Optimization and stabilization of the antioxidant properties from Alkanet (Alkanna tinctoria) with natural deep eutectic solvents

2.1 Plant material and chemicals

A. tinctoria root was collected from Iğdır province, Turkey. The root was cleaned, packed into sterilized brown bottles and kept at 4 °C for further use. Before analysis, the root was powdered by a disintegrator (sinbo, coffee and spice grinder, SCM 2934) and sieved. Distilled water was purified by a Millipore-Q system (Millipore Billerica, Massachusetts, USA).

Methanol (≥99.8%), 2,2- diphenyl-1-picrylhydrazyl (DPPH), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ, ≥99.0%), Trolox (97%), sodium nitrite (99–100.5%), hydrochloric acid (36.5–38%), (−)- epicatechin, gallic acid (≥99.0), lactic acid (≥85%) and sodium carbonate (99.5–100.5%) were purchased from Sigma-Aldrich. Ethanol (≥99.9%) was bought from Isolab. Folin-Ciocalteu reagent, aluminium chloride, iron (III) chloride, iron sulfate heptahydrate (≥99.5%) and formic acid (98–100%) were brought from Merck. Sodium acetate anhydrous (≥99.0%), glacial acetic acid (99.5%), potassium chloride (≥99.0%) and Sodium hydroxide (≥97.0%) were taken from Carlo erba.

2.2 NaDES preparation

The NaDESs were prepared following the procedures detailed in El Kantar et al. (2019) and Wei et al. (2015) with some changes. The NaDESs were synthesized as molar ratio of two different components consisting of hydrogen bond donor (HBD) and hydrogen bond acceptor (HBA), followed by the addition of 20% of distilled water. Sodium acetate was used as HBA and formic acid and lactic acid were used as HBD (Xu et al., 2015; Pal and Jadeja, 2020). In this study, NaDES components were placed in reaction flask by heating at 75 °C with constant stirring for 2 h30 min until a homogeneous liquid was obtained.

2.3 Rheology

The rheological characteristics of NaDESs were monitored based on the viscosities. The viscosity of NaDESs was determined with a Rheometer (Buchi, CH-9230 Flawil 1, Switzerland) fitted with a parallel geometry with 20 mm of diameter and gap 1 mm. The measurements were carried out as described in Aroso et al. (2017). All the measurements were performed in triplicate at 20 °C.

2.4 Fourier transformed infrared (FTIR) analysis

FTIR analysis of NaDESs and extracts was carried out using a FTIR Spectrometer ((Perkin Elmer, Spectrum-Two, USA, PEService 35). Diamond lens attenuated resistance was used. The spectrometer was adjusted in resolution 4 and by selecting the Norton-Beer (N-B) strong apodization function. The range of all spectra was between the wavenumbers of 4000 and 400 cm⁻¹. Before every spectrum, a background reference was taken using an empty cell to ensure no interferences. Then, spectrum intensity was transformed into relative transmittance, %T.

2.5 Extraction

The extraction of phytochemical compounds with antioxidant activity was achieved in both conventional and NaDESs media using an ultrasound. The ultrasound-assisted extractions (UAE) were performed in a sonication water bath (WUC-A03H, ultrasonic cleaner set, daihan scientific Co., Ltd). Distilled water, 70% ethanol and 80% methanol were considered in present work as conventional solvents due to their high extractability of polyphenols. 0.25 g of comminuted A. tinctoria root was mixed with 10 mL of conventional solvents and NaDESs. The sonication was carried out following a modified previous method (Zhao et al., 2020). The mixture was ultrasonicated in an ultrasonic bath at room temperature (25 °C) for 20 min. The samples were then filtered through Whatman filter paper No.1 thrice.

2.6 Determination of total phenolic content (TPC)

The TPC of A. tinctoria extracts was determined using a previous Folin-Ciocalteu method with modification (Maran, 2013; Zannou et al., 2020). Briefly, an aliquot of 150 μL of the diluted extract was mixed 750 mL of 10% Folin-Ciocalteu reagent. The mixture was shaken for 1 min before adding 600 mL of 7.5% sodium carbonate solution. The mixture was shaken again and placed in the dark for 2 h before reading absorbance. The absorbance was read at 760 nm in an UV-spectrophotometer (Thermo Spectronic) and the TPC was calculated from a calibration curve using gallic acid as a standard. The results were given as mg gallic acid equivalent (GAE) g⁻¹ dw.

2.7 Determination of total flavonoid content (TFC)

The TFC was determined using a modified protocol (Hossain and Shah, 2015; Lakka et al., 2019). 1 mL of the diluted solution was combined with 0.3 mL of 5% NaNO₂ and left to stand for 5 min, followed by the addition of 0.5 mL of 5% AlCl₃. The mixture was kept for 6 min before adding 0.5 mL of 1 M NaOH. After 10 min, the absorbance was read at 510 nm. The TFC was estimated based on a calibration curve using epicatechin as standard. The results were given as mg epicatechin equivalents (ECE) g⁻¹ dw.

2.8 Determination of the DPPH radical scavenging activity

The DPPH radical scavenging was determined using a modified method (Hossain and Shah, 2015; Singh et al., 2015). Briefly, an aliquot of 50 μL sample was added with 1 mL DPPH solution (0.06 mM in 80% methanol). The mixture was shaken and left to stand in dark for 1 h until the reaction completed. Thereafter, the absorbance at 517 nm was recorded. The DPPH solution was used as control. The reduction ratio of DPPH was determined with the following equation:

2.9 Determination of ferric reducing antioxidant power (FRAP)

FRAP assay was performed according to the procedure of Shang et al. (2019) and Singh et al. (2015). Briefly, an aliquot of 50 µL volume of sample was mixed with 950 mM of FRAP solution constituted of 100 mM acetate buffer: 10 mM FeCl₃: 10 mM TPTZ (2,4,6-tripyridyl-s-triazine). The assembly was shaken for about 5 min and the absorbance was read at 593 nm against a blank. The FRAP values of the extracts were calculated from the calibration curve using FeSO₄ as a standard. The results were given as mmol FeSO₄ equivalents (mmol ISE g⁻¹ dw).

2.10 Optimization plan

The optimization parameters of the NaDES were examined systematically using response surface methodology based on the three-level Box-Behnken design (Design expert software 11.0). The experimental design was carried out with three independent variables of X₁ (SAFA molar ratio), X₂ (water content) and X₃ (solvent-to-solid ratio). The actual and coded values of the independent variables were presented in Table 1. The association of the molar ratio of sodium acetate (1) to formic acid (1, 2 and 4), water content (10%, 45% and 80%) and solvent-to-solid ratio (10:0.25, 17:0.25 and 25:0.25 mL/g) were independent variables chosen for UAE. These variables were regrouped in 15 experimental points including three replicate at the central point. Total phenolic content, total flavonoid content and DPPH radical scavenging activity were investigated as the responses (Y) for the design experiment. The experimental points along with responses were given in Table 2. The analyses were performed in duplicate. The experimental data were fitted to the following quadratic polynomial model:

2.11 Thermal stability

5 mL of A. tinctoria extracts were put into the brown bottles with screw caps and placed in a preheated water bath at 80, 60 and 40 °C. The bottles were removed from the water bath after 2 h and cooled to room temperature. The first-order reaction rate constant (k) and activation energy (Ea) were used to express the kinetic modeling of degradation of TPC, TFC, DPPH radical scavenging and FRAP during the thermal treatment of A. tinctoria in NaDES (Dai et al., 2014; Olivares-tenorio et al., 2017).

2.12 Data and statistical analyses

The software Design Expert 11.0 (Trial version, Stat-Ease Inc., Mineapolis, USA) was used to design the experimentation along with data analysis. ANOVA was used to determine the statistical relationship between factors. The adequacy of the models obtained was ascertained by screening the R², adjusted R², coefficient of variation (CV) and the value of Fisher’s test (F-value). The significance of the models and regression coefficients were measured at p < 0.05. The relationship between independent variables and responses was checked by 3D graphics. The optimum conditions were determined according to the desirability function. One-way statistical analyses were carried out by ANOVA with post-hoc Duncan’s test using SPSS (version 21). The significance of the results was assessed at p ≤ 0.05.

3.1 Characterization of NaDESs

The viscosity of NaDESs destined for extraction phytochemical compounds is a key element. The very high viscosity would adversely impact the extraction yield of phytochemical compounds since it hinders the mass transfer to be appropriately achieved (Wei et al., 2015). In the present study, two NaDESs composed of hydrogen bond acceptor (sodium acetate) and hydrogen bond donor (lactic acid and formic acid) in 1:2 M ratios were formulated (Xu et al., 2015; Pal and Jadeja, 2019; Zhao et al., 2020). The viscosity of these NaDESs was tailored by adding 20% of distilled water to enhance the extraction abilities (Obluchinskaya et al., 2019; Rajha et al., 2019; Saha et al., 2019). The NaDES sodium acetate:lactic acid was encoded SALA₁₂ while sodium acetate:formic acid was SAFA₁₂. The viscosity determined for SAFA₁₂ was 46.25 ± 6.83 m.Pa while SALA₁₂ exhibited a viscosity of 56.49 ± 7.68 m.Pa (Table 3). The highest viscosity was detected with SALA₁₂ and the lowest with SAFA₁₂.

The pH is an essential factor influencing the performance of NaDESs (Duan et al., 2016; Chakroun et al., 2019). It exerts a crucial role in chelate formation and stability, affecting the recovery of targeted analyte (Zounr et al., 2018). The pH of SALA₁₂ and SAFA₁₂ were assigned in Table 3. As can be seen, the lowest pH was detected with SAFA₁₂ (pH 3), while the highest pH was found with SALA₁₂ (pH 3.41). Similarly, García et al. (2016) have reported pH 2 and pH 0.5 in choline chloride:lactic acid (1:2) and choline chloride:malonic acid, respectively.

The FTIR spectra (Fig. 1) were used to identify the functional groups and feature reactions that occur between HBA (sodium acetate) and HBD (lactic acid and formic acid). To the best of the authors’ knowledge, no study describing the FTIR frequencies of the studied NaDESs has been published so far. As shown in Fig. 1a, the FTIR analysis of SAFA₁₂ revealed that the peak at 3418.98 cm⁻¹ was associated to O— H stretching frequency of water denoting the effective presence of water in the medium. The peak at 2932.93 cm⁻1 could be associated to the stretching frequency of C— H of sodium acetate (Jones and Mclaren, 1954) and formic acid. The peak at 2540.17 cm⁻¹ was assigned to O— H stretching frequency of formic acid since the peaks ranged between 2500 and 3000 cm⁻¹ are due to the O— H stretching of the acid component. The strong frequencies obtained at 1705.71 and 1575.23 cm⁻¹ were related to the stretching C⚌O of formic acid and sodium acetate, respectively. The rock C—H of formic acid was obtained at 1375.18 cm⁻¹. Also, 1375.18 cm⁻1 could be considered as the stretching frequency of C—O of sodium acetate. The lower range of the region below 1200 cm⁻¹ generally represents different kinds of C—H, C—O and CH₃ vibrations (rocking, deformation and stretching) (Tripathi et al., 2015). These findings indicated the homogenization of the mixture and the formation of a strong network between the functional groups of NaDES sodium acetate:formic acid were effective.

Close

Fig. 1

FTIR spectrum of (a) SAFA₁₂ and (b) SALA₁₂.

Export to PPT

Fig. 1b presented the FTIR spectrum of SALA₁₂. As can be seen, the peak at 3380.30 cm⁻¹ was associated to O—H stretching frequency of water and lactic acid. This peak was found to be larger than O—H stretching peak obtained in Fig. 1a, since it rose from water and lactic acid. The peak at 2989.80 cm⁻1 was the stretching frequency of C—H of sodium acetate (Jones and McLaren, 1954) and lactic acid. The peak at 2569.00 cm⁻¹ was assigned to O—H stretching frequency of lactic acid. The strong frequencies obtained at 1708.05 and 1581.05 cm⁻¹ were assigned to the stretching C⚌O of lactic acid and sodium acetate, respectively. The C⚌O stretching band of lactic acid shifted to 1645.60 cm⁻¹. The band at 1454.04 cm⁻¹ was for CH₃ deformation present in sodium acetate and lactic acid, while the band at 1411.39 cm⁻¹ was C—O stretching of sodium acetate. The peak frequencies of 1370.27 and 1248.02 cm⁻¹ were related to the CH bend and OH bend, respectively. The bands ranged from 1200 to 950 cm⁻¹ explained the stretching modes of C—C and C—O functional groups of lactic acid. These results explained the effectiveness of the presence of NaDES components in the mixture and the formation of a strong network between the functional groups of NaDES sodium acetate:lactic acid.

3.2 Evaluation of NaDESs efficiency

There are many intrinsic characteristics notably diffusion, surface tension, viscosity, density, polarity, physicochemical interactions and solubility (Bi et al., 2013; Wei et al., 2015) which influence the extraction efficiency of NaDESs. The results showing the extraction performance of SAFA₁₂ and SALA₁₂ together with conventional solvents were presented in Table 3 and Fig. 2. The values of TPC, TFC, DPPH radical scavenging and FRAP values were ranged from 33.12 to 175.97 mg GAE/g, 1.93 to 4.78 mg EGE/g, 74.07 to 130.91 mmol TE/g and 55.47 to 170.43 mmol ISE/g, respectively. These results are in accordance with the values previously reported (Jaradat et al., 2018; Assimopoulou et al., 2004; Assimopoulou and Papageorgiou, 2005). As can be seen, the values of TPC, TFC, DPPH radical scavenging and FRAP were found to be higher in SAFA₁₂ and SALA₁₂. Similarly, the NaDES sodium acetate:lactic acid has provided the best extraction performance of phenolic compounds from mango (Pal and Jadeja, 2019). The conventional solvents such as water, 70% ethanol and 80% methanol displayed the lowest extractability performance compared to the NaDESs. The hydrophilic NaDESs have been revealed to be appropriate extraction solvents compared to the conventional solvents (El Kantar et al., 2019; Saha et al., 2019). The high extractability efficiency of SAFA₁₂ and SALA₁₂ could be linked with the multiple hydrogen-bonding networks generated during their formulation and the introduction of water to tailor their viscosity. The acid-based NaDESs have been reported to be adequate for the extraction of phenolic compounds, since they conferred strong H-bonding interactions amongst components and presented the low viscosity (Lakka et al., 2019; Y. Liu et al., 2019; Zhou et al., 2018). Furthermore, polyphenols are assigned to a type of hydrogen bond donor since they interact firmly with acetate anions leading to their higher solubility and satisfactory extraction performance (Bakirtzi et al., 2016). As shown in Table 3, SAFA₁₂ displayed the highest extraction yield of TPC (175.97 mg GAE/g), TFC (4.78 mg EGE/g) and DPPH radical scavenging (130.91 mgTE/g) from A. tinctoria root. This might be explained by the difference in the viscosities of NaDESs, since SAFA₁₂ displayed lower viscosity compared to SALA₁₂. Saha et al. (2019) have reported that lowering the viscosity enhances the cavitational process and allows the formation of a stronger H-bonding between the NaDESs and solute, enhancing the extraction yield and capacity. In addition, the higher performance of SAFA₁₂ could be related to the pH, since it displayed a lower pH than SALA₁₂. It has been demonstrated that the most acidic NaDESs have shown the highest yields of bioactive compounds with antioxidant properties (García et al., 2016; Duan et al., 2016). Therefore, the NaDES SAFA₁₂ was chosen for further investigation.

Close

Fig. 2

Extraction performance of NaDESs and conventional solvents.

Export to PPT

3.3 Optimization of NaDES and extraction conditions

The influence of molar ratio, water content and liquid-to-solid ratio was studied in order to optimize the values of TPC, TFC, DPPH radical scavenging and FRAP of A. tinctoria root. A three-level Box-Behnken design (Design expert software 11.0) was used to design the experimentation profile. A total of 15 experimental points with three replication in the central point was generated and investigated. The responses and the coded independent factors of all the experimental points were shown in Table 2.

3.3.2

3.3.2 Effects of independent variables on TPC

The linear and quadratic terms of the response surface methodology models developed for TPC of A. tinctoria root were shown in Table 4. The model showed that the molar ratio was the most significant linear variable having effects on TPC with F values of 32.10 (p < 0.0002), which means that it is needed for complete understanding of the behavior of this response. The linear values of water content and solvent ratio had no significant effects on TPC. The interaction of the linear units of water content and molar ratio had significant impacts on TPC with F value 19.28 (p < 0.0014). Likewise, only the quadratic term of water content had the most significant impact on TPC with F values of 10.17 (p < 0.0097). The 3D response surface graphics (Fig. 3) were generated to interpret the effects of the combinations of water content and molar ratio, of water content and solvent ratio and of molar ratio and solvent ratio on TPC. As can be seen, the couple of variables water content and molar ratio had significant effects on TPC and occurred a rapid augmentation of this response when increased and this behavior of TPC was most accentuated when the molar ratio was incremented. The availability of H-bond donors elevated when the molar ratio increased, leading to the formation of a stronger hydrogen network between NaDES components and solute. Such phenomenon facilitates the dissolution of phenolic compounds in NaDESs, improving the extraction efficiency (Peng et al., 2015; Saha et al., 2019). The increase in water content demonstrated a beneficial effect on TPC as the highest value was obtained 80% of water content (Fig. 3). Similar reaction has been observed during the optimization of polyphenols from olive leaf (Alañón et al., 2020), as the salt-based NaDES are too anhydrous forming a plastic solid at low water content. Lowering the viscosity by adding water enhanced the efficiency of sodium acetate:formic acid NaDES for the extraction of polyphenols of A. tinctoria. The quantity of water required to extract the maximum phenolic compounds depends on the intrinsic characteristics of NaDESs, the raw material and the targeted phytochemical compounds. For example, 10, 20, 25, 30, 50 and 60% of water content in various NaDESs have been reported as the appropriate required water to perform higher efficiency the extraction of phenolic compounds many plant materials (Barbieri et al., 2020; Buldo et al., 2019; Cui et al., 2015; Jancheva et al., 2017; Ozturk et al., 2018; Wei et al., 2015). Nonetheless, some quantity of water should not be overpassed (Alañón et al., 2020) else, the eutectic structure of the mixture would break down (Dai et al., 2013). The solvent ratio was almost indifferent to the response surface carried out for TPC, indicating a positive character in economic terms. Moreover, larger solvent volumes negatively impact the efficiency and the economic feasibility, generating unnecessary byproducts (Zhang et al., 2014).

Close

Fig. 3

Effect of independent variables on TPC.

Export to PPT

3.3.3

3.3.3 Effects of independent variables on TFC

The linear and quadratic terms of the response surface methodology models developed for TFC of A. tinctoria root were given in Table 4. The developed model showed that the linear term of molar ratio was the most significant linear unit having effects on TFC with F value 20.13 (p < 0.0006). The linear values of water content and solvent ratio had no significant effects on TFC. No quadratic term had significant on the TFC since all the quadratic terms displayed p-values higher than 0.05. The response surface plot showing the effect of molar ratio on TFC was presented in Fig. 4. As can be seen, the increase in molar ratio linearly generated an increase on the response TFC. The extraction of flavonoid compounds with NaDESs was reported to be strongly dependent to the extraction conditions, NaDESs components, molar ratio, solvent ratio and water content (Bajkacz and Adamek, 2017; Wei et al., 2015; Xiong et al., 2019; Zhuang et al., 2017).

Close

Fig. 4

Effect of independent variables on TFC.

Export to PPT

3.3.4

3.3.4 Effects of independent variables ratio on DPPH radical scavenging

The linear and quadratic terms of the response surface methodology models developed for DPPH radical scavenging of A. tinctoria root were presented in Table 4. The analysis of the model revealed that linear units of molar ratio and solvent ratio exhibited significant effects on DPPH radical scavenging with F values 22.25 (p < 0.0015) and 7.43 (p < 0.0260), respectively. The interaction of the linear units of water content and molar ratio had significant impacts on DPPH radical scavenging with F value 7.44 (p < 0.0259). Furthermore, the quadratic terms of water content and molar ratio were found to have important impacts on DPPH radical scavenging with F values of 7.40. (p < 0.0262) and 12.29 (p < 0.0080), respectively. The 3D response surface plots showing the behavior of DPPH radical scavenging against independent variables were given in Fig. 5. The couple variables molar ratio and water content showed a pronounced positive action occurring the augmentation in DPPH radical scavenging when increased. However, the pair variables of solvent ratio and water content and of molar ratio and solvent had moderate increasing effects independently of the linear variable solvent ratio. Similar behavior has been observed with TPC, suggesting that the DPPH values are dependent to the concentration of phenolic compounds at a certain degree. Such observations have been remarked for many foods and wastes (Makris et al., 2007; Pal and Jadeja, 2020, 2019).

Close

Fig. 5

Effect of independent variables on DPPH radical scavenging.

Export to PPT

3.4 Characterization of SAFA₁₄ and the optimum extract

The SAFA₁₄ and the extract obtained at the optimum conditions were characterized to determine their rheological properties and structural features. For the rheological characterization, the viscosities of these samples were measured. The viscosities of SAFA₁₄ and extract were determined as 8.6 ± 14.12 and 7.53 ± 3.26 m.Pa, respectively. As can been, the viscosities of SAFA₁₄ and extract were very close, explaining the complete dissolution of the sample with NaDES during the extraction process. Thus, this resulted in the augmentation of the extraction yield. Furthermore, the viscosity of the extract was slightly low due to the application of temperature (25 °C) during the sonication. For the evaluation of structural features, the analysis FTIR spectra revealed that the FTIR spectrum of SAFA₁₄ and SAFA₁₂ were very similar (Fig. 6), denoting that both spectra possessed the same functional groups and then same components. However, the intensities of the frequencies at 3418.98, 1705.71 cm⁻¹ were higher in SAFA₁₄ when compared with SAFA₁₂, indicating there are more water and acid in SAFA₁₄. The FTIR spectra of SAFA₁₄ and extract were shown in Fig. 7. As can be seen, the extract was a replicate of SAFA₁₄. This confirmed the results of rheological properties and indicated SAFA₁₄ inlaid the sample as its own component.

Close

Fig. 6

FTIR spectra of SAFA₁₂ and SAFA₁₄.

Export to PPT

Close

Fig. 7

FTIR spectra of SAFA₁₄ and extract.

Export to PPT

3.5 Thermal stability of TPC, TFC, DPPH radical scavenging and FRAP in NaDES

The thermal effects on TPC, TFC, DPPH radical scavenging and FRAP values of the NaDES extract of A. tinctoria root were determined by applying 40, 60 and 80 °C for 2 h. The first-order kinetics (k) and activation energy (Ea) expressing the thermal degradation of TPC, TFC and antioxidant properties of A. tinctoria root in NaDES were calculated. For the thermal stability experimentation, the samples were prepared with the optimum conditions figured out upon the response surface methodology consisting of sodium acetate:formic acid (1:4 M ratio), 45% water content and 25% mL solvent ratio (SAFA₁₄). The TPC, TFC, DPPH radical scavenging and FRAP values of A. tinctoria root extracts showed strong stability in NaDES during the experimentation. The values of TPC, DPPH radical scavenging and FRAP values increased from 390.16 to 404.82 mg GAE/g, 444.68 to 456.87 mmol TE/g and 537.63 to 568.20 mmol ISE/g during thermal treatment (Table 5). However, a slight decrease was observed at 80 °C, indicating the beginning of an obvious thermal degradation. Similar tendencies have reported during the thermal treatment of many agricultural products (Lupoae et al., 2019; Oancea et al., 2017, 2014; Zhan, 2018). Following the heat treatment, the highest degradation estimated at 17% loss of TFC occurred between 60 and 80 °C. This might be due to the prolonged heating time of 120 min. Our findings are consistent with the previous studies which have reported that the high temperature and prolonged magnitude of time caused the loss of flavonoids compounds (Oancea et al., 2017). The values of k varied from −6.91 · 10⁻⁶ to −0.74 · 10⁻⁶ s⁻¹ for TPC, from 1.80 · 10⁻⁶ to 29.91 · 10⁻⁶ s⁻¹, from −7.68 · 10⁻⁶ to −2.32 · 10⁻⁶ s⁻¹ for DPPH radical scavenging and from −7.68 · 10⁻⁶ to −2.32 · 10⁻⁶ for FRAP (Table 6). These k values were found to be very low compared k values previously mentioned for many fruits extracted either with water and ethanol or with hexane:acetone mixture (Oancea et al., 2017, 2014; Olivares-tenorio et al., 2017). Thus, phytochemical compounds together with the antioxidant activity of A. tinctoria root were more protected in SAFA₁₄ upon heat application. It has been proved bioactive compounds are stable in NaDESs during heat and light treatments as well as storage time (Dai et al., 2014). Moreover, low activation energies of 3.13 · 10⁴ J mol⁻¹ for TPC, 4.56 · 10⁴ J mol⁻¹ for TFC, 1.11 · 10⁴ J mol⁻¹ for and 1.94 · 10⁴ J mol⁻¹ for FRAP (Table 6). These Ea values indicated that the mass transfer during the extraction and heat process proceeded more easily in NaDES. In addition, the strong thermal stability observed for polyphenols and antioxidant activity extracted with NaDESs has been explained by the very strong network bonding occurring between extracts and NaDESs components (Dai et al., 2014).