Raw material, reagents and standards

Fresh chia seeds were purchased in a city local market (Zacatecas City, Mexico). The seeds were cleaned, ground and sieved to a particle size of 500 μm, packed in hermetic plastic bags and stored in the dark at room temperature (25 ºC). Folin-Ciocalteu reagent (2N), 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), 6-hydroxy-2,5,7,8-tetramethylchromane-2 carboxylic acid (Trolox), gallic acid, potassium persulphate and sodium carbonate were analytical grade (Sigma Chemical Co.; St. Louis, MO, USA); n-hexane and methanol were HPLC-grade (JT Baker).

Conventional extraction

Chia seeds flour (5 g) and 50 mL of n-hexane or methanol (1:10 w/v) was thoroughly mixed for 5, 10 or 15 min using an Ultraturrax homogenizer (T-25, IKA, Staufen, Germany) and then centrifuged (2701 x g; 4 °C) for 10 min. The supernatants were stored at 4 ºC until their analysis (24 h).

Ultrasound-assisted extraction (UAE)

Ultrasonic device set-up

An ultrasonic probe device (UP400S, Hielscher, Teltow, Germany) working under controlled temperature conditions (25 ºC) was used at 24 kHz, with maximum power output of 400W, and an emitter surface of 3.8 cm² (Figure 1). The ultrasonic probe was immersed 1.5 cm into the solution and the vessel was kept in the dark in order to avoid potential light-induced damages of the extract.

Figure 1 Experimental set-up for ultrasonic assisted extraction of chia (Salvia hispanica L.) seeds phenolic compounds. A: Portable Computer; B: ultrasonic probe system; C: jacketed extraction vessel; D: chiller. 

The effective ultrasonic power transferred into the medium (P) was determined using a calorimetric procedure (^{Hussam et al., 2013}) (Equation 1) registering the temperature rise each second, for the first 3 min of ultrasonic application. P was calculated as:

(1)

where M (kg) is the solvent mass, Cp (J kg^-1 °C) is the heat capacity of solvent and dT dt^-1 is the slope of the logged temperature-time curve. Determinations were carried out in triplicate.

Ultrasonic extraction

Experiments were carried out using the same chia seed flour (solvent ratio as before). Extract was placed in a jacketed extraction vessel and submitted to ultrasonic processing for 5, 10, and 15 min at amplitudes corresponding to 50, 75 and 100 % of power output (400 W). After extraction, the mixture was centrifuged (2701 x g 4 ºC) for 10 min. The supernatant obtained was stored (24 h) in opaque vials at 4 ºC until analyzed.

Determination of total phenolic compounds (TPC)

Total phenolic content (TPC) was determined using the Folin-Ciocalteu method described by ^{Hussam et al. (2013)} with some modifications. Briefly, 250 μL of extract was mixed with 15 mL deionized water and 1.25 mL of Folin-Ciocalteu phenol reagent. After 5 min, 3.75 mL of Na₂CO₃ (7.5 %) and bring to 25 mL with deionized water. Absorbance was measured at 765 nm in a spectrophotometer UV-Vis (Thermo Scientific 10S, Thermo Fisher Scientific Inc, USA). Response variable was the TPC in the extracts expressed as mg gallic acid equivalents (GAE) per sample (100 g) using a regression equation and a gallic acid calibration curve. At least three replicates were made for each extract.

Antioxidant capacity (AC) and ABTS ^+⋅ scavenging ability

ABTS scavenging ability was determined as described by ^{Fu et al. (2011)}. An ABTS radical cation (ABTS^+⋅) was generated by reacting an ABTS aqueous solution (7 mmol L^-1) with K₂S₂O₈ (2.45 mmol L^-1, final concentration) in the dark for 16 h, adjusting its absorbance at 734 nm to 0.700±0.1 with ethanol. After addition of 900 mL of diluted ABTS^⋅+ solution to 100 mL of extract, the absorbance reading was taken exactly 1 min after initial mixing. The percentage inhibition of absorbance at 734 nm is calculated and plotted as a function of concentration of antioxidants and of Trolox for the standard reference data. The Trolox Equivalent Antioxidant Capacity (TEAC) was subsequently calculated. Results were expressed as mg Trolox 100 g^-1. Experiments were run in triplicate.

Experimental design and statistical analysis

The experimental design was factorial (3x3x2) and three replicates. All values expressed are means ± standard deviation. The data were used for to one-way and multifactorial ANOVA, and the differences between means was determined by Tukey's test (p≤0.05) using Statgraphics^® Centurion XV (Statpoint Technologies Inc., Warrenton, VA, USA). Pearson's correlation test was carried out to determine the linear correlations between TPC and AC using the same software.

Conventional extraction

A significant difference (p<0.05) was observed between TPC of methanolic extracts and those obtained with hexane at all extraction times, with an average increment of 64.56 %; thus, phenolics in chia are thought to be polar in nature (Figure 2a). These results are in agreement with those of ^{González-Jiménez et al. (2010)}. In ethanol-extracted chia samples, ^{Reyes-Caudillo et al. (2008)} obtained TPC values of 88.00 mg per 100 g of sample, while ^{Marineli et al. (2014)} reported TPC of 94 mg 100 g^-1. A significant (p<0.05) difference in AC of methanol-extracted or hexane-extracted samples were observed (Figure 2B).

Figure 2 Effect of the type of solvent (n-hexane or methanol) at three different conventional extractions times on the average (A) TPC and (B) AC from samples of chia seeds. Average ± HSD (p≤ 0.05). 

In order to compare the effect of solvent, average values of total phenolics and antioxidant activity were obtained from all values by conventional extraction. Methanolic extracts showed an AC of 8.27±0.15 mg Trolox 100 g^-1 of chia (average value), whereas hexane extracts had an AC of 0.44±0.05 mg Trolox 100 g^{- 1} of chia (average value), which is consistent with TPC results. Hexane extracts are rich in highly-unstable polyunsaturated fatty acids (PUFA) (^{González-Jiménez et al. , 2010}). The oxidation-prone PUFAs would be responsible for such low AC values. These results agree with those obtained by ^{González-Jiménez et al. (2010)}, who reported that oil fractions of chia seeds showed a minor AC related to the low TPC obtained with hexane and the high PUFAs content. Results showed (Figure 2A) a significant difference (p≤0.05) between methanol extractions for 5 and 10 min in comparison with the 15 min extraction. Meanwhile, for hexane samples, a significant (p≤0.05) increase of TPC was observed in 5-min hexane-extracted samples, whereas AC (Figure 2B) did not exhibit a significant influence (p>0.05) of extraction time for both solvents used.

There is an influence of solvent polarity in the extraction process. Thus the higher TPC and AC were obtained in samples with methanol, while for the conventional extraction time, the TPC was decreasing in methanol samples and increasing in hexane samples, and the AC remained constant in both samples.

Ultrasonic assisted extraction

The effective ultrasonic power was in the range from 29.002 to 37.534 W. As expected, the ultrasound applied into the medium increased almost linearly (R² _Methanol = 0.8833; R² _n-Hexane = 0.9972) with supply of electric power to the transducer (Table 1).

Table 1 Ultrasonic power (W) applied to the medium as function of the percentage of the total electric power (400 W) and time (3min) of ultrasonic application. 

Data are means ± standard deviation (n=3)

Similar to conventional extraction, in order to compare the effect of solvent, average values of total phenolics and antioxidant activity were obtained from all values by UAE. Net ultrasonic power applied was always higher for hexane than methanol; this was attributed to differences in heat capacity. UAE results (Table 2) showed a significant difference (p≤0.05) on TPC and AC between methanolic extracts (149.95±20.39 mg of GAE 100 g^{- 1} and 9.90±0.05 mg Trolox 100 g^-1 of chia, average value) and hexane (60.31 ±5.99 mg of GAE 100 g^-1 and 1.11 ± 0.38 mg Trolox 100 g^-1 of chia, average value). From other studies (^{Cárcel et al., 2007}; ^{Hussam et al., 2013}), the ultrasonic power applied was considered one of the key factors affecting phenolics extraction efficiency. A higher ultrasonic power applied generates more intense cavitation which facilitates solvent penetration into the matrix and increases the extraction efficiency (^{Priego and Luque, 2004}).

Table 2 Influence of ultrasonic assisted extraction process parameters on the TPC and AC of Chia seeds with two different solvents (methanol and hexane). 

Data are means ± standard deviation (n=3). Different lower case letters in the same cell and the same column indicate significant differences between amplitudes at a time of extraction (p≤0.05); while upper case letters in a different cell within a single column indicate statistically significant difference between times for the same amplitude (p≤0.05)

Best extraction parameters of TPC were obtained at 100 % ultrasonic power for 15 min using methanol (194.06± 11.11 mg of GAE 100 g^-1 of chia); meanwhile, no influence (p>0.05) of ultrasonic application was observed for 5 and 15 min hexane extraction (68.14± 0.14 and 66.13± 2.14 mg of GAE 100 g^-1 of chia, respectively). Again, this was attributed to non-affinity of hexane polarity towards hydrophilic compounds of chia, as well as with the irradiation power, which might reduce the efficiency of ultrasonic energy transmitted into the medium due to an increase in the bubble numbers in solvent during cavitation (^{Filgueiras et al., 2000}; ^{Zhao et al., 2007}). As expected, methanolic extracts exhibit the highest AC ((9.90±0.05 mg Trolox 100 g^-1 of chia) (Table 2); however, no significant differences in ultrasonic power applied (p> 0.05) were observed. Besides, there was correlation between TPC and AC of chia seeds for both solvents used (R=0.939; n=66; p=p 0.000) (Figure 3).

Figure 3 Relationship between the antioxidant capacity (mg of Trolox 100 g^-1) and total phenolic compounds (mg of GAE 100 g^-1) of chia seed extracts with two different type of solvent (n-hexane and methanol).