presented by the rice bran defatted with hexane and submitted to a desolventizing process. Therefore, it can be inferred that the alcohol extraction process at 80 °C using IPA12 as the solvent affects the protein fraction of the rice bran similarly to the industrial desolventization process. Thermal analysis by DSC was carried out for samples of the raffinate phases from the sequential extractions and rice bran samples from the industrial process, with the results shown in Table 2. This table also shows the content of protein (%) on a dry basis for each of these samples. The denaturation temperatures ranged from 68.2 to 71.1 °C, without statistically significant differences. The lowest value was related to the raffinate phase from the first stage of extraction using Et6 at 80 °C, and the highest value was related to the raffinate also originating from the first extraction stage, but with IPA0 at 60 °C. Regarding the denaturation enthalpy, no statistically significant differences were observed among all the raffinate phases from the alcoholic extraction process and rice bran after the pelletization step. Sawada et al. (2014) characterized the thermal properties of defatted soybean meal obtained by ethanolic extraction and found that the process conditions dramatically affected the protein fraction. They observed endothermic peaks in the thermograms of the two major

Navarro et al. (2016), respectively. Sawada et al. (2014) observed that the NSI value decreased from 48 to 13% for extraction with Et6 at 90 °C, whereas Navarro et al. (2016) reported a dramatic decrease in the NSI from 31 to 0% when IPA12 was used to extract oil at 70 °C. The observations in this work are also consistent with the findings of Sessa et al. (1998), who evaluated the impact of the ethanol water content (0–30 mass %) on the extracted raffinate protein solubility. According to these authors, the aqueous alcoholic solvent exhibits both hydrophilic and hydrophobic characteristics, which destabilize the proteins by weakening the hydrophobic interactions between the nonpolar components and by increasing the interactions between the protein molecule and the solvent, i.e., the water molecules. From the results obtained for the oil extraction yield, in Fig. 1a, it was observed that the solvent Et6 showed lower extraction capability in a single contact stage between the raw material and the solvent, throughout the temperature range studied. It can be inferred that the larger amount of residual oil present in the raffinate phase could be responsible for the decreased protein solubility in this case. Indeed, Vojdani (1996) mentions that lipid compounds, through their non-polar aliphatic chains, can interact with the hydrophobic regions of proteins, thereby reducing their solubility. Thus, in order to investigate the actual influence of the extraction temperature, the water content in the solvent and the residual oil content in the raffinate phases on protein solubility, were performed sequential extraction experiments, with two stages. This type of experiment allowed for associating NSI with the solvent type (with or without hydration, resulting in a lower or higher residual oil content) and the process temperature. Table S1 shows the residual oil contents of the raffinate phases from the second stage of the sequential extraction process. The results are consistent with data on oil extraction yield, as shown in Fig. 1a. According to these data, the least powerful solvent for oil extraction was Et6; the raffinate phase corresponding to this experimental condition, after two successive extractions, had, on average, 5.31 ± 0.01% residual oil at a temperature of 60 °C, the highest level of residual oil among all conditions studied. At 80 °C, the raffinate phases from the extraction with Et0 and IPA0 showed the lowest values (0.9 ± 0.2 and 0.8 ± 0.1%, respectively) of residual oil. According to the data shown in Table 1, with special attention to the NSI values related to stages 1 and 2 of the sequential extractions, there were no statistical differences between the types of solvents at 80 °C. At 60 °C, with the exception of Et0, which showed no difference in the NSI values between the stages extraction, for the other solvents, the NSI values were very similar between stages. In this way, the results shown in Table 1, associated with the oil extraction yield results (Fig. 1a and Table S1), allow us to conclude that the NSI is not dependent on the presence of residual oil in the meal,

138

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

Table 2 Denaturation temperatures and enthalpies of the raffinate phases from each stage of sequential extraction process under different conditions and of samples from the industrial extraction process using hexane. Temperature (°C)

Solvent

60

Et0 Et6 IPA0 IPA12

80

Et0 Et6 IPA0 IPA12

Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage Stage

1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2

Rice bran Rice bran after pelleting Rice bran after extraction using hexane De-oiled rice bran after desolventizing

Protein (%, dry basis)

Td (°C)

ΔH (J/g protein)

15.56 ± 0.02EF 16.8 ± 0.2ABC 15.25 ± 0.03F 15.99 ± 0.02CDEF 15.6 ± 0.2EF 16.2 ± 0.2BCDE 15.5 ± 0.3EF 16.8 ± 0.2ABC 16.3 ± 0.4BCDE 17.50 ± 0.09A 15.73 ± 0.05DEF 16.7 ± 0.1ABC 16.3 ± 0.4BCDE 17.40 ± 0.03A 16.54 ± 0.01BC 17.0 ± 0.5AB

70 ± 1ABCD 70.6 ± 0.7ABC 69 ± 1BCD 69.7 ± 0.6ABCD 71.1 ± 0.7A 70.3 ± 0.9ABC 70.3 ± 0.7ABC 69 ± 2CD 69.5 ± 0.7ABCD 69.7 ± 0.8ABCD 68.2 ± 0.9D 68.4 ± 0.9D 70.9 ± 0.4AB 70.1 ± 0.4ABC 69.6 ± 0.7ABCD 68.4 ± 0.6D

5 ± 4A 3.5 ± 0.7A 3 ± 1A 2.7 ± 0.7A 4 ± 3A 2.6 ± 0.6A 4 ± 2A 3 ± 1A 3.8 ± 0.4A na 3 ± 1A 3 ± 1A 4 ± 2A 4.0 ± 0.9A 3 ± 2A 3 ± 2A

70 ± 1ABC 70 ± 2ABCD nd nd

na 5 ± 4A nd nd

13.0 14.1 17.5 16.5

± ± ± ±

0.5H 0.4G 0.4A 0.2BCD

These values are means ± standard deviation. Column values followed by different superscript are significantly different (p < 0.05) by DMRT. na: not available. nd: not detected.

soybean protein fractions. It was reported that as the water content in the ethanol solvent increased (from Et0 to Et6), the fraction 7S (mainly β-conglycinin) peaks disappeared, and the fraction 11S (glycinin fraction) denaturation peaks appeared at slightly lower temperatures. Furthermore, the denaturation enthalpies of the samples extracted with the aqueous solvent were much lower than those of the samples extracted with the absolute solvent. Therefore, it was inferred that increasing the water content of the ethanolic solvent led to the partial denaturation of the defatted soybean meal proteins (Sawada et al., 2014). Thermal data for whole rice bran or bran defatted with organic solvents are rare or even non-existent. The denaturation temperatures and enthalpies for different rice bran protein fractions, namely albumin, globulin, prolamin and glutelin (Adebiyi et al., 2008 Wang et al., 2014), and for protein concentrates, hydrolysates and isolates (Chinma et al., 2014; Tang et al., 2003; Wang et al., 1999) can be found in the literature. In general, the results vary considerably; the reported denaturation temperatures range from 46 to 108 °C, and the denaturation enthalpies range from 0.58 to 8.96 J/g protein. Therefore, the values determined in this study are consistent with the literature data. It is not possible, however, to correlate the thermal properties of the different raffinate phases with the oil extraction temperature or the type of solvent used. In fact, it should be noted that the use of different alcoholic solvents in the temperature range of 60 to 80 °C leads to less protein denaturation than the use of hexane and desolventization industrial processes; no endothermic peaks were observed in the thermograms of the products of these processes, making the calculations of the denaturation temperature and enthalpy impossible.

Table 3 FFA and phospholipid contents of rice bran oils obtained under different extraction conditions. Temperature (°C)

Solvent

FFA (%)

Phospholipids (%)

60

Et0 Et6 IPA0 IPA12 Et0 Et6 IPA0 IPA12

19.9 ± 0.4B 20.11 ± 0.06B 19.4 ± 0.2BC 18.4 ± 0.2CD 20.29 ± 0.04B 21.7 ± 0.4A 18 ± 1D 15 ± 1E 9.1 ± 0.6F 8.22 ± 0.05F

4.7 ± 0.2B 4.5 ± 0.3BC 3.52 ± 0.04D 4.16 ± 0.01C 4.83 ± 0.07B 5.5 ± 0.3A 4.66 ± 0.02B 4.8 ± 0.2B 5.23 ± 0.08F 5.72 ± 0.06F

80

Crude RBOa Degummed RBOa

These values are means ± standard deviation. Column values followed by different superscript are significantly different (p < 0.05) by DMRT. a Samples from the industrial extraction process using hexane.

with isopropanol at 75 °C and found that the FFA content of the oil increased with increasing alcohol concentration. According to these authors, this result might have been due to alcoholysis, in which acyl group exchange occurred between the acylglycerols and the alcohol (Zhang et al., 2002b). Table 3 also gives the FFA levels in the crude and degummed RBOs obtained by industrial extraction with hexane. The FFA contents of these oils were lower than those of the oils obtained by alcoholic extraction. The industrial RBO extraction procedure was performed immediately after thermally stabilizing the bran. This pretreatment is primarily employed to inactivate the bran enzymes. The FFA content results therefore indicate that this step is crucial for maintaining the oil quality, regardless of the extraction solvent used. Additionally, the phospholipid contents of the RBO samples obtained by alcoholic and industrial extraction were determined (Table 3). It should be noted that the extraction of this component from the oleaginous matrix was not significantly affected by the water content of the alcoholic solvent or the temperature. The phospholipid content ranged from 3.5% to 5.5%, which is consistent with the results of Orthoefer (2005). Furthermore, the phospholipid content in the industrial sample was lower than those in the alcoholic extraction samples.

3.3. Physicochemical characterization of RBO Table 3 lists the FFA contents of the RBOs obtained by extraction with alcoholic solvents and hexane. At 60 °C, the type of alcoholic solvent used in the extraction did not significantly affect the amount of FFA transferred from the solid matrix to the oil. In fact, for ethanol, the FFA content of the RBO was not affected by the solvent water content at any temperature. For isopropanol, the FFA content decreased with increasing temperature. At the highest temperature studied, IPA0 extracted the FFAs more effectively than IPA12. Similarly, Zhang et al. (2002b) extracted cottonseed oil from collets 139

Industrial Crops & Products 104 (2017) 133–143

140

701 ± 2E 769 ± 3A 702 ± 2E 735 ± 6C 718 ± 1D 741.6 ± 0.4B 736 ± 5BC 683 ± 4F 551 ± 4H 573 ± 5G 28.1 ± 0.1D 32.6 ± 0.2A 27.6 ± 0.2E 29.94 ± 0.02B 29.1 ± 0.3C 30.2 ± 0.2B 29.2 ± 0.1C 28.0 ± 0.1DE 23.0 ± 0.6G 23.90 ± 0.09F 57.4 ± 0.6G 54.2 ± 0.3H 63.9 ± 0.2E 67.1 ± 0.2C 70.0 ± 0.6B 66.6 ± 0.3CD 76.1 ± 0.3A 65.7 ± 0.3D 60 ± 1F 54.2 ± 0.7H Crude RBOa Degummed RBOa

80

These values are means ± standard deviation. Column values followed by different superscript are significantly different (p < 0.05) by DMRT. a Samples from the industrial extraction process using hexane.

65 ± 1G 96.0 ± 0.4B 86.7 ± 0.7D 91.2 ± 0.6C 67.1 ± 0.8F 91.5 ± 0.4C 98.1 ± 0.7A 85 ± 1E 54 ± 2I 58.2 ± 0.7H 3.76 ± 0.05CD 4.26 ± 0.03A 3.52 ± 0.02E 3.77 ± 0.03C 3.99 ± 0.05B 4.06 ± 0.03B 3.98 ± 0.05B 3.68 ± 0.01D 1.8 ± 0.1G 2.01 ± 0.02F 5.42 ± 0.01A 4.93 ± 0.08B 4.33 ± 0.07C 4.0 ± 0.3D 4.42 ± 0.03C 4.49 ± 0.04C 3.88 ± 0.06D 3.57 ± 0.09E 3.29 ± 0.03F 3.41 ± 0.05EF 13.34 ± 0.08F 13.9 ± 0.1C 13.5 ± 0.1DE 13.13 ± 0.07G 13.48 ± 0.03EF 14.16 ± 0.03B 13.7 ± 0.2D 12.26 ± 0.07H 14.21 ± 0.05B 14.39 ± 0.07A 38.5 ± 0.2D 41 ± 1B 39.2 ± 0.4D 41.2 ± 0.5B 38.9 ± 0.2D 42.1 ± 0.1B 40.3 ± 0.5C 41.3 ± 0.2B 44.1 ± 0.1A 44.3 ± 0.1A 30.0 ± 0.3A 27 ± 1C 28.3 ± 0.5B 27.03 ± 0.06C 30.3 ± 0.2A 26.0 ± 0.2D 27.8 ± 0.5B 28.2 ± 0.1B 22.6 ± 0.2E 22.38 ± 0.06E 12.72 ± 0.02E 12.7 ± 0.1E 14.6 ± 0.3C 14.6 ± 0.2C 12.97 ± 0.05E 13.3 ± 0.2D 14.4 ± 0.2C 14.7 ± 0.2C 15.79 ± 0.02A 15.50 ± 0.07B 0.01E 0.02A 0.01H 0.01G 0.01F 0.01B 0.02D 0.01C 0.01F 0.01F ± ± ± ± ± ± ± ± ± ± Et0 Et6 IPA0 IPA12 Et0 Et6 IPA0 IPA12 60

1.32 1.53 1.14 1.24 1.29 1.46 1.34 1.39 1.28 1.28

γ δ γ+β δ 24-Methylene cycloartanyl Cycloartenyl Campesteryl

Steryl ferulates (%) Oryzanol (%) Solvent Temperature (°C)

Table 4 Nutraceutical compounds and their fractions in RBO obtained under different extraction conditions.

β-Sitosteryl

Cycloartanyl

Tocopherols (mg/kg)

α

Tocotrienols (mg/kg)

α

The phospholipid contents of crude RBOs extracted with different solvents were reported in several studies. For industrial RBO extraction with hexane, the phospholipid content ranges from 1% to 4% (Balachandran et al., 2008; Indira et al., 2000; Rajam et al., 2005; Sengar et al., 2014; Sharif et al., 2013; Van Hoed et al., 2010). When dlimonene was used to extract RBO, the phospholipid content was in the range of 1.31–2.33% (Mamidipally and Liu, 2004; Liu and Mamidipally, 2005). Using chloroform/isopropanol and chloroform/methanol mixtures, RBOs with phospholipid contents of 4.5–4.9 and 4–6.7%, respectively, were obtained (Hemavathy and Prabhakar, 1987; Shin and Godber, 1996; Yoshida et al., 2011; Yoshida et al., 2012), which is consistent with the results of this study. Phospholipids have a polar phosphate group attached to a lipid group via electrostatic forces and hydrogen bonds, which may be transferred to the oil when more polar solvents, able to break these bonds and release them, are used (Brum et al., 2009). In this way, the dielectric constants (a measure of molecular polarity) of the solvents used in this study were calculated, as suggested by Tir et al. (2012) and Wohlfarth (2014) and are presented in Table S2. The differences among the values of the dielectric constants of hexane and the alcoholic solvents may be related to the lower content of phospholipids in the crude RBO from the industrial extraction process. In addition, it can be seen that, among the alcoholic solvents, Et6 had the highest dielectric constant value regardless of the temperature, which could explain the higher capacity of this solvent to extract phospholipids. Table 4 lists the contents of nutraceutical compounds (γ-oryzanol, tocopherols and tocotrienols) in the RBO samples obtained by alcoholic extraction and by industrial extraction with hexane. The tocol and steryl ferulate fractions are also given in this table, and Fig. 3 shows the mass spectral profile of five steryl ferulates in the crude RBO sample extracted with Et0 at 80 °C. For γ-oryzanol, the highest levels were observed in the samples extracted with Et6 at all the temperatures studied. These results are consistent with those of Kamimura et al. (in press), who studied the kinetics of rice bran oil extraction with ethanol containing 0 and 6% water at temperatures of 40–70 °C. Specifically, these researchers found that the highest γ-oryzanol yield was obtained by extraction with Et6 at the highest temperature studied. In general, the RBO samples obtained by alcoholic extraction had a higher γ-oryzanol content and higher steryl ferulate, tocopherol and tocotrienol fractions than those obtained by extraction with hexane. These results are consistent with those of Proctor and Bowen (1996), Hu et al. (1996), Chen and Bergman, 2005Chen and Bergman (2005), Imsanguan et al., 2008 and Zigoneanu et al., 2008. In these previous studies, the levels of minor compounds (γ-oryzanol and tocols) in the extracted crude oil increased when ethanol or isopropanol was used, with or without hexane, instead of hexane. The oryzanol contents of oils extracted with ethanol and isopropanol mixed with hexane were reported to be approximately 2.5–9.4 and 2.5–7.3 mg per g of rice bran, respectively (Hu et al., 1996; Chen and Bergman, 2005; Imsanguan et al., 2008). For the tocols, Hu et al., 1996 and Zigoneanu et al. (2008) measured total tocol (tocopherols and tocotrienols) contents in the ranges of 95.41–171.0 and 56.22–157.3 mg per kg of rice bran in oils extracted with isopropanol and hexane, respectively. The tocopherol and tocotrienol fractions were in the ranges of 5.37–80.9 mg and 4.6–164 mg per kg of rice bran, respectively, for the isopropanol solvent and 3.25–37.1 mg and 3.4–53.2 mg per kg of rice bran, respectively, for hexane (Chen and Bergman, 2005; Zigoneanu et al., 2008). It was reported that the steryl ferulate content and composition are affected by the germination period and environmental factors, but not by the maturity of the rice grains (Miller and Engel, 2006 Ziegler et al., 2016). Furthermore, according to Orthoefer (2005), the stabilization, rice bran storage and oil extraction methods affect the concentrations of the tocol fractions. Of these fractions, γ-tocotrienol is the most stable and can be stored for longer time periods. The rice variety and grinding process used to remove the husks and bran also influence the tocol

19.05 ± 0.08F 19.13 ± 0.07F 19.96 ± 0.03E 20.77 ± 0.09D 22.2 ± 0.1B 21.68 ± 0.06C 22.5 ± 0.2A 19.84 ± 0.07E 13.8 ± 0.2G 12.6 ± 0.1H

M.C. Capellini et al.

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

Fig. 3. Mass spectral profile of steryl ferulates in the crude RBO sample obtained by extraction with Et0 at 80 °C. Selected ion monitoring (SIM) of campesteryl ferulate (575.40 m/z), βsitosteryl ferulate (589.40 m/z), cycloartenyl ferulate (601.40 m/z), cycloartanyl ferulate (603.40 m/z) and 24-methylene cycloartanyl ferulate (615.40 m/z).

temperature also caused a decrease in protein solubility in the defatted meal. It can be concluded that substitution of hexane by ethanol or isopropanol is technically feasible. However, the conditions under which the extraction process is carried out must be carefully evaluated in order to obtain both a protein fraction and oil suitable for food applications and biodiesel production.

content. Table S3 shows the fatty acid composition results of oils from alcoholic extraction at 60 and 80 °C along with crude and degummed RBO from the industrial process with hexane. Based on the statistical analysis, there were no significant differences in the fatty acid contents, independent of the type of solvent or the temperature of extraction. In addition, the fatty acid compositions, iodine values and ratios between the unsaturated and saturated fatty acids were found to be in agreement with the RBO composition determined by Firestone (2006), i.e., the crude oils obtained using the alcoholic solvents in this work had the typical RBO composition. RBOs extracted with a chloroform/isopropanol mixture (Hemavathy and Prabhakar, 1987), supercritical CO2 (Zhao and Shishdcura, 1987), hexane (Zhao and Shishdcura, 1987), and a chloroform/methanol mixture (Yoshida et al., 2011) also had the same fatty acid composition. Terigar et al. (2011) extracted soybean and rice bran oils with ethanol under microwave irradiation and found that the fatty acid compositions of all the samples were typical of the type of oil extracted, regardless of the extraction temperature. Additionally, no significant compositional changes were observed in soybean oils extracted with ethanolic solvents (Sawada et al., 2014), corn germ/ bran oils extracted with ethanol or isopropanol (Navarro et al., 2016), oils extracted from green coffee beans (Freitas and Lago, 2007), sunflower oil extracted with ethanol or petroleum ether (Freitas and Lago, 2007), or oil extracted from Jatropha curcas L. seeds with absolute ethanol or n-hexane (Brossard-González et al., 2010). Therefore, these results show that alternative extraction solvents can be utilized to produce RBO for use as a food ingredient or in biodiesel production.

Acknowledgements The authors would like to thank EXTRAE/FEA/UNICAMP and LTA/ ZEA/FZEA/USP for several analyses, and Irgovel/Nutracea for the rice bran and oils donation. They also wish to acknowledge FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo–2010/ 13285-5, 2013/13339-6, 2014/09446-4, 2014/21252-0), CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico–303797/2016-9) for their financial support. 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.indcrop.2017.04.017. References Abraham, G., Hron, R.J., Kuk, M.S., Wan, P.J., 1993. Water accumulation in the alcohol extraction of cottonseed. J. Am. Oil Chem. Soc. 70, 207–208. Adebiyi, A.P., Adebiyi, A.O., Hasegawa, Y., Ogawa, T., Muramoto, K., 2008. Isolation and characterization of protein fractions from deoiled rice bran. Eur. Food Res. Technol. 228, 391–401. AOAC, 2007. Official Methods of Analysis, 18th ed. Association of Official Analytical Chemists, Washington, D.C. AOCS, 1998. Official Methods and Recommended Practices of the AOCS, 5th ed. American Oil Chemists Society, Champaign. Astbury, G.R., Bugand-Bugandet, J., Grollet, E., Stell, K.M., 2004. Flash points of aqueous solutions of flammable solvents. Symp. Ser. 505–522. Balachandran, C., Mayamol, P.N., Thomas, S., Sukumar, D., Sundaresan, A., Arumughan, C., 2008. An ecofriendly approach to process rice bran for high quality rice bran oil using supercritical carbon dioxide for nutraceutical applications. Bioresour. Technol. 99, 2905–2912. Baümler, E.R., Carrín, M.E., Carelli, A.A., 2015. Extraction of sunflower oil using ethanol as solvent. J. Food Eng. 178, 190–197. http://dx.doi.org/10.1016/j.jfoodeng.2016. 01.020. Bera, M.B., Mukherjee, R.K., 1989. Solubility, emulsifying, and foaming properties of rice bran protein concentrates. J. Food Sci. 54, 142–145. Bessa, L.C.B.A., Ferreira, M.C., Rodrigues, C.E.C., Batista, E.A.C., Meirelles, A.J.A., 2017. Simulation and process design of continuous countercurrent ethanolic extraction of

4. Conclusions From the results obtained in the solid-liquid extraction experiments of RBO using ethanol and isopropanol as alternative solvents, it was noted that, in general, the variable of the greatest impact was solvent hydration. A higher amount of water in the alcohol caused a decrease in oil extraction yield. On the other hand, increasing process temperature favored lipid compound extraction. In addition, regardless the process conditions, fatty acid composition remained typical for RBO, and the γoryzanol and tocols main fractions in the oils increased with alcoholic extraction. Regarding the protein fraction, increasing hydration of the alcoholic solvents and process temperature led to an increase in protein fraction transfer from the bran to the extract phase. Solvent hydration and 141

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

Navarro, S.L.B., Capellini, M.C., Aracava, K.K., Rodrigues, C.E.C., 2016. Corn germ-bran oils extracted with alcoholic solvents: Extraction yield, oil composition and evaluation of protein solubility of defatted meal. Food Bioprod. Process. 100, 185–194. Nesterenko, A., Alric, I., Silvestre, F., Durrieu, V., 2013. Vegetable proteins in microencapsulation: a review of recent interventions and their effectiveness. Ind. Crops Prod. 42, 469–479. Oliveira, R., Oliveira, V., Aracava, K.K., Rodrigues, C.E.C., 2012a. Effects of the extraction conditions on the yield and composition of rice bran oil extracted with ethanol—A response surface approach. Food Bioprod. Process 90, 22–31. Oliveira, C.M., Garavazo, B.R., Rodrigues, C.E.C., 2012b. Liquid-liquid equilibria for systems composed of rice bran oil and alcohol-rich solvents: Application to extraction and deacidification of oil. J. Food Eng. 110, 418–427. Orthoefer, F.T., 2005. Rice Bran Oil. In: Shahidi, F. (Ed.), Bailey's Industrial Oil and Fat Products. John Wiley and Sons, New Jersey, pp. 465–489. Proctor, A., Bowen, D.J., 1996. Ambient-temperature extraction of rice bran oil with hexane and isopropanol. J. Am. Oil Chem. Soc. 73, 811–813. Rajam, L., Soban Kumar, D.R., Sundaresan, A., Arumughan, C., 2005. A novel process for physically refining rice bran oil through simultaneous degumming and dewaxing. J. Am. Oil Chem. Soc. 82, 213–220. Rao, R.K., Arnold, L.K., 1956a. Alcoholic extraction of vegetable oils. II. Solubilities of corn, linseed, and tung oils in aqueous ethanol. J. Am. Oil Chem. Soc. 33, 82–84. Rao, R.K., Arnold, L.K., 1956b. Alcoholic extraction of vegetable oils. III. Solubilities of babassu, coconut, olive, palm, rapeseed, and sunflower seed oils in aqueous ethanol. J. Am. Oil Chem. Soc 33, 389–391. Rao, R.K., Arnold, L.K., 1957. Alcoholic extraction of vegetable oils, Part IV. Solubilities of vegetable oils in aqueous 2-propanol. J. Am. Oil Chem. Soc. 34, 401–404. Rao, R.K., Krishna, M.G., Zaheer, S.H., Arnold, L.K., 1955. Alcoholic extraction of vegetable oils. I. Solubilities of cottonseed, peanut, sesame, and soybean oils in aqueous ethanol. J. Am. Oil Chem. Soc. Rodrigues, C.E.C., Aracava, K.K., Abreu, F.N., 2010. Thermodynamic and statistical analysis of soybean oil extraction process using renewable solvent. Int. J. Food Sci. Technol. 45, 2407–2414. Rodrigues, C.E.C., Gonçalves, C.B., Marcon, E.C., Batista, E.A.C., Meirelles, A.J.A., 2014. Deacidification of rice bran oil by liquid–liquid extraction using a renewable solvent. Sep. Purif. Technol. 132, 84–92. Rodrigues, C.E.C., Oliveira, R., 2010. Response surface methodology applied to the analysis of rice bran oil extraction process with ethanol. Int. J. Food Sci. Technol. 45, 813–820. Sawada, M.M., Venâncio, L.L., Toda, T.A., Rodrigues, C.E.C., 2014. Effects of different alcoholic extraction conditions on soybean oil yield, fatty acid composition and protein solubility of defatted meal. Food Res. Int. 62, 662–670. Sengar, G., Kaushal, P., Sharma, H.K., Kaur, M., 2014. Degumming of rice bran oil. Rev. Chem. Eng. 30, 183–198. Sessa, D.J., Nelsen, T.C., Snyder, J.M., 1998. Effect of salts on soy storage proteins defatted with supercritical CO2 and alcohols. J. Am. Oil Chem. Soc. 75, 911–916. Sgarbieri, V.C., 1996. Proteínas em alimentos proteicos: propriedades, degradações, modificações. Livraria Varela, São Paulo. Sharif, M.K., Butt, M.S., Anjum, F.M., Khan, S.H., 2013. Rice bran: a novel functional ingredient. Crit. Rev. Food Sci. Nutr. 37-41, 608586. Shin, T.-S., Godber, J.S., 1996. Changes of endogenous antioxidants and fatty acid composition in irradiated rice bran during storage. J. Agric. Food Chem. 44, 567–573. Sineiro, J., Domínguez, H., Núñez, M.J., Lema, J.M., 1998. Ethanolic extraction of sunflower oil in a pulsing extractor. J. Am. Oil Chem. Soc. 75, 753–754. Tang, S., Hettiarachchy, N.S., Horax, R., Eswaranandam, S., 2003. Physicochemical properties and functionality of rice bran protein hydrolyzate prepared from heatstabilized defatted rice bran with the aid of enzymes. J. Food Sci. 68, 152–157. Terigar, B.G., Balasubramanian, S., Sabliov, C.M., Lima, M., Boldor, D., 2011. Soybean and rice bran oil extraction in a continuous microwave system: From laboratory- to pilot-scale. J. Food Eng. 104, 208–217. Tir, R., Dutta, P.C., Badjah-Hadj-Ahmed, A.Y., 2012. Effect of the extraction solvent polarity on the sesame seeds oil composition. Eur. J. Lipid Sci. Technol. 114, 1427–1438. Van Hoed, V., Ayala, J.V., Czarnowska, M., De Greyt, W., Verhé, R., 2010. Optimization of physical refining to produce rice bran oil with light color and high oryzanol content. JAOCS, J. Am. Oil Chem. Soc. 87, 1227–1234. Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583–3597. Vojdani, F., 1996. Solubility. In: Hall, G.M. (Ed.), Methods of Testing Protein Functionality. Chapman & Hall, London, pp. 11–61. Wang, M., Hettiarachchy, N.S., Qi, M., Burks, W., Siebenmorgen, T., 1999. Preparation and functional properties of rice bran protein isolate. J. Agric. Food Chem. 47, 411–416. Wang, C., Li, D., Xu, F., Hao, T., Zhang, M., 2014. Comparison of two methods for the extraction of fractionated rice bran protein. J. Chem. 1-10. http://dx.doi.org/10. 1155/2014/546345. Wohlfarth, C., 2014. Permittivity (dielectric constant) of liquids. In: Haynes, W.M. (Ed.), CRC Handbook of Chemistry and Physics. CRC Press, Boca Raton, FL, pp. 187–191. Xia, N., Wang, J., Yang, X., Yin, S., Qi, J., Hu, L., Zhou, X., 2012. Preparation and characterization of protein from heat-stabilized rice bran using hydrothermal cooking combined with amylase pretreatment. J. Food Eng. 110, 95–101. Yoshida, H., Kuriyama, I., Tomiyama-Sakamoto, Y., Mizushina, Y., 2012. Profiles of lipid components, fatty acids and triacylglycerol molecular species in lipids of rice bran cultivars. Food Sci. Technol. Res. 18, 219–226. Yoshida, H., Tanigawa, T., Yoshida, N., Kuriyama, I., Tomiyama, Y., Mizushina, Y., 2011.

rice bran oil. J. Food Eng. 202, 99–113. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. Brossard-González, C.O., Ferrari, R.A., Pighinelli, L., Park, J., 2010. Evaluación preliminar del etanol anhidro como solvente en la extracción de aceite de semillas de jatrofa (Jatropha curcas L.). Grasas y Aceites 61, 295–302. Brum, A.A.S., De Arruda, L.F., Regitano-d’Arce, M.A.B., 2009. Métodos de extração e qualidade da fração lipídica de matérias-primas de origem vegetal e animal. Quim. Nova 32, 849–854. Bustamante-Rangel, M., Delgado-Zamarreño, M.M., Sánchez-Pérez, A., CarabiasMartínez, R., 2007. Determination of tocopherols and tocotrienols in cereals by pressurized liquid extraction–liquid chromatography–mass spectrometry. Anal. Chim. Acta 587, 216–221. Chandi, G.K., Sogi, D.S., 2007. Functional properties of rice bran protein concentrates. J. Food Eng. 79, 592–597. Chen, M.H., Bergman, C.J., 2005. A rapid procedure for analysing rice bran tocopherol, tocotrienol and γ-oryzanol contents. J. Food Compos. Anal. 18, 319–331. Chien, J.T., Hoff, J.E., Lee, M.J., Lin, H.M., Chen, Y.J., Chen, L.F., 1990. Oil extraction of dried ground corn with ethanol. Chem. Eng. J. 43. Chinma, C.E., Ilowefah, M., Shammugasamy, B., Ramakrishnan, Y., Muhammad, K., 2014. Chemical, antioxidant, functional and thermal properties of rice bran proteins after yeast and natural fermentations. Int. J. Food Sci. Technol. 49, 2204–2213. Cuevas, M.S., Souza, P.T., Rodrigues, C.E.C., Meirelles, A.J.A., 2017. Quantification and determination of composition of steryl ferulates in refined rice bran oils using an UPLC-MS method. J. Am. Oil Chem. Soc. 94, 375–385. Duncan, D., 1955. Multiple range and multiple F tests. Biometrics 11, 1–42. FDA, 2012. US Food and Drug Administration - Guidance for Industry - Q3C Tables and Lists . Firestone, D., 2006. Physical and Chemical Characteristics of Oils, Fats, and Waxes, 2nd ed. AOCS Press, Washington, DC. Franco, D., Sineiro, J., Núñez, M.J., 2009. Analysis of variables and modeling of gevuina avellana oil extraction with ethanol near azeotrope conditions. J. Food Process Eng. 32, 664–681. Franco, D., Sineiro, J., Pinelo, M., Núñez, M.J., 2007. Ethanolic extraction of Rosa rubiginosa soluble substances: oil solubility equilibria and kinetic studies. J. Food Eng. 79, 150–157. Freitas, S.P., Lago, R.C.A., 2007. Equilibrium data for the extraction of coffee and sunflower oils with ethanol. Brazilian J. Food Technol. 10, 220–224. Gandhi, A.P., Joshi, K.C., Jha, K., Parihar, V.S., Srivastav, D.C., Raghunadh, P., Kawalkar, J., Jain, S.K., Tripathi, R.N., 2003. Studies on alternative solvents for the extraction of oil-l soybean. Int. J. Food Sci. Technol. 38, 369–375. Gunstone, F.D., 2005. Vegetable Oils. In: Shahidi, F. (Ed.), Bailey's Industrial Oil and Fat Products. John Wiley and Sons, New Jersey, pp. 213–267. Hemavathy, J., Prabhakar, J.V., 1987. Lipid composition of rice (Oryza sativa L.) bran. J. Am. Oil Chem. Soc. 64, 1016–1019. Hron, R.J., Kuk, M.S., Abraham, G., Wan, P.J., 1994. Ethanol extraction of oil, gossypol and aflatoxin from cottonseed. J. Am. Oil Chem. Soc. 71, 417–421. Hu, W., Henry, J., Shin, T., Godber, I.S., 1996. Comparison of isopropanol and hexane for extraction of vitamin E and oryzanols from stabilized rice bran. J. Am. Oil Chem. Soc. 73, 1653–1656. Imsanguan, P., Roaysubtawee, A., Borirak, R., Pongamphai, S., Douglas, S., Douglas, P.L., 2008. Extraction of α-tocopherol and γ-oryzanol from rice bran. LWT - Food Sci. Technol. 41, 1417–1424. Indira, T.N., Hemavathy, J., Khatoon, S., Gopala Krishna, A.G., Bhattacharya, S., 2000. Water degumming of rice bran oil: a response surface approach. J. Food Eng. 43, 83–90. IUPAC, 1979. Standard methods for the analysis of oils, fats and derivatives. In: Paquot, C. (Ed.), 6th ed. Pergamon Press, Oxford (part 1, sections I and II). Johnson, L.A., Lusas, E.W., 1983. Comparison of alternative solvents for oils extraction. J. Am. Oil Chem. Soc. 60, 229–242. Ju, Y.H., Vali, S.R., 2005. Rice bran oil as a potential resource for biodiesel: A review. J. Sci. Ind. Res. (India) 64, 866–882. Juliano, B.O., 1985. Rice: Chemistry and Technology, 2nd ed. American Association of Cereal Chemists, St. Paul, Minn., USA. Kamimura, J.A.A.M., Aracava, K.K., Rodrigues, C.E.C., 2017;al., in press. Experimental data and modeling of rice bran oil extraction kinetics using ethanol as solvent. Sep. Sci. Tech.. http://dx.doi.org/10.1080/01496395.2017.1307224. (In press). Khan, S.H., Butt, M.S., Sharif, M.K., 2011. Biological quality and safety assessment of rice bran protein isolates. Int. J. Food Sci. Technol. 46, 2366–2372. Kinsella, J., 1979. Functional Properties of soy protein. J. Am. Oil Chem. Soc. 56, 242–258. Lerma-García, M.J., Herrero-Martínez, J.M., Simó-Alfonso, E.F., Mendonça, C.R.B., Ramis-Ramos, G., 2009. Composition, industrial processing and applications of rice bran (-oryzanol. Food Chem. 115, 389–404. Liu, S.X., Mamidipally, P.K., 2005. Quality comparison of rice bran oil extracted with dlimonene and hexane. Cereal Chem. 82, 209–215. Mamidipally, P.K., Liu, S.X., 2004. First approach on rice bran oil extraction using limonene. Eur. J. Lipid Sci. Technol. 106, 122–125. Miller, A., Engel, K.-H., 2006. Content of gamma-oryzanol and composition of steryl ferulates in brown rice (Oryza sativa L.) of European origin. J. Agric. Food Chem. 54, 8127–8133. Morr, C.V., German, B., Kinsella, J.E., Regenstein, J.M., Buren, J.P.V., Kilara, A., Lewis, B.A., Mangino, M.E., 1985. A collaborative study to develop a standardized food protein solubility procedure. J. Food Sci. 50, 1715–1718. Nagendra Prasad, M.N., Sanjay, K.R., Shravya Khatokar, M., Vismaya, M.N., Nanjunda Swamy, S., 2011. Health Benefits of Rice Bran - A Review. J. Nutr. Food Sci. 01, 1–7.

142

Industrial Crops & Products 104 (2017) 133–143

M.C. Capellini et al.

carbon dioxide. Agric. Biol. Chem. 51, 1773–1777. Ziegler, J.U., Flockerzie, M., Longin, C.F.H., Würschum, T., Carle, R., Schweiggert, R.M., 2016. Development of lipophilic antioxidants and chloroplasts during the sprouting of diverse Triticum spp. J. Agric. Food Chem. 64, 913–922. Zigoneanu, I.G., Williams, L., Xu, Z., Sabliov, C.M., 2008. Determination of antioxidant components in rice bran oil extracted by microwave-assisted method. Bioresour. Technol. 99, 4910–4918. Zullaikah, S., Lai, C.-C., Vali, S.R., Ju, Y.-H., 2005. A two-step acid-catalyzed process for the production of biodiesel from rice bran oil. Bioresour. Technol. 96, 1889–1896.

Lipid components, fatty acid distributions of triacylglycerols and phospholipids in rice brans. Food Chem. 129, 479–484. Zhang, F., Rhee, K.C., Koseoglu, S.S., 2002a. Isopropyl alcohol extraction of cottonseed collets: efficiency and performance. J. Food Lipids 9, 147–160. Zhang, F., Rhee, K.C., Koseoglu, S.S., 2002b. Isopropyl alcohol extraction of cottonseed collets: crude oil composition. J. Food Lipids 9, 161–173. Zhang, H.-J., Zhang, H., Wang, L., Guo, X.-N., 2012. Preparation and functional properties of rice bran proteins from heat-stabilized defatted rice bran. Food Res. Int. 47, 359–363. Zhao, W., Shishdcura, A., 1987. Fractional extraction of rice bran oil with supercritical

143