Introduction

The turmeric rhizome (Curcuma longa L.) has been used since ancient times for medicinal, cultural, and culinary purposes (Esparza, 2021). Its high content of bioactive compounds, primarily curcuminoids, imparts antioxidant, anti-inflammatory, antimutagenic, chemopreventive, antiangiogenic, hepatoprotective, and antibiotic properties, among others (Fuloria et al., 2022; Papayrata et al., 2024; Wu et al., 2024).

Various methodologies have been employed to extract curcuminoids; however, challenges persist due to their low water solubility, chemical instability at neutral and alkaline pH, and photosensitivity (Fuloria et al., 2022). Conventional extraction methods, which typically involve toxic solvents, can yield high quantities but require long processing times, high energy consumption, and are not environmentally sustainable (Ahmed et al., 2023; Gökdemir et al., 2020). Alternatively, nonconventional technologies have been explored, including supercritical fluids (Sharma et al., 2023), ultrasound (Patil and Rathod, 2020), microwaves (Doldolova et al., 2021), enzyme-assisted extraction (Le-Tan and Jaeger, 2022), deep eutectic solvents (Doldolova et al., 2021), and ionic liquids (Gökdemir et al., 2020). While these methods can achieve yields comparable to or better than conventional techniques, they still involve significant time and cost.

Micellar extraction, also known as surfactant--mediated extraction, has demonstrated high efficiency, short processing times, and low environmental impact when applied to plant matrices (Azooz et al., 2021; Śliwa and Śliwa, 2021). However, its application to turmeric rhizomes has been limited. The micellar extraction process begins with the simultaneous micellization of the surfactant (micelle formation at the critical micellar concentration [CMC]) and the solubilization of bioactive compounds. In an aqueous medium, the hydrophobic part of the surfactant (alkyl chains) orients toward the micelle’s interior. In contrast, the hydrophilic part moves toward the surrounding aqueous solution (Śliwa and Śliwa, 2021). Micelles can aggregate into various structures (e.g., spherical micelles, worm-like micelles, vesicles, cubosomes), depending on the structural characteristics of the surfactant, pH, temperature, and other factors (Suga et al., 2016).

Extraction efficiency is influenced by the spatial configuration of the surfactant; specifically, the size and surface area of the micelle (Othman et al., 2019; Suliman et al., 2022; Yamini et al., 2020); the structure of the bioactive compounds (Śliwa and Śliwa, 2020); the length of the aliphatic chain of the surfactant (Vinarov et al., 2018); the use of ionic or nonionic surfactants and their combinations (Zarei, 2009); and the surfactant’s hydrophilic–lipophilic balance (Khani et al., 2019). Micelles exist in dynamic equilibrium with monomers in the aqueous phase, facilitating interactions between the bioactive compounds and the surfactant (Śliwa and Śliwa, 2021).

Solubilization primarily occurs through hydrophobic interactions between bioactive compounds and surfactant molecules (Śliwa and Śliwa, 2020). According to Śliwa and Śliwa (2020), polyphenol solubilization within micelles occurs in three stages: (1) slow diffusion of bioactive compounds to the micellar surface; (2) absorption; and (3) rapid incorporation into the micelle structure. The location of bioactive compounds within the micelle depends on their hydrophobicity: highly hydrophobic compounds migrate toward the nucleus; moderately hydrophobic bioactive compounds reside between the aliphatic chains and the polar regions; and hydrophilic molecules remain on the micellar surface (Rangel-Yagui et al., 2005). Additional factors influencing simultaneous solubilization and micellar extraction include hydrogen bonding and electrostatic interactions (Azooz et al., 2021; Śliwa et al., 2019).

Sodium oleate in aqueous solution exhibits a low Krafft temperature, indicating that micelle formation and, -consequently, micellar extraction can occur without heating (He et al., 2019). Furthermore, it inhibits enzymatic pathways responsible for curcuminoid conjugation, thereby enhancing the bioavailability of curcuminoids (Dong et al., 2017). Some researchers report high -efficiency, with phenol extraction yields of up to 82.0% from wastewater (Zhang et al., 2023).

In this context, the present study aimed to experimentally optimize the micellar extraction process for extracting bioactive compounds from turmeric using sodium oleate as surfactant, and to compare its performance with that of conventional ethanol-based extraction.

Materials and Methods

Materials and reagents

Turmeric rhizomes cultivated in La Julia, located in Montenegro, Quindío, Colombia, were used in this study. The micellar extract of turmeric was prepared using a mixture consisting of oleic acid (Sigma-Aldrich, St. Louis, MO, USA), sodium hydroxide (Panreac, Spain), glacial acetic acid (Scharlau, Spain), and deionized water.

Micellar extraction

The micellar extract was prepared following the methodology described by Śliwa and Śliwa (2020), with modifications. Initially, the micellar solution (MS) of sodium oleate (MS_SO) was prepared using the required amount of water, based on the turmeric/MSso mass ratio defined for each experiment. This solution was then mixed with the crushed turmeric using a Silverson L5M homogenizer (Silverson Machines Ltd., UK), equipped with a standard emulsifying mesh head, operating at 5000 rpm for the extraction times (t_E) specified in the experimental design.

Subsequently, the mixture was filtered through a cloth to remove unsolubilized lignocellulosic residue. The resulting supernatant was adjusted to a pH between 6 and 7 with glacial acetic acid and stored at 5°C for characterization.

The micellar extraction process was evaluated using response surface methodology with a face-centered central composite design (α = 1), comprising 19 experiments. The independent variables considered were: turmeric/MS_SO ratio (0.01–0.05 g/g); surfactant concentration (0.070–1.525%); and t_E (5–15 min). The dependent variables were: total phenolic content (TPC); antioxidant capacity (DPPH^• and ABTS^•+); curcumin (CUR); particle size (ϕ); and zeta potential (ζ).

The dependent variables were modelled using a -second-order polynomial model (Equation 1), where Y represents the dependent variable and A, B, and C are the independent variables, β₀ is the model constant, β_A, β_B, and β_C are the linear coefficients, β_AA, β_BB, and β_CC -represent the quadratic coefficients, and β_AB, β_AC, and β_BC are the linear interaction coefficients of the independent variables:

Y = β₀+β_AA + β_BB + β_CC + β_ABAB + β_ACAC + β_BCBC + β_AAA² + β_BBB² + β_CCC² (1)

A multiresponse experimental optimization was also conducted to determine the independent variables that ensure the physicochemical stability of the micellar extract, and to maximize the TPC, DPPH^•, ABTS^•+, and CUR content. The models were validated, and the mean relative error (MRE) was calculated using Equation 2. Experimental values were obtained from three replicate experiments under optimal conditions.

Conventional extraction

Micellar extraction was performed following the methodology described by Yang et al. (2020), using 80% ethanol (Sigma-Aldrich, USA). Initially, 2 g of crushed turmeric and 50 mL of 80% ethanol were placed in a plastic-covered Erlenmeyer flask and incubated for 60 min in a water bath maintained at 35°C. The flask was mounted on a horizontal shaking device (model SV 29/45, Memmert, Germany) operating at 105 oscillations/min. Subsequently, the mixture was centrifuged, and the supernatant was collected and stored at 5°C for characterization.

Extract characterization

Total phenolic content was determined using the Folin–Ciocalteu method, following the modified procedure of Pandey et al. (2021). In an amber container, 50 mL of the sample was diluted with 950 mL of ethanol. Then, 50 mL of the diluted sample was mixed with 800 mL of distilled water and 100 mL of 20% w/v Na₂CO₃ solution and left to stand in the dark for 5 min. Subsequently, 50 mL of 1 N Folin–Ciocalteu reagent (Sigma-Aldrich, St. Louis, MO, USA) was added, stirred, and allowed to react in the dark for 2 h. The absorbance was then measured at λ = 765 nm. TPC was expressed as mg gallic acid equivalent (GAE)/g turmeric dry basis (db), based on a calibration curve that related the absorbance to the concentration of a gallic acid standard (Sigma-Aldrich, USA), with the equation: Absorbance = 0.0033 (GAE concentration) – 0.1079 (R²= 0.991).

The antioxidant capacity was determined using the ABTS^•+ and DPPH^• methods. For ABTS^•+, the modified methodology of Hemlata et al. (2020) was followed. A stock solution of the ABTS^•+ radical was prepared by mixing ABTS reagent (Sigma-Aldrich, USA) with potassium persulfate (J.T. Baker, Madrid, Spain). The working solution was obtained by diluting an aliquot of the stock solution in ethanol until an absorbance of 0.700 ± 0.002 was reached. A mixture of 100 mL of sample and 900 mL of ethanol was prepared separately. From this mixture, 50 mL was combined with 950 mL of the working solution and left to stand in the dark for 7 min. The absorbance was then measured at λ = 734 nm.

For the DPPH^• assay, a modified version of the method by Hemlata et al. (2020) was used. A stock solution of the DPPH^• radical was prepared by dissolving DPPH reagent (Sigma-Aldrich, USA) in ethanol until an absorbance of 0.700 ± 0.002 was reached. A separate mixture of 200 mL of sample and 800 mL of ethanol was prepared, from which 50 mL was added to 1000 mL of the DPPH^• solution. The final mixture was left to stand in the dark for 30 min, and the absorbance was measured at λ = 517 nm.

Results from both assays were expressed as mg Trolox equivalent (TE)/g turmeric db, where calibration curves were previously constructed, recording the absorbance versus the concentration of Trolox standards (Sigma-Aldrich, USA): ABTS^•+ (Inhibition (%) = 0.151(TE concentration) – 1.431, R² = 0.9989) and DPPH^• (Inhibition (%) = 0.222(TE concentration) – 1.616, R² = 0.994).

Curcumin was quantified using the methodology described by Yusuf et al. (2021), employing a Shimadzu 20A high-performance liquid chromatograph equipped with a DGU 20A5 degasser, SPD–M20A diode array detector, RID-10A refractive index detector, LC-20AD autosampler, and CTO-10A oven. A Luna C18 column (5 μm, 250 × 4.6 mm) and an AJO 4287 precolumn (Phenomenex, Torrance, CA, USA) were used. An isocratic method was employed, utilizing a mobile phase composed of a 50:50 (v/v) mixture of 1% w/v aqueous acetic acid and gradient-grade acetonitrile (Merck KGaA, Germany). Initially, 100 µL of the sample was dissolved in 900 µL of the mobile phase, shaken, and filtered through a 0.45 µm polyvinylidene fluoride syringe filter (Whatman, UK). Finally, 20 µL of the resulting solution was injected into the chromatograph at a flow rate of 1 mL/min. Quantification of CUR was performed at λ = 426 nm, determined using a calibration curve constructed with chromatography-grade CUR (99.2% purity) (AK Scientific, USA): Area =199119.292 (CUR concentration) – 13148.923 (R² = 0.999). The detection and quantification limits of CUR were 0.311 and 0.942 mg/L, respectively.

The particle size (φ) and zeta potential (ζ) were determined using a Zetasizer Nano ZS90 analyzer (Malvern Instruments Ltd., Worcester, UK) at 25°C. Particle size was measured without dilution or acidification of the extract, using dynamic light scattering with an M3-PALS analyzer (Kim et al., 2019), and an absorption index of 0.05, using a DTS0012 cell. The refractive index was previously determined as 1.340. Zeta potential was measured using electrophoresis and laser Doppler velocimetry, employing a DTS 1060 capillary cell (Morteza-Semnani et al., 2022).

Data analysis

Data were analyzed using Statgraphics Centurion XVII software. Analysis of variance (ANOVA) was conducted at a significance level of 5%.

Results

Micellar extraction process

Table 1 presents the experimental design and the mean values of the dependent variables for the micellar extraction process. Additionally, Figure 1 -displays the surface and volume response graphs of bioactive compounds.

Total phenolic content

Mean TPC values ranged from 86.1 to 181.0 mg GAE/g turmeric db. ANOVA revealed significant differences (P < 0.05) only concerning the turmeric/MS_SO ratio. These results are comparable to those reported by Singh et al. (2022) for microwave-assisted extraction using 95% ethanol (178.4 ± 4.8 mg GAE/g turmeric db). They are notably higher than those reported by Ivanović et al. (2021) for hydrodistillation extraction at 130°C (2.80 ± 0.19 mg GAE/g turmeric), highlighting the influence of both the solvent and the extraction method.

Figure 1A illustrates the TPC response, showing an increase in TPC as the turmeric/MS_SO ratio decreases. This trend is attributed to enhanced solvation of bioactive compounds due to the increased presence of MS_SO, which provides a greater number of micelles for turmeric. These micelles facilitate the solubilization and subsequent extraction of bioactive compounds (Śliwa and Śliwa, 2020). Similar findings have been reported by Ahmed et al. (2023) and Niazmand et al. (2021).

The optimal conditions for maximizing the TPC (red zone) were: turmeric/MS_SO = 0.01–0.02; -surfactant = 0.07–1.04%; and t_E = 5–13 min. Skrypnik and Novikova (2020) reported higher TPC extraction at a Tween 80 concentration of 1.14% and an apple/MS ratio of 0.0096 g/mL, which closely aligns with the minimum value in the present study.

Antioxidant capacity

The mean values of DPPH^• and ABTS^•+ ranged between 17.1–68.5 and 35.8–176.2 mg TE/g turmeric db, respectively. These results are notably higher than those reported in the literature. For instance, Papayrata et al. (2024) reported a DPPH^• value of approximately 19 mg TE/g turmeric db using 80% methanol at 37°C for 12 h. Wu et al. (2024) reported an ABTS^•+ value of 0.425 mg TE/g dry basis (db) via ethanolic extraction at room temperature for 24 h. Additionally, Ivanović et al. (2021) and Şahin (2018) reported ABTS^•+ values of 40 mg and 29.2 ± 1.1 mg TE/g db, respectively, using ultrasound-assisted ethanol extraction.

These findings suggest that micellar extraction is capable of extracting both lipophilic and hydrophilic bioactive compounds, whereas ethanol primarily targets phenolic bioactive compounds due to the polarity of the solvent (Calle Chumo et al., 2022; Rumpf et al., 2023). Turmeric is known to be rich in both lipophilic and hydrophilic bioactive compounds (Friesen et al., 2019), which explains the higher ABTS^•+ values observed in comparison to DPPH^•. Since DPPH^• is soluble in organic media, it predominantly reacts with lipophilic bioactive compounds, resulting in lower sensitivity to certain antioxidants (Rumpf et al., 2023). In contrast, ABTS^•+ is more suitable for evaluating antioxidant capacity, especially in samples containing hydrophilic compounds or pigments (Faria et al., 2021).

ANOVA revealed significant differences (P < 0.05) for both assays (DPPH^• and ABTS^•+) concerning turmeric/MS_SO ratio and surfactant. Additionally, DPPH^• showed significant variation with the linear interaction between turmeric/MSSO–surfactant, while the quadratic interaction between tE and ABTS^•+ significantly influenced ABTS^•+. Figures 1B and 1C display trends for DPPH^• and ABTS^•+ that closely mirror those observed for TPC. The influence of the turmeric/MS_SO ratio on antioxidant capacity is evident, as higher surfactant levels enhance the solubilization of bioactive compounds by increasing the availability of sodium oleate (Motikar et al., 2021; Śliwa and Śliwa, 2020). Motikar et al. (2021) observed a similar effect in the micellar extraction of pomegranate peel/MS (1:30 → 1:100), where DPPH^• increased from 78.2 to 84.4%. ABTS^•+ values increased consistently with increasing surfactant concentration, whereas DPPH^• increased only at a high turmeric/MSSO ratio (0.05). Motikar et al. (2021) also reported a similar trend for ABTS^•+ when increasing surfactant concentration (4 → 8% v/v).

For DPPH^•, the positive interaction between turmeric/MS_SO and surfactant is demonstrated by its increase at lower values of both variables. In contrast, the ABTS^•+ response exhibits a curvilinear trend concerning the surfactant–surfactant interaction, initially increasing and then reaching an asymptote at higher surfactant values (1.525%), indicating a stronger interaction between surfactant and bioactive compounds.

The optimal conditions for maximizing antioxidant capacity (red zone) were as follows:

• DPPH^•: turmeric/MS_SO = 0.01–0.02; surfactant and t_E across the entire range

• ABTS^•+: turmeric/MS_SO = 0.01–0.02; surfactant = 0.555–1.525%; t_E across the full range

Curcumin

The mean CUR values ranged from 3.3 to 11.6 mg/g turmeric dry basis (db). ANOVA revealed significant differences (P < 0.05) concerning both surfactant and surfactant–surfactant interactions. To date, there are no reports in the literature on the micellar extraction of CUR using sodium oleate as a surfactant. Park et al. (2022) reported CUR values ranging from 8.3–9.4 mg/g turmeric db using 70% ethanol for 2 h at 100°C, attributing the variation to differences in rhizome drying methods. Gilda et al. (2010) reported a higher yield of 20.3 mg/g turmeric db using Gelucire 44/14 as surfactant.

Other studies have highlighted the considerable variability in CUR content, which is influenced by the initial bioactive compounds composition in turmeric. This composition is affected by environmental factors such as light exposure, ultraviolet radiation, and temperature (Li et al., 2020a). Additionally, extraction efficiency is influenced by several factors, including the polarity, viscosity, and surface tension of the solvent, as well as the extraction temperature and the solute–solvent ratio, among others (Array et al., 2018; Patil et al., 2019; Sepahpour et al., 2018).

Figure 1D presents the response surface graphs for CUR. A curvilinear (convex) increase in CUR is observed with increasing surfactant, attributed to the affinity between micelles and CUR; similar to the trends observed for TPC, DPPH^•, and ABTS^•+. Additionally, the surfactant–surfactant interaction reveals a minimum CUR value when the surfactant levels range between 0.070 and 0.555%.

Alibade et al. (2020) reported a similar trend in the extraction of antioxidants from residual by-products during cloud point micellar extraction, although they selected lower surfactant levels due to cost considerations. Similarly, Guo et al. (2019) observed an increase in polyphenols and alkaloids extracted from mulberry leaves during cloud point micellar extraction using Triton X-114.

The optimal conditions for maximizing CUR (red zone) within the studied region were: turmeric/MS_SO = 0.01–0.045, surfactant = 1.525%, and t_E across the entire range.

Elsewhere, Figure 2 presents the surface and volume response graphs of particle size and zeta potential, as well as the pH behavior in the micellar extraction process.

Particle size

The mean particle size values ranged from 165.5 to 484.2 nm. ANOVA revealed significant differences (P < 0.05) concerning surfactant, the surfactant–-surfactant interaction, and the turmeric/MS_SO–surfactant linear interaction. Gontsarik et al. (2018) and Suga et al. (2016) have linked the particle size of MS to pH. In this context, Figure 2C shows an increase in the pH of the extract with increasing surfactant, with the highest values observed when turmeric/MS_SO was between 0.03 and 0.05. Suga et al. (2016) reported different structural forms in aqueous sodium oleate solutions depending on pH: cubosomes at pH 7.5, vesicles at pH 8.5, and micelles (ϕ < 50 nm) and vesicles (ϕ < 1000 nm) at pH 9.7–10.1. Additionally, Gontsarik et al. (2018) reported particle size values ranging from 200 to 300 nm in cubosomes formed from oleic acid self-assembled systems with human cathelicidin LL-37 in aqueous solution.

Figure 2A presents the particle size response graphs during the micellar extraction of turmeric, where a decrease in particle size is observed with increasing surfactant. This trend is consistent with previous findings and is attributed to the enhanced formation of micellar structures of sodium oleate (Ogino et al., 1988), particularly when concentrations exceed the CMC of 0.033% w/w (Sergeev et al., 2022). Additionally, the largest particle size values occur when the extract exhibits a low pH.

Smaller particle size values suggest a greater number of micelles, which enhances the extraction of bioactive compounds due to an increased interaction with surfactant and improved extract stability (Li et al., 2020b; Wu et al., 2016). Furthermore, the turmeric/MS–-surfactant interaction was positive, indicating that particle size increases when both turmeric/MS_SO and surfactant decrease. This condition corresponds to a high electrical potential (ζ ≈ –80 and –90 mV), which may contribute to the expansion of micellar structures. Finally, the conditions of the independent variables that minimize zeta potential are illustrated in the blue zone, where surfactant = 0.555–0.525%, and turmeric/MS_SO and t_E span the entire range.

Zeta potential

The mean zeta potential values ranged from –100.3 to –30.6 mV, indicating a strong negative electrical potential that supports the physicochemical stability of the system due to the predominance of repulsive forces (|ζ| > 30 mV) (Pinheiro et al., 2020; Sun et al., 2016). ANOVA revealed significant differences (P < 0.05) concerning turmeric/MS_SO and surfactant, as well as the turmeric/MS_SO–surfactant and surfactant–surfactant interactions. The high negative electrical potential is primarily attributed to the increase in surfactant (Bhattarai et al., 2018) and the alkaline pH of the extracts (7.8–10.1) (Cacua et al., 2019). Figure 2B presents the zeta potential response graphs during the micellar extraction of turmeric, showing an increase in electric potential with increasing surfactant, particularly when turmeric/MS_SO is formulated at 0.01. This behavior is attributed to the anionic nature of sodium oleate, which enhances the negative electric potential in the surface layer (Bhattarai et al., 2018).

The presence of co-ions in the MS also influences micellar structure and zeta potential (Santos et al., 2017), suggesting that ions contributed by turmeric play a significant role in this phenomenon. A lower negative electric potential is observed at low surfactant levels (0.07–0.555%), along with increasing turmeric/MS_SO (red zone), likely due to the higher content of bioactive compounds and their interaction with the MS. Śliwa and Śliwa (2020) and Śliwa et al. (2019) described this condition as a neutralizing effect on zeta potential, where lower surfactant (fewer micelles) is more susceptible to the incorporation of bioactive compounds, thereby reducing the surface charge.

The t_E was not a critical factor in the micellar extraction of turmeric in the findings reported by Zhang et al. (2016) in microwave-assisted micellar extraction of Sophora flavescens Ait alkaloids, and by Mandal and Lahiri (2022) in cloud point micellar extraction of beryllium and chromium. However, other studies have identified t_E as an influential variable (Motikar et al., 2021; Skrypnik and Novikova, 2020). Finally, the conditions that favored a higher negative electric potential (blue zone) were: -turmeric/MS_SO = 0.01–0.03; t_E = 15 min; and surfactant across the entire range.

Mathematical modelling and experimental optimization of multiple responses

Table 2 presents the regression and determination coefficients (R²) for the models of the dependent variables. Most of the R² values account for over 80% of the variability in the experimental data, indicating a good model fit. However, the model for CUR showed the lowest acceptable R² value (76.0%), likely due to its sensitivity to processing conditions or other factors that introduce high variability, thereby limiting its predictive accuracy.

Additionally, ANOVA results indicated that most models exhibited statistically significant effects (P < 0.05), except the CUR (P > 0.05), suggesting that the behavior of CUR is only partially explained by the model. This limitation may be attributed to the potential degradation of CUR under alkaline conditions (Cano-Higuita et al., 2015).

Table 3 presents the criteria, impacts, and sensitivity parameters established for the experimental optimization of the micellar extraction of bioactive compounds from turmeric. The overall desirability reached 70.9%, with the optimal conditions for the independent variables defined as follows: turmeric/MS_SO = 0.01 g/g; surfactant = 1.525%; and t_E = 6.64 min. Additionally, the validation of the mathematical models yielded MRE values of up to 16.2%, which is considered acceptable according to Castaño-Peláez et al. (2022). These results demonstrate a good degree of fit for the models, consistent with the desirability function used for multiresponse optimization. Moreover, they validate the predictive ability of the developed models, confirming the effectiveness of the statistical approach applied for their optimization.

Comparison of extraction methodologies

Figure 1 presents a comparison of the TPC, DPPH^•, ABTS^•+, and CUR results from conventional extraction and micellar extraction, along with the corresponding P-values from the t-test. The results show that the mean values of TPC, DPPH^•, and ABTS^•+ were significantly higher in the micellar extraction compared to conventional extraction (P < 0.05), whereas the CUR content was significantly lower (P < 0.05).

The higher TPC, DPPH^•, and ABTS^•+ values observed in the micellar extract are attributed to the ability of micelles to interact with a broad range of bioactive compounds, both lipophilic and hydrophilic, when amphiphilic molecules are employed during their extraction (Ziyatdinova and Budnikov, 2021). The TPC results suggest that micellar extraction has a greater capacity to extract phenolic bioactive compounds beyond CUR, including demethoxycurcumin, bisdemethoxycurcumin, cyclocurcumin, β-sesquiphellandrene, α-curcumene, ar-turmerone, and α/β-turmerone, among others (Tyagi et al., 2015). Ziyatdinova et al. (2016) reported comparable values in ultrasound-assisted micellar extraction (surfactant: Brij 35) for phenolic antioxidants extracted from cinnamon, which showed an improvement of over 100% compared to conventional extraction methods. These findings were further validated in extracts from cloves, oregano, and black pepper. Similarly, compared to traditional extraction techniques, He et al. (2015) achieved a yield improvement of over 90% in the purification of -lecithin from black beans using micellar extraction based on sodium bis(2-ethylhexyl) sulfosuccinate.

However, the Folin–Ciocalteu method is known to be affected by interferences from reducing sugars, which can lead to an overestimation of TPC (Briones Muñoz and Riera, 2020). In this context, the micellar extraction may promote a certain degree of alkaline hydrolysis of the starch present, resulting in the formation of reducing sugars due to the elevated pH and increased extraction temperature (Lawag et al., 2023). Furthermore, the difference between ABTS^•+ and DPPH^• values is more pronounced in the micellar extraction than in the conventional extracts. This difference is likely due to the broader range of bioactive compounds extracted by the micellar extraction compared to ethanol-based extraction, which is limited by its polarity (Calle Chumo et al., 2022; Ziyatdinova and Budnikov, 2021).

The higher extraction of CUR with ethanol is attributed to the strong chemical affinity between them, as reported by various authors (Fernández-Marín et al., 2021; Rezaei et al., 2023; Yang et al., 2020). Additionally, it has been demonstrated that the degradation of CUR is pH-dependent, occurring more slowly under acidic conditions (pH < 7) and accelerating in alkaline environments, particularly at pH values near 10.8. Additionally, elevated temperatures further promote its decomposition (Aboudiab et al., 2020). These findings suggest that the extraction conditions used in this study (alkaline pH and the temperature increase associated with high-speed homogenization) could have contributed to an accelerated degradation rate of CUR. However, the health benefits of CUR have been shown to increase when it acts synergistically with other bioactive compounds present in turmeric (Panda et al., 2021). In general, the results were influenced by several factors, including kinetic parameters, radical sensitivity, adduct formation, reaction mechanisms, and the characteristics of the extract, among others (Ilyasov et al., 2020; Rumpf et al., 2023).

Table 4 presents the energy requirements and costs associated with micellar and conventional CUR extraction. The data indicate that micellar extraction results in lower energy consumption (by 68.3%), yield (by 28.8%), and production costs (by 85.0%) compared to the conventional extraction methods. Importantly, micellar extraction is easier to scale up, and the necessary equipment is readily available. In contrast, conventional extraction relies on heating in water bath, which poses limitations in terms of space and production capacity.

Several studies have evaluated the extraction of bioactive compounds using green technologies, showing a wide range of extraction yields. For instance, microwave--assisted extraction using acetone achieved a yield of 53.9% (Sahne et al., 2016), while subcritical water extraction modified with ethanol reached up to 152% (Rezaei et al., 2023), although this method requires high temperatures (90°C) and increased pressure (2 MPa). Similarly, supercritical CO₂ extraction has shown yields around 102.4%; however, it requires elevated pressures (25 MPa) and prolonged extraction times (90 min), leading to high energy consumption (Chhouk et al., 2017).

In comparison, the micellar extraction developed in this study achieved a yield of 75.2%, comparable to that obtained with ionic liquids: 74.3% without ultrasound and 86.7% with ultrasound assistance (Patil et al., 2021). In addition to the favorable yield, micellar extraction presents several advantages over other nonconventional techniques: (1) it can be easily scaled up using conventional equipment, (2) it requires very short extraction times (below 7 min), and (3) it operates under mild conditions without the need for high pressure or temperature. These benefits, combined with its low energy consumption and the use of food-grade, safe compounds, highlight micellar extraction as a promising and sustainable green technology for the recovery of bioactive compounds.

Conclusion

In general, the micellar extraction of turmeric is primarily influenced by the turmeric/MS_SO and the surfactant, with optimal performance observed at higher levels of both micellar sodium oleate and surfactant. These conditions enhance the interaction between MS_SO and the bioactive compounds of turmeric. Additionally, higher surfactant concentrations increase the number of micelles available to interact with turmeric, thereby improving solubilization of bioactive compounds and the overall efficiency of the micellar extraction process. Contrary to the expectations, t_E was not a critical factor, suggesting that the solubilization of bioactive compounds occurs rapidly.

Micellar extraction demonstrated higher TPC and antioxidant capacity compared to conventional extraction, while also requiring less energy and incurring lower associated costs. In this context, micellar extraction emerges as a cost-effective and environmentally friendly technology that delivers acceptable yields. It is therefore considered a promising alternative for industrial-scale applications.