Extraction and evaluation of antioxidant potential of the extracts obtained from tamarind seeds (Tamarindus indica), sweet variety

Journal of Food Engineering 173 (2016) 116e123

Contents lists available at ScienceDirect

Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng

Extraction and evaluation of antioxidant potential of the extracts obtained from tamarind seeds (Tamarindus indica), sweet variety ^
ulia Maria Cardoso Lima Reis a, *, Claudio Dariva b, Gizelle Angela Pa Barroso Vieira c, Haiko Hense a
polis, SC, Brazil Department of Chemical and Food Engineering, Federal University of Santa Catarina, EQA/UFSC, C. P. 476, CEP 88040-900, Floriano Department Process Engineering, University Tiradentes, Aracaju, SE, Brazil c ~ Department of Chemical e IF SERTAO-PE, Petrolina, PE, Brazil a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 May 2015 Received in revised form 29 October 2015 Accepted 2 November 2015 Available online 6 November 2015

The objective of this study was to evaluate the antioxidant activity of tamarind seed extracts, sweet variety. The extracts were obtained by sub and supercritical CO2 different concentrations of ethanol. The extractions with pure CO2 were performed in the range of 20e50
C and 100e250 bar and with ethanol under the conditions of 50
C and 250 bar as well at 40
C and ambient pressure. The yields of extractions with pure CO2 of better quality and quantity of antioxidant were obtained under the conditions of 50
C and 250 bar. However, the phenolic content of the tamarind seeds extract, sweet variety, obtained with pure ethanol at 50
C and 250 bar was superior to all the others, including the synthetic antioxidant butyl hydroxytoluene (BHT). Therefore, the ethanolic extracts of tamarind seed, sweet variety, are potential sources of natural antioxidants for the food industry. © 2015 Elsevier Ltd. All rights reserved.

Chemical compounds studied in this article: 9,12-Octadecadienoic acid (PubChem CID: 5280450) Cis-10-heptadecenoic (PubChem CID: 4385123) n-Hexadecanoic acid (PubChem CID: 985) Cis-9-octadecenoic acid (PubChem CID: 445639) Oleic acid (PubChem CID: 445639) Cis-11-octadecenoic acid (PubChem CID: 5282761) 1,2 benzenedicarboxylic acid (PubChem CID: 1017) Ethylbenzene (PubChem CID: 7500) Decane (PubChem CID: 15600) Trifluoroacetate (PubChem CID: 84468) Keywords: Sweet brazilian tamarind seed Antioxidant Supercritical CO2

1. Introduction Tamarind (Tamarindus indica) is a fruit plant that belongs to the legume family, native to equatorial Africa, India and Southeast Asia and grows in tropical and subtropical regions, with ideal average temperature of 25
C. It is considered an ideal tree for semi-arid regions, tolerating 5e6 months in dry conditions but does not

* Corresponding author. Tel.: þ55 87 2101 4330; fax: þ55 87 2101 4328. E-mail address: [email protected] (P.M. Cardoso Lima Reis). http://dx.doi.org/10.1016/j.jfoodeng.2015.11.001 0260-8774/© 2015 Elsevier Ltd. All rights reserved.

survive at low temperatures (Pereira et al., 2011). There are different varieties of Tamarindus indica and they can be divided into acidic and sweet. Acidic varieties are commonly found in most countries, therefore easily develops into warm, sunny locations. The varieties of sweet type are not readily available. In Thailand, two types of Tamarind are found in abundance, so-called sweet and sour varieties (Sudjaroen et al., 2005). The sweet variety is rarely found in Brazil, but in the interior of
, the Bebedouro farm has a planted area Bahia, in the city of Sento Se of “sweet” Tamarind, which is the name that was given to the fruit by not presenting the characteristic acidity of the common

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

tamarind. There are few studies in the literature with sweet variety of Tamarindus indica. Fruits, leaves and seeds are natural sources of antioxidants and several studies have bet on this alternative to replacing synthetic antioxidants (Ramalho and Jorge, 2006). The extraction with supercritical or subcritical fluid has as one of its advantages the obtaining of solvent-free extract with high purity and, as a disadvantage, the high cost due to the use of materials and structure resistant to high pressures (Brunner, 1994). Tsuda et al. (1994) identified four antioxidants in the seed coat of Indian tamarind: 2-hydroxy-30 ,40 dihydroxyacetophenome; methyl 3, 4-hidydroxybenzoate; 3,4-dihydrophenyl acetate and ()-epicatechin. Sudjaroen et al. (2005) identified the profile of polyphenolics in Tamarind pericarp, the extracts obtained by Soxhlet with methanol, was dominated by proanthocyanidins (73.4%) in various forms (þ)-catechin (2.0%), procyanidin B2 (8.2%), ( )-epicatechin (9.4%), procyanidin trimer (11.3%), procyanidin tetramer (22.2%), procyanidin pentamer (11.6%), procyanidin hexamer (12.8%) along with taxifolin (7.4%), apigenin (2.0%), eriodictyol (6.9%), luteolin (5.0%) and naringenin (1.4%) of total phenols. The content of tamarind seeds comprised only procyanidins, represented (%) mainly by oligomeric procyanidin tetramer (30.2), procyanidin hexamer (23.8), procyanidin trimer (18.1), procyanidin pentamer (17.6) with lower amounts of procyanidin B2 (5.5) and () epicatechin (4.8). Luengthanaphol et al. (2004) found only work on sweet tamarind variety supercritical fluid extractions performed on the range of 35e80
C and 10e30 MPa and met -() epicatequinana (z22.mu.g of ()- epicatechin, per 100 g of the tamarind seed coat sweet Thai. The use of a co-solvent of 10% ethanol resulted in a much higher yield of ()- epicatechin, (z13 mg/100 g), under the best conditions found to be 40
C and 10 MPa. Mathematical modeling is crucial for predicting the behavior of the extraction process on an industrial scale, providing the data needed to build equipment and projection of industrial plants, contributing to the evaluation of the economic viability of this process (Mezzomo et al., 2009). In Brazil, the extraction and characterization of the compounds present in the seed of tamarind are unprecedented, requiring further studies on the and the economic viability of this residue. The objectives of this work were to obtain the Tamarindus indica seed extracts, sweet variety, with pure CO2 and with ethanol in different concentrations in sub and supercritical conditions (temperatures of 20, 35 and 50
C and pressures of 100, 175 and 250 bar), and evaluation of its antioxidant activity and fatty acid profile.

117

then weighed masses retained on each sieve for the calculation of average particle diameter as described by Gomide (1983); Determination of aApparent specific mass (ra) was determined from the ratio of the sample mass used in extractor and the volume occupied by the fixed bed, including the pores of the bed and not the inner pores of the particles. Real specific mass (rr), was estimated using a gas He pycnometer (AccuPyc II 1340, Micrometrics, 2009). Bed porosity (3 ) was determined from the ratio by the real and apparent specific mass of tamarind seeds sample, sweet variety, including the pores of the bed and the interior of the particles. 2.2. Supercritical extraction equipment The supercritical extraction equipment consists of a CO2 cylinder with 99.9% purity (Praxair, Brazil), a heating tape (Teledyne Isco, model 500 D, syringe pump) a liquid pump Alliance HPLC Lab Series III, stainless steel extractor and two thermostatic baths. The first bath (Julabo, model F32), used to cool the solvent prior to entering the syringe pump and the second bath (Quimis, scientific devices Ltda., Model Q214-M2) to maintain the desired temperatures in the heated extractor. 2.3. Preliminary supercritical extraction The preliminary extractions with wet and dry samples were held in intermediate conditions of temperature and pressure, 35
C and 175 bar, and subsequently used in the yield tests for 4 h. The yields of extracts obtained from the use of wet and dry samples determined the choice of which sample would be used for subsequent tests. 2.4. Kinetic experiment e preliminary extraction curve and kinetic parameters Kinetic curve was constructed by collecting the solute at time intervals of 15e30 min at 250 bar and 50
C, maximum conditions subsequently used in the experimental design, for approximately 7 h with a flow rate of 2 mL/min, to reach the diffusion period (DCP) of the extraction curve. From the analysis of the kinetic curve, the extraction time for global yield extraction was determined. The extraction curves were plotted using the software Microsoft Excel Home and Business 2010, version 14.0.4760.1000, for yield evaluation, being the mass of accumulated extract in the function of extraction time. 2.5. Determination of apparent solubility

2. Materials and methods 2.1. Raw material
e Bahia, The raw material was collected in the city of Sento Se Brazil, in the months of December 2011 and January 2012. The tamarind seeds, sweet variety, were manually removed from the pulp and stored in the refrigerator at 4
C. The seeds were dried at 40
C for 24 h and stored in a transparent bottle coated with foil protecting it from light and heat. Seeds were ground in a knife mill and separated in a sieve shaker 24e48 mesh. The moisture content was determined by the method of direct drying at 105
C, described by Institute Adolfo Lutz (Brazil, 2005). Average particle diameter (ds) was determined by sifting 100.00 g of powdered seeds for 30 min and the mass retained in each sieve was used in the calculations, as described by Gomide (1983) in semi analytical balance (Quimis, model Q520-5200), accurate to two decimals, were weighed. The sieves were shaken for 30 min, and

Solubility is a required parameter for mathematical modeling in
(1994) and Sovova
modified by Martínez and Martínez, Sovova (2008) models. This was called preliminary because it is not the main objective of the work. This apparent solubility was obtained by the dynamic method of extraction, where solubility is represented by the slope of the extraction curves, in the stage of constant rate (CER). According to Danielski (2002), the time of contact between the phases needed to achieve balance is between 0.9 g/min to 1.4 g/min for the oleoresins solubility in supercritical CO2. Thus, the flow rate determined for this experiment was 1.1 ± 0.2 g/mL, promoting the saturation of the solvent with the oil at the outlet of the extractor. The extraction kinetic curve was used for determination of the apparent solubility in this experiment. 2.6. Obtaining extracts The extracts were obtained by Soxhlet, as described by the

118

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

Institute Adolfo Lutz (Brazil, 2005), and sub and supercritical CO2, pure and combined with ethanol. A 22 factorial design experiments with triplicate at the central point was used to verify the effects of temperature (T) and pressure (P) in the global yield of extracts. The temperatures used were 20
C, 35
C and 50
C and pressures were 100, 175 and 240 bar. These values were selected to cover a wide range of temperature, pressure and specific mass of CO2 solvent. In the condition of 250 bar and 50
C, the extractions were made with different concentrations of CO2 in ethanol, according to Table 2. The extracts were stored in amber bottles, covered with aluminum foil, under refrigeration.

2.7. Determining the specific mass of the supercritical solvent (pco2) Specific mass of the CO2 in each condition was determined using the Equation of Angus et al. (1976).

2.9. Antioxidant activity 2.9.1. Determination of total phenolic compounds (TPC) Phenolic compounds determination was performed using the Folin-Ciocalteu method (SINGLETON; Rossi and Singleton, 1965; Peschel et al., 2006). The reaction mixture was composed by 0.1 mL of extract (1667 mg/L), 7.9 mL of distilled water, 0.5 mL of FolineCiocalteau reagent (a mixture of phosphomolybdate and phosphotungstate) and 1.5 mL of 20% sodium carbonate. The flasks were shaken, held for 2 h, and the absorbance was measured at 765 nm. The TPC was calculated using a standard curve (R2 ¼ 0.97), previously prepared with gallic acid as standard, according to Equation (2). The TPC analyses were carried out in triplicate and the results were expressed as milligrams of gallic acid equivalent (GAE) per gram of the extract (mg GAE/g). C ¼ 709.48*A

(2)

C ¼ concentration; A ¼ absorbance;

2.8. Mathematical modeling For mathematical modeling of the results the models of Martínez et al. (2003), Esquível et al., 1999, Crank (1975) and
(1994) modified by Martínez and Martínez, (2008) were Sovova used. It was performed using the software Mass Transfer, developed by Correia et al. (2006), which uses the maximum likelihood method to minimize the sum of the calculated squared deviations (Kitzberger et al., 2009). The kinetic parameters (rate of mass transfer of step at constant rate (CER) and the times tCER and tFER) were calculated according to Mezzomo et al. (2009) and Benelli et al. (2010). The rate of mass transfer at CER (MCER) step was obtained from linear regression of the curve during the extraction CER, while the solute concentration in the solvent phase at CER (YCER) step was determined by the ratio between MCER and QCO2. The times tCER and tFER represent the end of the constant and decreasing steps of extraction, respectively. For the application of the models, parameters that were experimentally determined or calculated from the experimental data are required. The tCER calculated through a global optimization method, while the MCER and yields from extraction in step CER (XCER) were determined from the tCER and adjustment of the tendency of the curves of extraction in step CER. From the value of MCER one can obtain the value of YCER through Equation (1): YCER ¼ MCER/QCO2

(1)

2.9.2. Determination of antioxidant activity by DPPH method The maximum absorption occurs at 517 nm. The sample stock solution (1 mg/mL) was diluted at concentrations of 500, 250, 125, 50, 25, 10 and 5 mg/mL in a solution of DPPH concentration 0.3 mM and incubated for 30 min at 22
C (Mensor et al., 2001). The concentration of the sample necessary to capture 50% of the free radical DPPH (EC50 e Effective Concentration) is calculated by linear and exponential regression analysis (Mensor et al., 2001). 2.10. Chromatographic analysis coupled with mass spectrometry The gas chromatograph coupled to mass spectrometer (GC/MS, Varian Saturn 2100D) technology “Ion Trap” and sampling system “purge and trap”, equipped with a Rtx 5MS capillary column (30 m  0.25 mm e film 0,25 um), split at a ratio of 1:30. The carrier gas is helium (He), with a constant flow of 1.2 mL/min). The column was heated to 60
C for 3 min, programmed at 5
C per min to 220
C, and maintained at 220
C for 5 min. The extracts were diluted in ethyl acetate and inject at 250
C. The chemical compounds identification was based on the comparison of the mass spectrum of the substance with the GCeMS system database (Data Analysis and Technical Graphics e Origin ®) and retention index. The quantification of the chemical composition of the extracts was checked by the relative proportion to the areas obtained for each compound in the respective chromatograms. 2.11. Statistical analysis

Table 1 Kinetic parameters of extraction of tamarind, sweet variety, at 250 bar and 50
C. Parameters

a

t (min) mb (g) Xoc (%) Md (g/min) Ye (kg/kg) a b c d e f g h

Extraction periods CERf

FERg

DCPh

0e105 0.4097 1.64 0.0039 3.8945  103

105e200 0.2021 0.81 0.0021 e

>200 0.060 0.24 0.0015 e

t: duration of the extraction step. m: mass of extracted solute. Xo: extraction yield. M: extraction rate. Y: concentration of solute in the solvent phase. CER:constant rate step. FER: step of decreasing rate. DCP: diffusional rate step.

The results of all experiments were evaluated by analysis of variance (ANOVA) at the 5% level of significance (p < 0.05) with the aid of the Statistica 7.0 software (Statsoft Inc., USA). 3. Results and discussion 3.1. Preliminary tests The moisture content and volatile substances of the tamarind seeds in natura and dried were 10.38 ± 0.08% and 5.00 ± 0.085%, respectively. The results obtained in the raw material characterization were: average particle diameter of 397.6 mm, real specific mass of 1.45 ± 0.004 g/cm3; apparent specific mass of 0.40 g/cm3; bed porosity of 0.72 and mass of solid to form a fixed bed of 25.00 g. The volume occupied by the sample in the fixed bed, necessary for

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

119

Table 2 Results of the extracts of the seed of tamarind, sweet variety. Extracts Levels T T (
C) Levels P P (bar) 1 2 3 4 5 6 7 8 9 10 11 12 13 BHT

1 þ1 1 þ1 0 0 0

20 50 20 50 35 35 35 50 50 50 50 50 40 e

1 1 þ1 þ1 0 0 0

100 100 250 250 175 175 175 250 250 250 250 250 atmospheric e

rCO2 (g/cm3) (1) Solvent 0.85653 0.38535 0.96350 0.83497 0.84394 0.84394 0.84394 0.83497 0.83497 0.83497 0.83497 e e e

Xo (%)

CO2 CO2 CO2 CO2 CO2 CO2 CO2 CO2 þ 10% ethanol CO2 þ 25% ethanol CO2 þ 50% ethanol CO2 þ 75% ethanol 100% ethanol 100% ethanol (Soxhlet)

1.62ab 0.00b 2.62a 2.58ab 2.36ab 2.49ab 2.57ab 3,47c 3,44c 7,62d 20.27e 21.44e 20.69e

(2)

% AA (500 mg/mL) 31.45f e 28.21g 31.15f 53.91c 45.24d 41.90e 25.13h 97.18a 97.77a 97.61a 96.70a 97.00a 89.70b

(2) (3)

EC50 (mg/mL) 159.45cd e 307.63ab 97.72de 100.55de 190.47bcd 213.14bcd 370.82a 12.34e 8.92e e e e 261.00abc

(2) (3)

TPC (mg EAG/g

extrato)

(2) (4)

6.13hi e 21.92g 25.19g 9.26h 6.0hi 4.76j 25.6g 307.36e 361.42d 420.8b 471.05a 387.43c 268f

(1)

Angus et al., 1976. Equal letters do not differ significantly (p < 0.05). Antioxidant activity and EC50 by DPPH. (4) Total phenolic content (TPC), expressed as gallic acid equivalents. * Values determined by Benelli et al. (2010). (2) (3)

calculating the apparent specific mass was of 62.34 cm3. These measures ensure sufficient quantities of extracts for quantification and further analysis, and they also meet the necessary relationship between the height and diameter of the fixed bed to despise the mass transfer in the radial and axial directions, facilitating solvent flow with speed in the axial direction, condition
(1994) modified by Martínez and for application of the Sovova Martínez, (2008) model. These characteristics are also important for the standardization of the sample of tamarind seed, sweet variety, used during the experiments, facilitating the discussion of the parameters and results of supercritical extraction. 3.2. Yield curves of preliminary supercritical extraction The yield obtained in the extractions performed with the wet and dry samples were 2.47% and 2.77%, respectively. Tamarind seed dried at 40
C for 24 h gave a yield of approximately 12% higher than the moist seed. Probably, the humidity present in the sample compete with the supercritical fluid and influenced the performance of the extraction, therefore, the raw material used in the extraction was dried in a forced air dryer. 3.3. Kinetic curve and kinetic parameters of the extraction To evaluate the extraction yield curves and the influence of parameters, such as temperature and pressure, it is necessary to set a time of extraction which can be determined by the phases of the kinetic curve. The experiment to obtain the extraction kinetic curve was performed at 250 bar and 50
C and a solvent rate flow of 1.2 ± 0.2 g/ min, as shown in Fig. 1: In the constant rate of extraction (CER) step, the solute present in the particle surface is transferred by convection to the solvent; this step is between 0 and 105 min. In descending (FER) step, there is a depletion of solute on the particle surface and the mass transfer by diffusion initiate. The solvent penetrates the solid matrix through the free space (porosity), solubilizing the solute and returning to the particle surface; this step is between 105 and 200 min. The last stage has an extraction rate almost nil, called diffusional (DCP) step, i. e., the curve approaches to the value that is

the theoretical content of extractable solute (Xo) and it occurs from 200 min. Therefore, the time to reach the diffusive step is approximately 3.4 h. Thus to the yield curves analysis (Xo) the extractions were performed for 4 h, to guarantee that the scope of the diffusive step was reached. The kinetics parameters of extraction are presented in Table 1.
(1994) modified by To the mathematical model of Sovova Martínez and Martínez, (2008) is also necessary the apparent solubility, which was adopted in this work as 3.8945  103 kg/kg, as described in item 2.6. 3.4. Extraction yield curves (Xo) The extraction yields obtained by sub and supercritical extractions are presented in Table 2. According to the Table 2, the highest yield was 2.62% obtained at 250 bar and 20
C. This results was statistically similar to the other yields obtained in the same pressure and a temperature (50
C, 2.58%), and also at the mid point (175 bar, 35
C) at a 5% level of significance. There was no extract for supercritical extraction under conditions of 50
C and 100 bar with CO2, due to the low specific mass of the solvent, 0.38535 g/cm3, which decreases the solvating power of it. However, in a subcritical state, conditions of 20
C and 100 bar, a

Fig. 1. Extraction kinetics curve of seed of tamarind, sweet variety, at 250 bar and 50
C.

120

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

yellow oil with 1.62% of yield was obtained, confirming the positive influence of the specific mass on the yield of extracts that had increased to 0.85653 g/cm3. According to ANOVA (Table 3) the effect of temperature, pressure, as well as the temperature þ pressure interaction is significant in the overall yield of the Supercritical Fluid Extraction (SFE) at a 5% level of significance. In other words, the effect of pressure depends on the temperature and vice versa. It was noted that the process optimization of yield extraction with sub and supercritical CO2 of the tamarind seed, sweet variety, is found in regions of high pressure and high temperatures or high pressures and subcritical temperatures. Therefore, CO2 extractions combined with ethanol were carried out at 250 bar and 50
C. 3.5. Influence of temperature on the overall yield According to Carvalho-Junior et al. (2003) there are two major effects of temperature on solubility and yield: the solvent specific mass and vapor pressure of the solute. It was observed in this experiment that the increased specific mass positively influenced the yield of the extract, under the pressure of 250 bar where the extraction took place at 20
C (subcritical) and 50
C (supercritical), and the 0.9635 g/cm3 and 0.8349 g/cm3 specific mass, respectively, extraction yield increased with increase specific mass. The vapor pressure of the solute increases with an increase in temperature, but in this experiment, the solvent specific mass was the dominant effect on the extraction process. 3.6. Influence of pressure on overall yield The extract yield increased from 1.62% to 2.62% (around 60% more) at 20
C when the pressure was increased from 100 bar to 250 bar, respectively. This can be explained by the increased specific mass of the solvent with pressure, in other words, there is an increase in the solvating power of CO2 (solubility) with the specific mass (Brunner, 1994). This behavior is commonly published in several papers with supercritical extraction with pure carbon dioxide (Benelli et al., 2010; Oliveira, 2010; Andrade, 2011). 3.7. Use of ethanol as a solvent in the extraction process Ethanol has been used as co-solvent in several studies on the supercritical extraction of antioxidants in small concentrations. Therefore, in this work, we performed extractions combining carbon dioxide with an increasing in the ethanol concentrations, from 10% to 100% by mole fraction. The aim was to evaluate the yield and the amount of polar compounds that could be extracted from the tamarind seed, sweet variety. The yields obtained with CO2 combined with ethanol are presented in Table 2. The highest extraction yields with ethanol were obtained when using only ethanol or when combined with CO2, in a proportion of 3:1 since there is no significant difference at the 5% level by Tukey's

test. With the use of ethanol the extracts obtained had a characteristic of yellow oil mixed with a reddish brown liquid. The significant increase in the yields from 2.50% to 21.44% by using ethanol as a solvent can be explained by the increase in solubility of polar compounds from tamarind seeds in the mixture CO2 þ ethanol. However, increasing the solubility decreased the selectivity of the process and vice versa. It is noticed that there was no significant difference in the extraction yield of the mixture CO2 þ ethanol at concentrations of 10% and 25%. This may be because the mole fraction of ethanol increase, and hampered the extraction of nonpolar compounds, whereas ethanol is polar and preferentially extracts polar compounds (Oliveira, 2010). The concentration of ethanol from 50% to 75% increased the yield significantly. However, there was no significant difference in the antioxidant activity, confirmed by DPPH method. Therefore, the increase in solubility reduces the selectivity of the extraction process and vice versa. Comparing the highest yields (20.27%, 21.44% e 20.69%) (Table 2) obtained from the extraction with CO2 þ ethanol or by Soxhlet there is no significant difference at the 5% level according to the Tukey's test. 3.8. Mathematical modeling The yield curve of supercritical extraction with pure CO2 at 250 bar and 50
C was chosen to represent the modeling of the experimental data since at this condition the highest yield was obtained and was also used in the extraction with ethanol. According to Fig. 2, the experimental data was best fitted by
(1994) modified Martínez et al. (2003) model, followed by Sovova by Martínez and Martínez, 2008, as a result of the mean square error (MSE) 1.72  105 and 0.0006, respectively. The model of Esquível et al., 1999, describes the effects of thermodynamic and mass transfer, not being possible to evaluate the influence of different mechanisms of mass transfer. The modeled curve has the shape of a hyperbola, not similar to the experimental data (MSE of 0.0087) characteristics of systems in which the solute has an easy contact with the solvent and thus readily extracted (Sousa et al., 2005). The curve of the Crank (1975) model was not similar to the experimental data, despite the MSE 0.0075, occurring deviations from the experimental points. This can be explained by deviations

Table 3 Effects of temperature (T) and pressure (P) on the yield of SFE tamarind seeds, sweet variety, evaluated by ANOVA test. Effect

SQ(1)

F(2)

p(3)

Pressure (bar) Temperature (K) Interaction P  T Error SQ(1) total

3.203026 0.692390 0.624416 0.022083 5.547170

290.0941 62.7089 56.5526 e e

0.003429 0.015557 0.017227 e e

SQ(1) ¼ sum of squares; F(2) ¼ test statistic; p(3) ¼ probability.

Fig. 2. Curve ESC seed of tamarind, sweet variety, modeled 250 bar/323.15 K, with a flow rate of 3.85 ± 0.5 g/min.

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

from the experimental points, positive in CER step and negative in FER step, leading to a false impression of a good fit of the model. The Sovov
a (1994) modified by Martínez and Martínez, (2008) model presents good fits to the experimental data during periods of constant rate, describing the phenomenon, mass transfer, which occurs in supercritical extraction. This indicates that the preliminary determination of solubility was adequate (Gupta and Shim, 2007). 3.9. Determination of total phenolic compounds (TPC) The larger amount of antioxidant compounds, extracted with CO2, was at 250 bar and 50
C with 25.19 ± 0.6 mg GAE/gextract. It was considered statistically equal to the amount found in the condition of 250 bar and 20
C, according to Tukey's test at 5% significance. According to Table 2 both, pressure and temperature, had a significant influence on the extraction of these compounds. In the presence of two solvents, the greatest amount was found in the condition with 75% of ethanol (420.8 ± 2.5 mg GAE/gextrato). This behavior can be explained by the high polarity of ethanol, which contributes to the extraction of phenolic compounds, since they possess chemical characteristics similar to ethanol, facilitating their extraction. Although the extraction yield with 75% ethanol was very similar to the Soxhlet extraction, the quality of the extract is different, being found 420.8 mg GAE/gextrato and 387.43 mg GAE/gextrato, respectively. Probably the percentage of CO2 and the temperature and pressure conditions favoring extraction, since ethanol is at supercritical temperature and subcritical pressure, 6.06 MPa and 241
C (Nimed, 2009). The highest amount of extracted phenolic compounds occurred at 250 bar and 50
C with only ethanol, 471.05 mg ± 3.2 EAG/gextrato, being approximately 22% higher than that extracted by the Soxhlet method (387, 43 ± 1.4 mg GAE/gextrato). This means that the conditions of pressure and temperature influenced positively in obtaining these compounds, being necessary to conduct a study on the economic viability of the investments required for the use of

121

these conditions, since the operation of an industrial plant and high-pressure temperature stills very expensive compared to conventional extraction. Comparing the antioxidant activity of butyl hydroxytoluene (BHT), a synthetic compound with known antioxidant activity, identified by Benelli et al. (2010), with the extracts of seed of tamarind, variety sweet, according to Table 2, one observes an amount of phenolic compounds significantly higher in extracts obtained with 25%, 75% and 100% ethanol, both at supercritical extraction method and by Soxhlet, demonstrating the potential of seed of tamarind, variety sweet, for future testing in the food industry as a natural antioxidant, considering, of course, the sensory characteristics of the product. 3.10. Determination of antioxidant activity by DPPH Table 2 shows the values of the antioxidant activity on the highest tested concentration of the sample (500 ug/mL) in AA % and EC50. According to Jesus (2010), the EC50 is the effective concentration to reach 50% of antioxidant activity. Therefore, the lower the EC50 value, the lower the concentration required for 50% of antioxidant activity (AA) according to the DPPH method, and the higher the antioxidant activity of the system under study. According to the results, the samples that had the highest antioxidant activities were extracted with ethanol at a concentration of 25%, 50%, 75% and 100% with 97,18%, 97,77%, 97,61% and 97,79%, respectively, at 250 bar and 50
C. These results and that one obtained by Soxhlet also do not show significant difference at level of 5% by Tukey's test. This can be attributed to the polarity of ethanol that solubilized the polar compounds, among them phenolic compounds. This antioxidant activity was greater than of the synthetic antioxidant BHT (89.7 ± 0.5 ug/mL), making the tamarind seeds extract, sweet variety, an antioxidant to be tested in foods. For the samples extracted with 75% and 100% of ethanol, as well as Soxhlet, was not possible to calculate the EC50 values. The sample dilutions used to construct the DPPH scavenging curve did not

Table 4 Shows the major compounds found, in which thirty-five chemical compounds in the extracts of sweet tamarind seed were identified. Chemical compounds

Butanoic acid, 3-methyl, ethyl ester Ethylbenzene o-Xyleno 1-Butanol, 3-methyl, acetate Benzene, 1,2,4 trimethyl Decane Oxalic acid, 2-ethyllhexyl hexyl ester Benzene, (1-methylethyl) n-Hexadecanoic acid Pentadecanoic acid 14-methyl Hexadecanoic acid 9,12-Octadecadienoic acid cis 9 e Octadecenoic acid cis Vacenic acid cis-10-Heptadecenoic acid trans 13 e Octadecenoic acid Heptadecanoic acid, 16 methyl Octadecanoic acid 17-Chloro-7 heptadecin Ethyl 9-hexadecenoate Ethyl oleate Pentadecanoic acid, 3-methylbuthyl 10.1Bicyclo tridec-1-ene Tetratriacontane a

Relative area percentage (%).

Retention time (min)

Extraction with CO2 and ethanol 250 bar/50
C

Extractions with pure CO2 100 bar 20
Ca

175 bar 35
Ca

175 bar e 35
Ca

250 bar e 20
Ca

250 bar e 50
Ca

10% ethanola

25% ethanola

50% ethanola

75% ethanola

6.236 6.596 6.890 7.139 12.877 13.057 13.802 13.850 29.133 31.789

2.89 3.98 3.42 2.34 1.53 3.14 0.80 0.65 9.73

3.45 4.67 4.02 2.78 1.70 3.45 1.10 1.52 8.88

3.99 5.32 4.57 3.32 2.00 4.09

5.83 7.69 6.57 4.78 2.82 5.84 1.63 1.79 7.12

4.29 5.52 4.56 3.45 2.08 4.18 1.14 1.42 8.93

2.77 3.78 3.00 2.21 1.37 2.81 0.76 0.82 10.22

5.86 7.69 6.56 4.76 2.83 5.87 1.60 1.96 6.77

7.78 10.29 8.92 6.41 3.86 8.04 2.46 2.80 3.14

3.68 4.49 3.83 2.98

34.824 34.895 34.949 34.983 35.105 35.616 35.724 35.819 35.957 35.994 38.382 43.734 47.925

28.60 0.96

24.07

23.30

22.26

19.41

18.41

1.40 9.05

100% Ethanol

Soxhlet ethanol

3.66

8.25 26.85

17.47 13.06

23.88

28.57

18.31

22.43

14.88

5.72

11.24

23.24

38.27

17.52 25.87 9.00

2.86

3.12

3.27

23.21 34.50 24.31

9.33

2.76

0.92

1.05

1.21

2.18 1.21 2.94 8.66

1.48 2.49

1.60 2.09

1.57 2.35

1.37 2.17

1.23 2.27

1.26 2.63

2.00

1.15 1.84

122

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123

reach the reliable region to calculate such values using the method based on the principle of right triangle, as described by Alexander et al. (1999). Among the samples extracted with pure supercritical CO2, the best antioxidant activity was found in that one obtained at 250 bar and 50
C, still three times lower compared to extracts with ethanol. Luzia and Jorge, (2011) obtained an ethanolic extract of tamarind seed, acid variety, with 75.93% of AA (500 mg/mL) an ethanol extract of tamarind seed, acid variety, and 204.72 of EC50was found (Luzia and Jorge, 2011). Despite the difference in the variety and the method of extraction, these results were similar to the values of this work. 3.11. Fatty acid profile found in the extracts of tamarind seed, sweet variety The main fatty acid compounds (Table 4) found in the supercritical extracts were: 9, 12-octadecadienoic acid (linoleic acid), cis
ico and n-hexadecanoic acid (palmitic acid). 10-heptadeceno Interestingly, the compound cis-9-octadecenoic acid, commonly known as oleic acid, was identified only in the subcritical extraction (at 20
C and 250 bar) and in the extract obtained by Soxhlet. There is a great interest in natural antioxidants sources like seeds, leaves and fruits. Monounsaturated fatty acids, such as linoleic and oleic acids, have been studied as natural antioxidants and proposed as potential substitutes for synthetic antioxidants in specific sectors of food preservation (Miranda, 2010). The major compounds found in extracts with CO2 þ ethanol did not show much similarity in its polarity. For the condition with 10% ethanol, the predominant compounds were 9, 12-octadecadienoic acid (linoleic acid) and n-hexadecanoic (palmitic acid). In the condition of 25% and 75% ethanol the main compounds were 9, 12octadecadienoic acid (linoleic acid) and cis-11-octadecenoic acid. With 50% of ethanol, ethylbenzene and decane. However, for all the extracts, linoleic acid is present in a significant percentage of area. There was no extraction of oleic or stearic acid in the ethanolic extracts but, cis 11-octadecenoic acid, 1,2 benzenedicarboxylic acid, trifluoroacetate of oleyl alcohol and ethyl oleate. These results agree with those reported by Luzia and Jorge (2011), in tamarind seed, acid variety, which reported linoleic acid as the main component, besides palmitic, stearic, and oleic acids, saturated, mono-unsaturated and poly-unsaturated compounds. The cited compounds has an action in the removal of bad cholesterol (LDL). 4. Conclusion The tamarind seed, sweet variety, is a promising source of natural antioxidants. In this work the best yield and antioxidant activity were obtained with 100% of pressurized ethanol at 250 bar and 50
C in comparison with those obtained by Sohxlet. The mathematical model of Martínez et al. (2003), was the one that best fitted the experimental data of the extracts obtained by sub and supercritical CO2 with a MSE of 1.7245  105. It is an important tool for process optimization and scale-up. Also, fatty acids compounds with potential antioxidant activity were found in the extracts such as linoleic and oleic acid, as well as others like cis-10heptadecenoic acid and n-hexadecanoic acid (palmitic acid). Acknowledgments We would like to thank the Foundation for Science and Technology of the State of Pernambuco and CAPES for the financial ~ support, and the IFSERTAO-PE and Tiradentes University for the technical support provided.

References Alexander, B., Browse, D.J., Reading, S.J., Benjamin, I.S., 1999. A simple and accurate mathematical method for calculation of the EC50. J. Pharmacol. Toxicol. Methods 41 (2e3), 55e58. ~o das Te
cnicas de Extraça ~o e do Potencial AntiAndrade, D. A. de, 2011. Avaliaça
(Coffea ar
oxante dos Extratos Obtidos a Partir de Casca e de Borra de cafe abica L.). Dissertation (Masters in Food Engineering). Department of Chemical Engineering and Food Engineering, Federal University of Santa Catarina,
polis. Floriano Angus, S., Armstrong, B., De Reuck, K.M., 1976. International Thermodynamic Tables of the Fluid State: Carbon dioxide. Oxford. Benelli, P., Riehl, C.A.S., Sm^ anica Jr., A., Sm^ aniac, E.F.A., Ferreira, S.R.S., 2010. Bioactive extracts of orange (Citrus sinensis L. Osbeck) pomace obtained by SFE and extract composition. J. Supercrit. Fluids 55, 132e141. Brazil, 2005. Ministry of Health National Health Surveillance Agency. Physicochemical Methods for Food Analysis. MS Publisher, Brasilia, p. 1017 (Brasília). Brunner, G., 1994. Gas Extraction: An Introduction to Fundamentals of Supercritical Fluids and the Applications to Separation Process, vol. 4. Steinkopff, Darmasta dt. Carvalho-Junior, R.S., Moura, L.S., Quispe-Condori, S., Rosa, P.T.V., Ming, L.C., Meireles, M.A.A., 2003. Determination of the global yield isotherms for the system rosemary (Rosmarinus officinalis) þ CO2. In: International Symposium on Super Critical Fluids, 6th, Versailles. Proceedings of the 6th International Symposium on Supercritical Fluids. Versailles:ISASF, pp. 129e137. Correia, J., Michielin, E.M.Z., Ferreira, S.R.S., 2006. Study of Models for Mass Transfer for Supercritical Fluid Extraction Processes. Report (Undergraduate Research PIBIC-CNPq). Department of Chemical Engineering and Food Engineering,
polis. Federal University of Santa Catarina, Floriano Crank, J., 1975. The Mathematics of Diffusion. Claredon Press, Oxford, 85pp. Danielski, L., 2002. Solubility of Marigold Oleoresins (Calendula officinalis L.) and Horsetail (Esquisetum Arvense) in Supercritical CO2, 142 pp. (Masters in Food Engineering). Department of Chemical Engineering and Food Engineering,
polis. Federal University of Santa Catarina, Floriano Esquível, M.M., Bernardo-Gil, M.G., King, M.B., 1999. Mathematical models for supercritical extraction of olive husk oil. J. Supercrit. Fluids 16, 43e58. ~es Com Sistemas So
lidos Granulares, vol. 1. Catalogaca ~o da Gomide, R., 1983. Operaco C^ amara Brasileira de Publicac~ ao de Livros, S~ ao Paulo. Gupta, R.B., Shim, J., 2007. Solubility in Supercritical Carbon dioxide. CRC Press. Jesus, S. P. de, 2010. Recovery of Y-oryzanol from Waste of the Rice Oil Industry Using Supercritical Extraction: Evaluation of Process Parameters and Characterization of the Extracts. Dissertation (Masters in Food Engineering). Department of Chemical Engineering and Food Engineering, Federal University of
polis. Santa Catarina, Floriano Kitzberger, C.S.G., Lomonaco, R.H., Michielin, E.M.Z., Danielski, L., Correia, J., Ferreira, S.R.S., 2009. Supercritical fluid extraction of shiitake oil: curve modeling and extract composition. J. Food Eng. 90, 35e43. Luengthanaphol, S., Mongkholkhajornsilp, D., Douglas, S., Douglas, P.L., Pengsopa, L., Pongamphai, S., 2004. Extraction of antioxidants from sweet Thai tamarind seed coateepreliminary experiments. J. Food Eng. 63, 247e252. Luzia, D.M.M., Jorge, N., 2011. Antioxidant activity, fatty acid profile and tocopherols ^nc. Tecnol. Aliment. Campinas 31 (2), 497e501. of Tamarindus indica L. seeds. Cie Martínez, J., Martínez, J.M., 2008. Fitting the Sovov
a’s supercritical fluid extraction model by means of e global optimization tool. Comput. Chem. Eng. 32, 1735e1745. Martínez, J., Monteiro, A.R., Rosa, P.T.V., Marques, M.O.M., Meireles, M.A.A., 2003. Multicomponent model to describe extraction of ginger oleoresin with supercritical carbon dioxide. Ind. Eng. Chem. Res. 42, 1057e1063. ~o, G.G., Reis, A.S., Santos, T.C., Coube, C.S., Leit~ Mensor, L.L., Menezes, F.S., Leita ao, S., 2001. Screening of Brazilian plant extracts for antioxidant activity by the use of DPPH free radical method. Phytother. Res. 15, 27e30. Mezzomo, N., Martínez, J., Ferreira, S.R.S., 2009. Supercritical fluid extraction of
rsica) almond oil: kinetics, mathematical modeling and scalepeach (Prunus pe up. J. Supercrit. Fluids 51, 10e16. Miranda, C. A. S. F. de, 2010. Effects of Temperature (T) and Pressure (P) on the Yield of SFE Tamarind Seeds, Sweet Variety, Evaluated by ANOVA Test. Dissertation (Masters in Agrochemical Engineering). Federal University of Lavras, Lavras. Nimed, G., 2009. Evaluation of Supercritical Carbon dioxide Solvent and Subcritical Propane in the Extraction of Sunflower Oil. Toledo: UNIOESTE. Dissertation (Masters in Chemical Engineering). Center for Engineering and Exact Sciences,
. State University of West Parana Oliveira, D.A., 2010. Phytochemical and Biological Characterization of Extracts Obtained from Grape Pomace (Vitis vinifera) Varieties Merlot and Syrah. Dissertation (Masters in Food Engineering). Department of Chemical Engineering and
polis. Food Engineering, Federal University of Santa Catarina, Floriano ~o, A.A., Alves, P.R.B., 2011. A Cultura do Tamarindeiro Pereira, P.C., Melo, B., Franza (Tamarindus indica L.). I Semana do Curso de Engenharia Ambiental. Instituto de ^ncias Agr
Cie arias da Universidade Federal de Uberl^ andia e Campus Umarama e MG.
Peschel, W., Sanchez-Rabaneda, F., Diekmann, W., Plescher, A., Gartzía, I.,
nez, D., Lamuela-Ravento
s, R., Buxaderas, S., Codina, C., 2006. An industrial Jime approach in the search of natural antioxidants from vegetable and fruit wastes. Food Chem. 97, 137e150. ~o Paulo). Ramalho, V.C., Jorge, N., 2006. Foto antioxidants. Quím. Nova 29 (4) (Sa

P.M. Cardoso Lima Reis et al. / Journal of Food Engineering 173 (2016) 116e123 Rossi, J.A.J., Singleton, V.L., 1965. Colorimetry of total phenolics with phosphomolybdic phosphotungstic acid reagents. Am. J. Enol. Vitic. 16, 144e158. Sousa, E.M.B.D., Martínez, J., Chiavone-Filho, O., Rosa, P.T.V., Meireles, M.A.A., 2005. Extraction of volatile oil from Croton zehntneri Pax et Hoff with pressurized CO2: solubility, composition and kinetics. J. Food Eng. 69 (3), 325e333.
, H., 1994. Rate of the vegetable oil extraction with supercritical CO2: I. Sovova Modelling of extraction curves. Chem. Eng. Sci. 49 (3), 409e414.

123

Sudjaroen, Y., Haubner, R., Wurtele, G., Hull, W.E., Erben, G., Spiegelhalder, B., Changbumrung, S., Bartsch, H., Owen, R.W., 2005. Isolation and structure elucidation of phenolic antioxidants from Tamarind (Tamarindus indica L.) seeds and pericarp. Food Chem. Toxicol. 43, 1673e1682. Tsuda, T., Watanabe, M., Ohshima, K., Yamamoto, A., Kawakishi, S., Osawa, T., 1994. Antioxidative components isolated from the seed of Tamarind (Tamarindus indica L.). J. Agric. Food Chem. 42, 2671e2674.