Available online at www.sciencedirect.com
J. of Supercritical Fluids 45 (2008) 322–331
Supercritical carbon dioxide extraction and deacidification of rice bran oil Chao-Rui Chen a , Chih-Hung Wang a , Ling-Ya Wang a , Zih-Hao Hong a , Shuo-Hsiu Chen b , Wai-Jane Ho c , Chieh-Ming J. Chang a,∗ a
Department of Chemical Engineering, National Chung Hsing University, 250, Kuo-Kuang Road, Taichung 402, Taiwan, ROC b Department of Research and Development, Taiwan Supercritical Technology Company, Ltd., 346 Yuan-Tsao Road, Changhua 502, Taiwan, ROC c Department of Bioresources, Dayeh University, 112, Shanjiao Road, Changhua 515, Taiwan, ROC Received 7 November 2007; received in revised form 20 December 2007; accepted 4 January 2008
Abstract This study examined pilot-scale extraction and lab-scale deacidification of rice bran oil by using supercritical carbon dioxide (SC-CO2 ). Two purest gamma-oryzanols (␥-oryzanols) (>98 wt%) were initially obtained by preparative reverse-phase high-performance liquid chromatography. Supercritical carbon dioxide extraction at 300 bar and 313 K from 1.03 kg powdered rice bran indicated a total yield of oil of 15.7% with a free fatty acids content of 3.75%, obtained from 20.5 kg of carbon dioxide in 8 h. In the SC-CO2 deacidification, pressure ranged from 200 bar to 300 bar, temperature ranged from 343 K to 363 K and consumption of carbon dioxide ranged from 900 g to 2700 g: the efficiency of removal of free fatty acids from 13 g extracted oil in deacidification at 250 bar and 353 K reached 97.8% using 2700 g of carbon dioxide. Finally, three-factor center composite scheme of response surface methodology was employed in designing a SC-CO2 deacidification system, which demonstrated that the pressure and consumption of carbon dioxide are significant in retaining triglycerides and in removing free fatty acids from rice bran oil. © 2008 Elsevier B.V. All rights reserved. Keywords: Supercritical carbon dioxide; Rice bran; Rice bran oil; Gamma-oryzanols; Extraction distillation; Deacidification
1. Introduction Rice bran consists of 11–15% proteins, 34–62% carbohydrates, 7–11% crude fibers, 7–10% ashes and 15–20% lipids, which are a by-product of the rice-refining process [1]. This material comprises nutritional and non-nutritional compounds that benefit humans. Rice bran oil contains 95.6% saponifiable lipids, including glycolipid and phospholipids; and 4.2% unsaponifiable lipids, including tocopherols, tocotrienols, gamma-oryzanol (␥-oryzanols), sterols and carotenoids [2]. The saponifiable lipids are mainly triglycerides. However, these triglycerides are easily hydrolyzed by lipase to form free fatty acids. The ␥-oryzanols content in the rice bran oil is approximately 1.8–3%, according to the experimental data of Hu et al. [3]. Xu and Godber [4] adopted a low-pressure normal phase-silica column to obtain oryzanols-containing fractions, which were
∗
Corresponding author. Tel.: +886 4 2285 2592; fax: +886 4 2286 0231. E-mail address:
[email protected] (C.-M.J. Chang).
0896-8446/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.supflu.2008.01.006
further partitioned using a preparative normal phase HPLC column. They have reported that 10 molecular structures of ␥oryzanols were identified using a reverse-phase HPLC column that is coupled with a GC–mass chromatograph. Cycloartenyl ferulate, 24-methylenecycloartanyl ferulate and campesteryl ferulate are evidently major ␥-oryzanols in rice bran. ␥-Oryzanols have certain biological and physiological abilities, such as antioxidation [5], anti-blood cholesterol [6] and anti-carcinogenic [7–10]. The LC–MS/MS method has been frequently adopted in elucidating the structures of ␥-oryzanols [11]. St¨oggl et al. identified and quantified tocopherols, ␥-oryzanols and cartenoids in rice bran [12]. Aguilar-Garcia et al. performed a biological study of the correlation between the quantities of ␥-oryzanols and anti-oxidant capacity [13]. The edible rice bran oil only allows a maximal acid value of 0.2, in general equivalent to 0.1 wt% free fatty acid. The Indian regulation for the refined rice bran oil allows a maximal acid value of 0.5, which is equivalent to 0.25 wt% free fatty acid [14]. Triglycerides in the wasted rice bran oil had been processed into biodiesel by acid-catalyzed trans-esterification [15]. In recent years, supercritical fluid extractions of powdered rice bran have
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
shown that the odor and the flavor of extracted oil are superior to that obtained by Soxhlet n-hexane extraction [16]. Deacidification of rice bran oil using supercritical carbon dioxide (SC-CO2 ) has been recognized as a very environmentally friendly process [17]. Kim et al. investigated the time-related mass transfer kinetics of oil components that migrated between solid phase (rice bran) and fluid phase (carbon dioxide) in supercritical carbon dioxide extraction [18]. Chang et al. studied supercritical carbon dioxide extraction kinetics and high-pressure vapor–liquid phase equilibrium measurements between oil compounds and carbon dioxide [19,20]. Several multi-stage supercritical fluid deacidifications of rice bran oil have been conducted to remove free fatty acids or triglycerides from raffinate oil using a packed column at a middle-high pressure at 333–353 K to increase concentrations of ␥-oryzanols and phytosterols in the oil [21–25]. Local chemical agencies could supply no pure standard of ␥oryzanols other than cycloartenyl ferulate. Hence, the objective of this work is to obtain pure 24-methylenecycloartanyl ferulate and campesteryl ferulate from a mixed ␥-oryzanols standard using a semi-preparative HPLC method. Then, the supercritical carbon dioxide extraction of rice bran oil from powdered rice bran, followed by SC-CO2 deacidification using a response surface methodology were studied. 2. Materials and methods 2.1. Reagents and materials A local Da-quan rice bran producer (Taichung, Taiwan) donated two 10 kg respective batches of fresh rice bran powder, which was stored in a cooler at 269 K and de-activated in an oven at 353 K within 90 s before use. The rice bran powder was smaller than 2 mm by sieving through a 10 mesh international type stainless steel screen. The water content of this rice bran powder was 9.5%, as determined using a moisture analyzer (A&D, AD4714A, Japan). De-ionized water was obtained using a Milli-Q reverse osmosis purification system. Analytical grade reagents – 99.95% carbon dioxide (Toyo gas, Taiwan), 99.7% nitrogen (Toyo gas, Taiwan), 99.8% hydrogen (Toyo gas, Taiwan), 99% air (Toyo gas, Taiwan), 99.9% ethyl acetate (Mallinckrodt, USA), 99.9% n-hexane (Mallinckrodt, USA), 99.9% methanol (Mallinckrodt, USA), 99.9% dichloromethane (Mallinckrodt, USA), 99.5% ethyl ether (Mallinckrodt, USA), 99.5% isopropanol (Mallinckrodt, USA), 99.9% acetonitrile (J.T. Baker, USA), 99.8% acetic acid (Merck, Germany) and 95% ethanol (Taiwan Sugar Co. Ltd., Taiwan) – were purchased from a chemicals supplier and used without further treatment. Authentic standards, including 99% cycloartenyl ferulate (Wako, Japan), 95% mixed standard ␥-oryzanols (Wako, Japan) and a fatty acid methyl ester standard kit of FAME-6 (Supelco, USA), were used herein only for quantification. 2.2. Isolation of γ-oryzanols using HPLC chromatography A semi-preparative HPLC system has one C18 column (YMC, 250 mm × 10 mm I.D.) that is connected to a UV detector (785A, Perkin-Elmer) via a high-pressure pump (410,
323
Fig. 1. Information block diagram of preparing purest ␥-oryzanols using semipreparative HPLC.
Perkin-Elmer). Fig. 1 shows an information block diagram of preparing purest ␥-oryzanols using this HPLC method. The 1.5 mL (500 ppm) mixed standard of ␥-oryzanols dissolved in methanol was injected into a 5 mL loop and this sample was partitioned using a mixed solvent of acetonitrile, dichrolomethane and acetic acid (90:6:4) at a flow rate of 5 mL/min. The eluates were detected at a wavelength of 330 nm and four isolated compounds were collected with different retention times. Ten milligrams of both 24-methylenecycloartanyl ferulate and campesteryl ferulate were obtained. Their chemical structures were identified using a 400 MHz 1 H NMR spectrophotometer. 2.3. Quantification of γ-oryzanols, free fatty acids and triglycerides HPLC quantifications of four ␥-oryzanols (cycloartenyl ferulate, 24-methylenecycloartanyl ferulate, campesteryl ferulate and sitosteryl ferulate) and three free fatty acids (oleic acid, linoleic acid and linolenic acid) were performed using a reversephase analytical column (YMC RP-18, 5 m, 250 mm × 4.6 mm I.D., Japan). The column was linked to a UV/Vis detector (L4200H, Hitachi, Japan) via an intelligent pump (L-7100, Hitachi, Japan) and software was used to control the interface (D-7000, Hitachi, Japan). The column temperature was maintained at 313 K using a column oven (TCM-004657, Waters-Millipore, USA). The UV absorption of the four ␥-oryzanols and the three free fatty acid samples was detected at a wavelength of 330 nm and 210 nm. The injection volume of the sample was 20 L. One mobile phase of the mixed solvent of 90% acetonitrile, 6% dichrolomethane and 4% acetic acid was used in the analysis of ␥-oryzanols. The other mobile phase of the mixed solvent with 85% acetonitrile, 5% methanol and 10% de-ionized water with
324
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
Fig. 2. Schematic flow diagram of SC-CO2 extraction of rice bran oil. (1) CO2 cylinder, (2) pump, (3) chiller, (4-1–4-2) CO2 cleaner, (5) mixer, (6-1–6-2) pre-heater, (7) extractor, (8) separator, (9-1–9-2) circulator, (10) wet gas meter, (11-1–11-4) gauge, (12-1–12-6) metering valve, and (13-1–13-3) vent valve.
1% acetic acid was used to analyze the free fatty acids. The Rsquare correlation coefficients of the calibration curves of both ␥-oryzanols and free fatty acids exceeded 0.99. The limits of detection of ␥-oryzanols and free fatty acids were 0.125 g/g and 0.545 g/g, respectively.
GC quantification of seven triglycerides was performed using a non-polar capillary column (007-CW, Quadrex, USA) in a gas chromatographer (GC-14B, Shimadzu, Japan). The column temperature was initially set to 443 K, and programmed to increase to 488 K at 5 K/min, then to 496 K at 2 K/min, and finally to
Fig. 3. Schematic flow diagram of SC-CO2 deacidification of rice bran oil. (1) CO2 cylinder, (2) CO2 cleanup column, (3) constant temperature circulator, (4) high pressure pump, (5) pressure gauge, (6) hot plate, (7) oil bath, (8) extraction vessel, (9-1–9-2) backpressure regulator, (10) micro-metering valve, (11-1–11-4) metering valve, (12) separator, (13) collection flask, (14) wet gas meter, (15) temperature display, (16) temperature controller, and (17) thermocouple.
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
325
Fig. 4. Four hundred megahertz 1 H NMR spectra of (a) purest 99% cycloartenyl ferulate; (b) purest 98.5% 24-methylenecycloartanyl ferulate; (c) purest 98% campesteryl ferulate.
503 K at 1 K/min. The split ratio was 3.4:1. Injection and flame ionization detector temperatures were set to 553 K. The limit of detection of fatty acid methyl esters was 0.148 g/g. 2.4. Classical solvents extraction In Soxhlet extractions, four 15 g samples of rice bran powder were individually loaded into a 250 mL reflux Soxhlet extractor and extracted using 300 mL n-hexane for 4 h, respectively. All of the extracts were collected and weighed. The total amount of extract, the extraction efficiencies and the concentration fac-
tors of ␥-oryzanols, free fatty acids, triglycerides were then calculated by the following equations: TY =
Ri =
weight of the extracted oil × 100%; total yield. weight of feeding material
weight of i component in the extracted oil × weight of i component in Soxhlet extracted oil oryzanols 100%, i = FFAs ; extraction efficiency. TGs
(1)
(2)
326
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
extraction efficiency of i component , total yield oryzanols i = FFAs ; concentration factor.
βi =
(3)
TGs
2.5. Supercritical carbon dioxide extraction of rice bran oil Fig. 2 displays a schematic flow diagram of pilot-scale SCCO2 extraction of rice bran oil from powdered rice bran. A mass of rice bran powder varying from 0.6 kg to 1 kg was individually packed inside a 5 L stainless steel tubular extractor. Liquid CO2 from a cylinder (1) was passed through a chiller (3) at 277 K, and was compressed to the desired working pressure using a highpressure pump (2) and heated to supercritical conditions using a pre-heater (6-1). This carbon dioxide flowed into the extractor (7); came into contact with the rice bran powder, and was used to extract the oil. A heating circulator (9-1) was maintained at a constant temperature; a metering valve (12-5) located at the outlet was manually adjusted to maintain constant extraction pressure. A drop in the pressured drove the oil-laden CO2 into a 1 L separator (8) following the extraction. The SC-CO2 extract then precipitated in a separator that was maintained at 50 bar and 308 K. The amount of low-pressure CO2 was measured using a wet gas meter (10) and thus returned to the ambient conditions. At the end of each experiment, the extracted oil was collected through a metering valve (13-3). The oil was weighed and stored at 273 K before use. The total yield, the extraction efficiency, the concentration factors of ␥-oryzanols, free fatty acids and triglycerides in the extracts were then calculated. 2.6. Supercritical carbon dioxide deacidification of rice bran oil Fig. 3 presents a schematic flow diagram of the SC-CO2 deacidification of rice bran oil. The type stainless steel packed material was directly loaded into a 275 mL (2.2 cm I.D. × 75 cm L) high-pressure distillation column (8). Then, 13 g of the rice bran oil obtained from the SC-CO2 extraction at 300 bar and 313 K was loaded and prepared for the SC-CO2 deacidification. Liquid CO2 from a cylinder (1) was passed through a cooling bath (3) at 277 K, preheated by a hot plate (6) through an oil bath (7), and was compressed using a syringe pump (4). This carbon dioxide, maintained at a flow rate of 10 g/min,
flowed into the deacidification column whose pressure was maintained by a back-pressure regulator (9-1). A heating element, equipped with a PID temperature controller (16), was thermo-statically maintained at a constant temperature. Following SC-CO2 deacidification, a drop in pressure drove the acid-laden CO2 into a separator (12), and the gas was then expanded through a spiral-type nozzle at 50 bar. The amount of low-pressure CO2 was measured using a wet gas meter (14) before the gas was returned to the ambient conditions. Following this process, the deacidified oil was collected in a flask (13) after depressurization using a metering valve (11-3) and was then ready for analysis and calculation. In addition, the free fatty acids-enriched extracted oil was also collected in the sample vial by opening a metering valve (11-2). 3. Results and discussion 3.1. Three purified γ-oryzanols The 99% pure cycloartenyl ferulate was purchased from a chemicals supplier. The 98.5% pure 24-methylenecycloartanyl ferulate and the 98% pure campesteryl ferulate were obtained by semi-preparative HPLC chromatography. Fig. 4(a)–(c) shows the 1 H NMR chemical shifts of cycloartenyl ferulate, 24-methylenecycloartanyl ferulate and campesteryl ferulate, respectively. The 24-methylenecycloartanyl ferulate chemical shifts pattern obtained using 1 H NMR in CDCl3 , δ was 7.60 (d, J = 15.8 Hz), 7.08 (dd, J = 1.8, 8.1 Hz), 7.04 (d, J = 1.8 Hz), 6.92 (d, J = 8.4 Hz), 6.30 (d, J = 16.1 Hz), 5.85 (br s), 4.72 (br s), 4.71 (m), 4.67 (br s), 3.94 (s), 2.24 (sept., J = 7.0 Hz), 1.04 (d, J = 7.0 Hz), 1.03 (d, J = 6.6 Hz), 0.98 (s), 0.92 (s), 0.9 (d, J = 5.5 Hz), 0.9 (s), 0.60 (d, J = 4.0 Hz, endo) and 0.37 (d, J = 4.0 Hz, exo). The campesteryl ferulate chemical shifts pattern revealed by 1 H NMR in CDCl3 , δ was 7.60 (d, J = 15.8 Hz), 7.07 (dd, J = 1.8, 8.4 Hz), 7.03 (d, J = 1.8 Hz), 6.91 (d, J = 8.4 Hz), 6.28 (d, J = 16.1 Hz), 5.84 (br s), 5.40 (m), 4.71 (m), 3.92 (s), 2.41 (br s), 2.39 (br s), 1.05 (s), 0.92 (d, J = 6.6 Hz), 0.85 (d, J = 6.2 Hz), 0.81 (d, J = 6.8 Hz), 0.78 (d, J = 6.6 Hz) and 0.69 (s). Yasukawa et al. reported these three NMR patterns [7]. 3.2. Soxhlet extractions of rice bran oil Table 1 presents experimental data on Soxhlet n-hexane extractions of rice bran oil. Feeds #1, #2, #3 were from the
Table 1 Experimental data on Soxhlet n-hexane extractions of rice bran oil from 15 g of four rice bran powder Feed #
Woil (g)
TY (%)
WOry (mg/goil )
βOry
WFFA (mg/goil )
βFFA
WTG (mg/goil )
WFFA + WTG (mg/goil )
βTG
Wothers (mg/goil )
1 2 3 4
3.00 2.90 2.82 2.50
20.0 19.3 18.8 16.7
15.2 11.4 11.4 10.5
5.00 5.18 5.32 5.99
95.0 152 180 38.9
5.00 5.18 5.32 5.99
800 752 726 875
895 904 906 914
5.00 5.18 5.32 5.99
89.8 84.6 82.6 75.6
#1 and #4 feed: fresh rice bran; #2 and #3 feed: aged rice bran; Woil : weight of extracted oil; TY: total oil yield = (Woil /WRB ) × 100%; WRB : weight of rice bran; WOry : concentration of oryzanols; WFFA : concentration of free fatty acids; WTG : concentration of triglycerides; βOry : oryzanols concentration factor = ROry /TY; βFFA : free fatty acids concentration factor = RFFA /TY; βTG : triglycerides concentration factor = RTG /TY; Wothers : concentration of waxes, glycolipids and phospholipids; ROry : oryzanols recovery; RFFA : free fatty acids recovery; RTG : triglycerides recovery; ROry = RFFA = RTG = 100%.
Table 2 Experimental data on SC-CO2 extractions of rice bran oil from 35 g rice bran powder at 300 bar and 313 K Experiment #
Woil (g)
TY (%)
WOry (mg/goil )
ROry (%)
βOry
1a
6.15 ± 0.01 5.59
17.6 ± 0.1 16.0
15.3 ± 0.1 8.1
88.3 ± 0.5 73.7
5.04 ± 0.02 4.61
2b
WFFA (mg/goil ) 106 ± 1 39.8
RFFA (%)
βFFA
WTG (mg/goil )
RTG (%)
βTG
Wothers (mg/goil )
97.7 ± 0.3 98.6
5.56 ± 0.03 6.16
831 ± 1 876
91.0 ± 0.4 96.5
5.19 ± 0.01 6.03
48.8 ± 1.0 76.2
Table 3 Experimental data on SC-CO2 extractions of rice bran oil from 0.6 to 1.03 kg powder at 300 bar and 313 K Experimenta #
WRB (kg)
WCO2 (kg)
3 4
0.60 1.03 ± 0.01
12.1 20.5 ± 0.2
Woil (g) 90.7 157 ± 5
TY (%)
WOry (mg/goil )
ROry (%)
βOry
WFFA (mg/goil )
RFFA (%)
βFFA
15.1 15.7 ± 0.5
6.0 6.3 ± 0.1
52.0 56.3 ± 2.4
3.44 3.58 ± 0.04
35.4 37.5 ± 0.8
82.3 90.7 ± 4.5
5.45 5.76 ± 0.13
WTG (mg/goil ) 864 866 ± 7
RTG (%)
βTG
Wothers (mg/goil )
89.0 93.4 ± 3.7
5.89 5.93 ± 0.05
94.6 89.9 ± 7.2
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
Woil : weight of extracted oil; TY: total oil yield = (Woil /WRB ) × 100%; WRB : weight of rice bran; WOry : concentration of oryzanols; WFFA : concentration of free fatty acids; WTG : concentration of triglycerides; ROry : oryzanols extraction efficiency = [(WOry × TY)/(WOry,Soxhlet × TYSoxhlet )] × 100%; RFFA : free fatty acids extraction efficiency = [(WFFA × TY)/(WFFA,Soxhlet × TYSoxhlet )] × 100%; RTG : triglycerides extraction efficiency = [(WTG × TY)/(WTG,Soxhlet × TYSoxhlet )] × 100%; βOry : oryzanols concentration factor = ROry /TY; βFFA : free fatty acids concentration factor = RFFA /TY; βTG : triglycerides concentration factor = RTG /TY; Wothers : concentration of waxes, glycolipids and phospholipids. a First feed in Table 1. b Fourth feed in Table 1.
WRB : weight of rice bran; WCO2 : weight of carbon dioxide; Woil : weight of extracted oil; TY: total oil yield = (Woil /WRB ) × 100%; WOry : concentration of oryzanols; WFFA : concentration of free fatty acids; WTG : concentration of triglycerides; ROry : oryzanols extraction efficiency = [(WOry × TY)/(WOry,Soxhlet × TYSoxhlet )] × 100%; RFFA : free fatty acids extraction efficiency = [(WFFA × TY)/(WFFA,Soxhlet × TYSoxhlet )] × 100%; RTG : triglycerides extraction efficiency = [(WTG × TY)/(WTG,Soxhlet × TYSoxhlet )] × 100%; βOry : oryzanols concentration factor = ROry /TY; βFFA : free fatty acids concentration factor = RFFA /TY; βTG : triglycerides concentration factor = RTG /TY; Wothers : concentration of waxes, glycolipids and phospholipids. a Fourth feed in Table 1.
327
SC-CO2 deacidification at 10 g/min; P: pressure; T: temperature; WCO2 : weight of carbon dioxide; Woil : weight of extracted oil; Roil : oil retention = Woil /Woil,feed × 100%; Ory: oryzanols concentration = (WOry /Woil ) × 100%; FFA: free fatty acids concentration = (WFFA /Woil ) × 100%; TG: triglycerides concentration = (WTG /Woil ) × 100%; ROry : oryzanols recovery = (WOry /WOry,feed ) × 100%; RFFA : free fatty acids recovery = (WFFA /WFFA,feed ) × 100%; RTG : triglycerides recovery = (WTG /WTG,feed ) × 100%; βOry : oryzanols concentration factor = Ory/Oryfeed ; βFFA : free fatty acids concentration factor = FFA/FFAfeed ; βTG : triglycerides concentration factor = TG/TGfeed ; RRFFA : free fatty acids removal = (WFFA,feed − WFFA )/WFFA,feed × 100%. Bold values present the best run.
63.8 93.7 85.3 55.4 83.9 95.5 70.5 93.4 97.8 92.7 90.2 99.9 97.2 88.5 99.3 1.10 1.11 1.09 1.09 1.09 1.10 1.11 1.08 1.10 1.09 1.08 1.10 1.09 1.09 1.09 89.7 85.7 86.7 83.5 79.7 84.0 86.3 84.5 85.5 81.0 74.6 70.1 84.6 85.1 90.6 Oil contained 0.63% Ory, 3.75% FFA, 86.6% TG before deacidification 1(F) 900 200 343 11.4 87.6 2(F) 2700 200 343 10.9 83.9 3(A) 1800 200 353 11.2 85.8 4(F) 900 200 363 11.5 88.1 5(F) 2700 200 363 11.2 85.8 6(A) 1800 250 343 10.2 78.1 7(A) 900 250 353 11.3 86.9 8(C) 1800 250 353 11.1 84.9 9(A) 2700 250 353 10.7 82.2 10(A) 1800 250 363 10.8 83.1 11(F) 900 300 343 9.2 70.8 12(F) 2700 300 343 6.5 49.7 13(A) 1800 300 353 9.2 70.9 14(F) 900 300 363 10.5 80.9 15(F) 2700 300 363 9.1 70.1
0.64 0.64 0.63 0.59 0.58 0.67 0.62 0.62 0.65 0.61 0.66 0.88 0.75 0.66 0.81
1.96 0.35 0.80 2.50 0.89 0.30 1.62 0.38 0.13 0.42 0.82 0.01 0.21 0.73 0.05
95.2 95.8 94.4 93.9 94.2 95.2 95.8 93.5 94.9 94.6 93.3 94.9 93.9 94.6 94.2
45.9 7.8 18.4 59.1 20.4 6.3 37.7 8.6 2.9 9.3 15.5 0.1 4.0 15.7 0.9
96.4 92.7 93.9 96.0 93.6 86.2 96.4 92.1 90.2 90.8 76.2 54.6 76.6 88.2 76.1
1.02 1.02 1.00 0.94 0.92 1.06 0.98 0.98 1.03 0.97 1.05 1.40 1.19 1.05 1.29
0.52 0.09 0.21 0.67 0.24 0.08 0.43 0.10 0.03 0.11 0.22 0.003 0.06 0.20 0.01
RRFFA (%) TG FFA Ory RTG (%) RFFA (%) ROry (%) TG (%) FFA (%) Ory (%) Roil (%) Woil (g) T (K) P (bar) WCO2 (g) RSM #
Table 4 Experimental designed SC-CO2 deacidifications of rice bran oil obtained by SC-CO2 extraction at 300 bar and 313 K
55.9 78.6 73.2 48.8 72.0 74.6 61.3 79.3 80.4 77.0 63.9 49.7 68.9 71.6 69.6
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331 Roil × RRFFA
328
first batch of rice bran. The #1 feed was fresh and the other two feeds were aged for 6 months. The #4 fresh feed was from the second batch of rice bran. The aging process by the lipase transformation at room temperature results in relatively high content of free fatty acids in the rice bran oil. The maximal concentration of ␥-oryzanols in the oil extracted from these four samples was 15.2 mg/goil , after in 4 h of the 300 mL Soxhlet n-hexane extraction of 15 g rice bran. Triglycerides and free fatty acids are two major components in the extracted oil. Table 1 indicates that the summation of these two compounds reached 90%. The lowest content of free fatty acids 38.9 mg/goil , was found in the fourth extracted oil. 3.3. SC-CO2 extractions of rice bran oil Table 2 presents experimental data on the SC-CO2 extraction of rice bran oil from 35 g of rice bran powder. The oils extracted in Experiments #1 and #2 in Table 2 were produced from the first and fourth feed materials in Table 1, respectively. Total oil yield exceeded 16% upon extraction at 300 bar and 313 K when 2450 g of carbon dioxide was used. Table 3 presents experimental data on the SC-CO2 extraction of rice bran oil from 0.6 kg to 1.03 kg of powder. The total oil yield exceeded 15% upon extraction at 300 bar and 313 K using a constant solvent-to-feed ratio of 20. The concentrations of ␥-oryzanols, free fatty acids and triglycerides remain unchanged. The oil extracted in Experiment #4 was used for following SC-CO2 deacidifications. 3.4. Experimentally designed SC-CO2 deacidifications of rice bran oil A center composite scheme of pressure, temperature and CO2 consumption of response surface methodology (RSM), configured with six axial points, eight factor points and one center point, was designed for the SC-CO2 deacidification of 13 g of rice bran oil, obtained in Experiment #4, as indicated in Table 3. The flow rate of CO2 was constant at 10 g/min and the height of the packed column was fixed at 45 cm with 67.5 g of packing material. Table 4 displays experimental data on this RSM designed SC-CO2 deacidification at pressures from 200 bar to 300 bar and temperatures from 343 K to 333 K, using 900 g to 2700 g of carbon dioxide. In SC-CO2 deacidification experiments, the amount of remaining triglycerides and the removal efficiencies of free fatty acids are two major variables of interest. The free fatty acid content, 0.13%, in the deacidified oil was obtained at 250 bar, 353 K and 2700 g of CO2 extraction. This experiment demonstrated that the retention efficiency of oil and the removal efficiency of free fatty acids were 82.2% and 97.8%, respectively. The product of these two responses reached 80.4, which is the highest value among all 15 RSM experiments. Further examination of these data revealed that the concentration factors of oryzanols and triglycerides increased, but the concentration factors of free fatty acids decreased to zero (datum #9), implying that active compounds in the deacidified oil were concentrated and the free fatty acid content in the oil was substantially decreased.
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
329
Fig. 5. Three-dimensional responded experimental data on removal efficiency of free fatty acids using SC-CO2 deacidification, datum #10 in Table 4 (F-testing: R2 = 0.9788, S.D. = 3.30). (a) WCO2 : 1800 g, (b) temperature: 363 K, and (c) pressure: 250 bar.
Fig. 6. Three-dimensional responded experimental data on retention efficiency of oil using SC-CO2 deacidification, datum #8 in Table 4 (F-testing: R2 = 0.9798, S.D. = 2.44). (a) WCO2 : 1800 g, (b) temperature: 353 K, and (c) pressure: 250 bar.
330
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331
Fig. 7. Response surface methodology optimization of the multiple value of oil retention and free fatty acids removal efficiency responses (F-testing: R2 = 0.9451, S.D. = 4.04). (a) WCO2 : 2160 g, (b) temperature: 363 K, and (c) pressure: 260 bar.
Fig. 5 shows that effects of pressure and the amount of consumed CO2 are important to the removal efficiency of free fatty acids. Fig. 6 reveals that the effect of pressure is more significant than that of CO2 consumption. The effect of temperature is insignificant because the operative region is close to the crossover pressure and the solubility of triglycerides in supercritical carbon dioxide increases as the fluid density increases with pressure. Fig. 7 plots the effects of the pressure and CO2 consumption associated with a multiple response of the retention of oil and the FFA removal efficiency. The value of this response is optimal at 260 bar, 363 K and with 2160 g of CO2 consumed. The CO2 consumption and pressure both affect the retention of oil and the removal of free fatty acids.
and 353 K with 2700 g of CO2 consumed. The three-factor center composite designed SC-CO2 deacidifications of rice bran oil demonstrate that carbon dioxide consumption and pressure significantly influence the removal efficiency of free fatty acids and the retention efficiency of triglycerides. Acknowledgments The authors would like to thank the National Science Council of the Republic of China, Taiwan, for financially supporting this research under Contract No. NSC96-2628-E005-085-MY2. This work is also partially supported by the Ministry of Education, Taiwan, ROC, under the ATU plan.
4. Conclusions References The purest 24-methylenecycloartanyl ferulate and campesteryl ferulate were successfully isolated from a mixture of ␥-oryzanols using a preparative C18 high-performance liquid chromatography method. The 15.7% of rice bran oil was obtained by using a pilot-scale SC-CO2 extraction at 300 bar and 313 K, the recovery of extracted oil was 94.0%, which was close to that by using a lab-scale SC-CO2 extraction at the same operation conditions. The oryzanol and free fatty acid contents in the extracted oil were 0.63% and 3.75%, respectively. An oil retention efficiency of 82.2% and an FFA removal efficiency of 97.8% were achieved using SC-CO2 deacidification at 250 bar
[1] B.O. Juliano, P.A. Hicks, Rice functional properties and rice food products, Food Rev. Int. 12 (1996) 71–103. [2] R.N. Sayre, Rice Bran as a Source of Edible Oil and Higher Value Chemicals, Western Regional Research Center, ARS, USDA, 1998. [3] W. Hu, J.H. Wells, T.S. Shin, J.S. Godber, Comparison of isopropanol and hexane for extraction of vitamin E and oryzanols from stabilized rice bran, J. Am. Oil Chem. Soc. 73 (1996) 1653– 1656. [4] Z. Xu, J.S. Godber, Purification and identification of components of ␥-oryzanols in rice bran oil, J. Agric. Food Chem. 47 (1999) 2724– 2728.
C.-R. Chen et al. / J. of Supercritical Fluids 45 (2008) 322–331 [5] J.K. Duve, P.J. White, Extraction and identification of antioxidants in oats, J. Am. Oil Chem. Soc. 68 (1991) 365–370. [6] G.S. Seetharamaiah, N. Chandrasekhara, Studies on hypocholesterolemic activity of rice bran oil, Atherosclerosis 78 (1989) 219–223. [7] K. Yasukawa, T. Akihisa, Y. Kimura, T. Tamura, M. Takido, Inhibitory effect of cycloartenol ferulate, a component of rice bran, on tumor promotion in two-stage carcinogenesis in mouse skin, Biol. Pharm. Bull. 21 (1998) 1072–1076. [8] S.H. Nam, S.P. Choi, M.Y. Kang, N. Kozukue, M. Friedman, Antioxidative antimutagenic, and anticarcinogenic activities of rice bran extracts in chemical and cell assays, J. Agric. Food Chem. 53 (2005) 816– 822. [9] T. Akihisa, K. Yasukawa, M. Yamaura, M. Ukiya, Y. Kimura, N. Shimizu, K. Arai, Triterpene alcohol and sterol ferulates from rice bran and their anti-inflammatory effects, J. Agric. Food Chem. 48 (2000) 2313–2319. [10] H.F. Luo, Q. Li, S. Yu, T.M. Badger, N. Fang, Cytotoxic hydroxylated triterpene alcohol ferulates from rice bran, J. Nat. Prod. 68 (2005) 94–97. [11] N. Fang, S. Yu, T.M. Badger, Characterization of triterpene alcohol and sterol ferulates in rice bran using LC–MS/MS, J. Agric. Food Chem. 51 (2003) 3260–3267. [12] W. St¨oggl, C. Huck, S. Wongyai, H. Scherz, G. Bonn, Simultaneous determination of carotenoids, tocopherols, and ␥-oryzanols in crude rice bran oil by liquid chromatography coupled to diode array and mass spectrometric detection employing silica C30 stationary phases, J. Sep. Sci. 28 (2005) 1712–1718. [13] C. Aguilar-Garcia, G. Gavino, M. Baraga˜no-Mosqueda, P. Hevia, V.C. Gavino, Correlation of tocopherol, tocotrienol, ␥-oryzanols and total polyphenol content in rice bran with different antioxidant capacity assays, Food Chem. 102 (2007) 1228–1232. [14] A.G.G. Krishna, K.H. Hemakumar, S. Khatoon, Study on the composition of rice bran oil and its higher free fatty acids value, J. Am. Oil Chem. Soc. 83 (2006) 117–120.
331
[15] S. Zullaikah, C.C. Lai, S.R. Vali, Y.H. Ju, A two-step acid-catalyzed process for the production of biodiesel from rice bran oil, Bioresour. Technol. 96 (2005) 1889–1896. [16] Z. Shen, M.V. Palmer, S.S.T. Ting, R.J. Fairclough, Pilot scale extraction of rice bran oil with dense carbon dioxide, J. Agric. Food Chem. 44 (1996) 3033–3039. [17] B.M. Bhosle, R. Subramanian, New approaches in deacidification of edible oils—a review, J. Food Eng. 69 (2005) 481–494. [18] H.J. Kim, S.B. Lee, K.A. Park, I.K. Hong, Characterization of extraction and separation of rice bran oil rich in EFA using SFE process, Sep. Purif. Technol. 15 (1999) 1–8. [19] C.J. Chang, C.C. Chen, High-pressure densities and P-T–x-y diagrams for carbon dioxide + linalool and carbon dioxide + limonene, Fluid Phase Equilib. 163 (1999) 119–126. [20] C.J. Chang, M.S. Lee, B.C. Li, P.Y. Chen, Vapor–liquid equilibria of CO2 with four unsaturated fatty acid esters at elevated pressure, Fluid Phase Equilib. 233 (2005) 56–65. [21] Z. Shen, M.V. Palmer, S.S.T. Ting, R.J. Fairclough, Pilot scale extraction and fractionation of rice bran oil using supercritical carbon dioxide, J. Agric. Food Chem. 45 (1997) 4540–4544. [22] N.T. Dunford, J.W. King, Phytosterol enrichment of rice bran oil by a supercritical carbon dioxide fractionation technique, J. Food Sci. 65 (2000) 1395–1399. [23] N.T. Dunford, J.W. King, Thermal gradient deacidification of crude rice bran oil utilizing supercritical carbon dioxide, J. Am. Oil Chem. Soc. 78 (2001) 121–125. [24] N.T. Dunford, J.A. Teel, J.W. King, A continuous countercurrent supercritical fluid deacidification process for phytosterol ester fortification in rice bran oil, Food Res. Int. 36 (2003) 175–181. [25] L. Danielski, C. Zetzl, H. Hense, G. Brunner, A process line for the production of raffinated rice oil from rice bran, J. Supercrit. Fluids 34 (2005) 133–141.
