Utilizing Natural Deep Eutectic Solvents (NADESs) for Sustainable Phytonutrient Recovery: Optimization and Multi-Matrix Extraction of Bioactive Compounds

Featured Application

The optimized NADES-based extraction method developed in this study presents a sustainable alternative for the nutraceutical, pharmaceutical, and cosmetic industries to recover polyphenols (e.g., from Lonicera japonica), betalains (e.g., from beetroot), anthocyanins (e.g., from chokeberry pomace), and other bioactive compounds from diverse plant matrices. By replacing traditional volatile organic solvents with eco-friendly NADESs, this approach enhances extraction efficiency for hydrophilic phytochemicals while reducing environmental impact, aligning with green chemistry principles for industrial-scale phytochemical recovery.

Abstract

Bioactive phytochemicals, such as polyphenols, play vital roles in human health, but conventional extraction methods rely on hazardous solvents. This study establishes natural deep eutectic solvents (NADESs) as versatile and environmentally friendly alternatives for recovering a variety of bioactive compounds from plant materials. Five choline chloride-based NADESs were evaluated for their effectiveness in extracting betalains (from beetroot), carotenoids (from carrot and sweet potato), anthocyanins (from chokeberry pomace and red onion), and polyphenols (from Lonicera japonica flowers, hop cones, rowan berries, and spent coffee grounds). Notably, NADES2 outperformed water in betalain recovery (179.86 mg of betanin/100 g of beetroot), while NADES4 (choline chloride-urea, 1:2 molar ratio) matched the polyphenol extraction efficiency of ethanol. Using L. japonica flowers as a model for optimization, Response Surface Methodology (RSM) identified the solvent ratio and temperature as critical extraction parameters, using high ratios (12:1–15:1 v/w) and moderate heat (55–75 °C) to maximize recovery. NADES4 emerged as a high-performing solvent, achieving a total phenolic content (TPC) of 75.94 mg chlorogenic acid/g and antioxidant activity of 451.00 µmol Trolox/g under the following conditions: 60% aqueous dilution, 15:1 solvent ratio, and 80 °C, 30 min. These findings highlight NADESs as a green, tunable solvent system for phytochemical extraction across plant species, offering enhanced efficiency, reduced environmental impact, and alignment with sustainable practices.

1. Introduction

Bioactive phytochemicals, which include a wide variety of compounds such as polyphenols, terpenes, terpenoids, alkaloids, carotenoids, and betalains, play a crucial role in both plant defense and human health [1]. Phenolic compounds represent a diverse group of secondary metabolites synthesized by plants and other organisms, characterized by the presence of a benzene ring with one or more hydroxyl (-OH) groups directly bonded to it. The main classes of polyphenols comprise flavonoids, phenolic acids, stilbenes, and lignans, collectively forming a structurally diverse family with an estimated 5000 to 8000 distinct structural variants [2]. These bioactive molecules are pivotal in reducing oxidative stress, a critical factor in developing chronic diseases such as cardiovascular disorders, diabetes, and cancer [3]. Anthocyanins, a type of flavonoids responsible for the vibrant violet, blue, and red pigmentation in many plants, demonstrate potent free radical-scavenging properties [4]. Carotenoids, which are lipid-soluble pigments, contribute to provitamin A activity and protect cells from oxidative damage [5]. Meanwhile, betalains, nitrogen-containing pigments, are valued for their antioxidant and anti-inflammatory properties [6]. Together, these compounds highlight the importance of phytochemicals in both plant resilience and human health.

Traditional extraction methods like solvent extraction and maceration have been widely utilized to obtain bioactive compounds from plants. Many techniques, such as Soxhlet extraction, require high temperatures for solvent boiling and extended processing times [7]. However, these techniques often rely on solvents, including methanol, ethanol, and acetone, which are volatile organic compounds (VOCs) that contribute to air pollution and are linked to climate change [8,9]. Moreover, VOCs pose environmental and health risks due to their toxicity and potential carcinogenic properties. Some solvents can be dangerous if ingested or inhaled, and they may cause skin irritation upon contact. Certain solvents, like carbon tetrachloride, methylene chloride, and toluene, are identified as carcinogenic [10]. Furthermore, many conventional solvents are flammable, and their disposal poses challenges and high costs, potentially leading to environmental contamination [11].

In response to these challenges, the principles of green chemistry have emerged as a transformative paradigm. Green chemistry advocates for eco-friendly solvents and energy-efficient processes. It focuses on creating chemical products and processes that minimize or completely eliminate hazardous substances. Paul T. Anastas and John C. Warner first introduced the Twelve Principles of Green Chemistry in 1998 [12]. In the quest for new solvents, the journey is significantly longer. The first attempts to develop new solvents began at the beginning of the 20th century. Walden, in 1914, described the first ionic liquid, ethylammonium nitrate [CH₃CH₂NH₃]⁺[NO₃]⁻, a new type of solvent that is characterized by a liquid state at ambient conditions, with low volatility, composed of organic cations and inorganic anions [13].

The subsequent milestone was the identification of deep eutectic solvents (DESs). The first DES reported by Abbott et al. [14] was a 2:1 mixture of urea (melting point of 133 °C) and choline chloride (melting point of 302 °C), resulting in a melting point of only 12 °C. DESs are typically formed by combining hydrogen bond donors (HBDs, such as organic acids and sugars) with hydrogen bond acceptors (HBAs, such as quaternary ammonium salts) in specific molar ratios. This mixture is highly versatile due to the wide variety of compounds that can fulfill these roles and the potential applications for these mixtures [15].

Finally, NADESs, i.e., natural deep eutectic solvents, were proposed by Choi et al. [16] as a missing piece of the puzzle linking solvents composed of natural components, safe processes, and user safety. These mixtures have emerged as sustainable, eco-friendly alternatives to traditional organic solvents for extracting phytochemicals. They enhance the solubility and stability of extracts while improving the bioactivity and bioavailability of various drugs and medicines. Additionally, the main benefits compared to conventional solvents include being biodegradable, non-toxic, and recyclable. Their tunable properties—such as polarity and viscosity—enable customization through adjustments in component ratios or by the incorporation of water [17].

This study evaluates five distinct natural deep eutectic solvents (NADESs) for their effectiveness in extracting polyphenols, carotenoids, anthocyanins, and betalains from various plant materials. Furthermore, the current study addressed solubility issues related to chemically diverse phytochemicals, such as polar anthocyanins and lipid-soluble carotenoids, by customizing choline chloride-based systems with organic acids, polyols, carbohydrates, or urea. While a preliminary screening showed promising results for L. japonica, further optimization using Response Surface Methodology (RSM), conducted specifically on L. japonica flowers as the model system, was applied to maximize total phenolic content (TPC) and antioxidant activity. This optimization identified optimal NADES (composition and water addition) and process parameters (temperature, time, and solvent-to-material ratio).

2. Materials and Methods

2.1. Materials

The following plant and agro-food waste materials were used in this study: dried flowers of Japanese honeysuckle (Lonicera japonica) (Nanga, Błękwit, Poland), rowanberries (Sorbus aucuparia, collected near Skierniewice, Łódź Voivodeship, Poland, in September 2024; subsequently washed, frozen, and freeze-dried for 24 h in the Christ Gamma 1–16 apparatus (Osterode am Harz, Germany)), spent coffee grounds sourced from a local café in Warsaw, hop cones (collected from a garden in the Warsaw area and dried at 40 °C), chokeberry pomace (Greenherb Company, Wysoka, Poland), along with dried red onion, sweet potato, carrot (Planteon sp. z o.o., Borków Stary, Poland), and dried beetroot (SYS Foods Sp. z o.o., Kamionka, Poland). Following this, all the materials were ground using the IKA Tube Mill 100 control (IKA Werke GmbH, Staufen im Breisgau, Germany).

To create deep eutectic solvents, the following substances were utilized: choline chloride, urea, glucose, sorbitol, glycerol, and citric acid. All substances and other chemicals were sourced from Sigma-Aldrich (Poznań, Poland).

2.2. NADES Preparation

To prepare NADESs, choline chloride served as the hydrogen bond acceptor, while glucose, glycerol, citric acid, urea, and sorbitol functioned as hydrogen bond donors. These components were mixed in specific molar ratios [18,19], as shown in

Table 1. Composition of choline chloride-based natural deep eutectic solvents (NADESs).

Mixture	HBA–HBD	Molar Ratio (HBA–HBD)	pH (After Addition of 20% Water)
NADES1	Choline chloride–Glucose	2:1	4.80 ± 0.07
NADES2	Choline chloride–Glycerol	1:2	6.67 ± 0.04
NADES3	Choline chloride–Citric acid	2:1	0.83 ± 0.05
NADES4	Choline chloride–Urea	1:2	9.07 ± 0.06
NADES5	Choline chloride–Sorbitol	1:1	3.92 ± 0.08

Abbreviations: NADES—natural deep eutectic solvent, HBA—hydrogen bond acceptor, and HBD—hydrogen bond donor.

.

NADESs were obtained using a direct heating method. Precise masses of components were combined in sealed glass containers. The mixtures were then submerged in a temperature-controlled water bath at 70 °C for 2 h to facilitate eutectic formation. To ensure homogeneity, the mixture was periodically agitated manually during heating. After the thermal treatment, 20% (w/w) distilled water was gradually added to the cooled mixture while stirring continuously to reduce the viscosity of the resulting mixtures. The pH values of each NADES were measured using an Elmetron CP-551 pH meter (Zabrze, Poland) at 25 °C. The device was calibrated with standard pH buffer solutions.

2.3. Betalain Content Determination

Betalain analysis was performed using spectrophotometry, following the modified method detailed by Niemira and Galus [20]. Ground dried beetroot samples (0.1 g) were extracted using 15 mL of either the tested NADES or water, which served as a control solvent, due to the hydrophilic properties of betalains. The falcons were mixed vigorously with a vortex for 2 min and then mixed with an RM-2M rotator (neoLab, Heidelberg, Germany) for 10 min. Subsequently, the samples were centrifuged (8000 rpm, 10 min, MPW-351R, Warsaw, Poland). Absorbances were measured at 476, 538, and 600 nm using a Rayleigh UV-1601 spectrophotometer (BRAIC, Beijing, China) with the pure solvent employed for extraction as the blank. The contents of red pigment (mg of betanin/100 g of dry mass (d.m.)) and yellow pigment (mg of vulgaxanthin/100 g d.m.) in the obtained extracts were then calculated.

2.4. Carotenoid Content Determination

The carotenoid levels in dried sweet potato and carrot samples were measured using a spectrophotometric method based on a β-carotene standard curve. One gram of the ground sample was combined with 10 mL of the tested solvents, and, for comparison, either water or acetone was utilized. Following the same protocol as before, the falcon tubes were vortexed for 2 min, mixed for 10 min, and centrifuged, and the absorbance of the resulting solutions was recorded at 454 nm. Carotenoid content was reported as mg of β-carotene/g d.m.

2.5. Anthocyanin Determination

One gram of chokeberry pomace or dried red onion was placed into falcons, and 10 mL of NADES, water, ethanol, or a solution of ethanol–0.1 M HCl (85:15 v/v) was added. The suspensions were vortexed for one minute, followed by a 30 min incubation in a water bath at 40 °C. After the incubation, the falcons were vortexed again for one minute and then centrifuged at 8000 rpm for 10 min. Then, the monomeric anthocyanin content was determined using the pH-differential method adapted from Jiang et al. [21]. Briefly, 0.3 mL of the anthocyanin extract was mixed separately with 9.7 mL of pH 1.0 (0.025 M potassium chloride buffer) and pH 4.5 (0.4 M sodium acetate buffer) solutions. After being equilibrated in the dark at room temperature for 30 min, absorbance was measured at 525 nm and 700 nm using a spectrophotometer. Total monomeric anthocyanins were subsequently calculated as cyanidin-3-glucoside equivalents (mg of cyanidin-3-glucoside/g d.m.).

2.6. Total Phenolic Content

The total phenolic content (TPC) in the extracts was assessed using the Folin–Ciocalteu method as outlined by Rybak et al. [22]. This phenolic compound content was expressed in terms of chlorogenic acid equivalents (mg CGA/g). The absorbance measurements at 750 nm followed the extraction of phenolics from L. japonica flowers, rowanberries, hop cones, and spent coffee grounds. In accordance with the previous methods, one gram of ground materials was placed in falcon tubes with 10 mL of NADES, water, or 80% ethanol as solvents. The extraction process was completed with 2 min of vortexing, 10 min of mixing with a rotator, and, then, 20 mL of water was added before the samples were centrifuged (8000 rpm, 10 min). Subsequently, due to precipitation during measurements, the obtained extracts were diluted approximately 500-fold.

2.7. Antioxidant Activity Measurements

The extracts derived from L. japonica flowers were also tested for antioxidant activity using the DPPH radical scavenging assay [23]. A 0.004% solution of DPPH in methanol was prepared. In glass tubes, 0.3 mL of the sample was mixed with 2.7 mL of the DPPH solution. After 30 min of incubation in the dark, the absorbances were measured at 517 nm. The antioxidant activity was expressed as µmol Trolox/g d.m.

2.8. Optimization of Water Addition to NADES

To investigate how the aqueous dilution of NADES4 affects both the extraction efficiency of phenolic compounds and the antioxidant potential in L. japonica extracts, NADES4, made of choline chloride and urea in a 1:2 molar ratio, was synthesized again by heating the components at 70 °C in a water bath for 2 h until a homogeneous liquid formed. Aqueous NADES4 solutions were prepared by diluting the pure NADES4 with distilled water to create mixtures containing 10–100% NADES4 (w/w) at 10% intervals (10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100%). After preparing the solutions, phenolic compound extraction was again accomplished using the method described above, and TPC and DPPH radical scavenging activity were measured.

2.9. Statistical Analysis

The obtained results were analyzed statistically using Statistica 13.3 software (TIBCO Software Inc., Palo Alto, CA, USA). A one-way analysis of variance (ANOVA) and Tukey’s post hoc test were applied. The significance level was α = 0.05.

Furthermore, due to the promising results, the extraction process for phenolic compounds from L. japonica leaves was optimized using experimental designs for Response Surface Methodology. A three-factor, three-level experiment was planned using Box–Behnken designs (

Table 2. Coded levels and decoded values of Box–Behnken design.

Factors	Name	Units	Low (−1)	Medium (0)	High (+1)
1	Temperature	°C	40	60	80
2	Time	min	10	30	50
3	Liquid/Solid Ratio	v/m	5	10	15

). The ten-minute mixing of samples was substituted with incubation in a water bath. The following factors were analyzed: material-NADES4 ratio (1:5, 1:10, 1:15 m/v), incubation temperature (40, 60, and 80 °C), and incubation time (10, 30, and 50 min). The phenolic content and antioxidant activity of the extracts were measured as detailed in Section 2.6 and Section 2.7, respectively.

3. Results and Discussion

3.1. Screening of NADES for Phytochemical Recovery Across Diverse Plant Matrices

The current study evaluated five natural deep eutectic solvents (NADESs) formulated with choline chloride and other natural compounds. These synthesized NADESs were utilized in extraction protocols and compared with traditional solvents such as ethanol, acetone, and water for recovering bioactive compounds. The efficiency of solvents in extracting red (betanin) and yellow (vulgaxanthin) pigments from dried beetroot was evaluated, as shown in Figure 1. Among the tested solvents, NADES2 exhibited the highest extraction efficiency for both pigments, yielding 179.86 mg of betanin/100 g d.m. and 170.77 mg of vulgaxanthin/100 g d.m. For red pigment extraction, NADES4 (163.73 mg/100 g d.m.) also exhibited strong performance without any statistical differences. NADES5 (137.23 mg/100 g d.m.), conventional water extraction (125.40 mg/100 g d.m.), and NADES1 (119.31 mg/100 g d.m.) showed moderate efficacy, while NADES3 (68.11 mg/100 g d.m.) was the least effective. In yellow pigment extraction, NADES4 (134.82 mg/100 g d.m.) and NADES5 (129.01 mg/100 g d.m.) followed NADES2 in performance, and water (100.38 mg/100 g d.m.) outperformed NADES1 (83.73 mg/100 g d.m.), while NADES3 (24.42 mg/100 g d.m.) again ranked lowest, highlighting its poor suitability for pigment recovery.

Other research also investigated betalain extraction methods from beetroot. Hernández-Aguirre et al. [24] created a deep eutectic solvent (DES) by combining magnesium chloride hexahydrate with urea and tested this at pH levels of 3 and 7, comparing it to water under the same conditions. They found that the DES (2:1 ratio, pH 3) facilitated the highest recovery of betalain at 3.99 mg/g, compared to 3.55 mg/g for water [24].

In a study conducted by Kaba et al. [25], the authors prepared 10 distinct DES mixtures using choline chloride or various hydrogen bond acceptors, including ascorbic, lactic, or malic acids, and combined them with hydrogen bond donors like ethylene glycol, butanediol, acetic acid, glycerol, glucose, sorbitol, and xylitol. The obtained mixtures were utilized for extracting betalains from beetroot. The central composite design was used to optimize the recovery protocol. A choline chloride–glucose combination with a molar ratio of 1:0.75, a water content of 30.83%, and a temperature of 30 °C produced 1192.17 mg/kg of betalain, 738.83 mg/kg of betacyanin, and 453.34 mg/kg of betaxanthin. Moreover, the authors recognized that the stability and bioaccessibility of betalains extracted using NADESs have improved.

The leafy Amaranth (Amaranthus L.) is another example of a plant rich in betalains. Due to the low extraction yield with conventional solvents, Gupta et al. [26] developed a green method utilizing ultrasound-assisted extraction with NADESs. The combination of lactic acid and fructose outperformed four other NADESs, yielding 79.68 mg of betalains per 100 g of fresh weight. A significant yield enhancement was observed following the use of ultrasonication, achieving 297.28 mg per 100 g, which represents a 12-fold increase over conventional methods and 3.7 times higher than the NADES alone. While early DES systems demonstrated betalain extraction potential, the NADES offers distinct advantages, including biocompatibility and alignment with green chemistry principles.

The efficiency of solvents in extracting carotenoids from dried carrots and sweet potatoes is presented in Figure 2. Unfortunately, acetone outperformed all other solvents, yielding the highest carotenoid content for both matrices: 0.020 mg/g in carrots and 0.040 mg/g in sweet potatoes. These values significantly exceeded those of NADESs and water. Among the NADES formulations, NADES4 demonstrated moderate efficacy, extracting 0.009 mg/g from carrots and 0.018 mg/g from sweet potatoes. Water showed limited efficiency, yielding 0.015 mg/g for both matrices, and it outperformed most NADESs. The results indicate that acetone is superior, likely because of its more non-polar properties compared to the NADES, which aligns with carotenoid solubility. While NADES4 showed promise, its lower efficiency compared to acetone suggests room for formulation refinement. The substantial underperformance of NADES1, NADES2, NADES3, and NADES5 underscores the importance of solvent composition in tailoring extraction protocols.

Carotenoids, such as β-carotene, are highly non-polar compounds requiring solvents with compatible polarity to facilitate dissolution. Polar NADESs, composed of hydrophilic components like choline chloride, urea, and organic acids, form hydrogen-bonding networks that favor interactions with polar phytochemicals but exhibit poor affinity for hydrophobic molecules. This poor performance stems from a hydrophobicity mismatch, as carotenoids require solvents with non-polar affinity. Researchers have utilized a range of eutectic mixtures, known as HDESs (hydrophobic deep eutectic solvents), for extracting lipophilic compounds from various food matrices, including pumpkin (Cucurbita maxima), carrot peels, and Citrus sinensis peels [27,28,29].

Stupar et al. [27] introduced fatty acid-based and menthol–fatty acid eutectic solvents to enhance the solubility and extraction efficiency of β-carotene. Among the twelve hydrophobic NADES formulations tested, a caprylic acid to capric acid ratio of 3:1 demonstrated the best performance, i.e., β-carotene solubility of 200.77 µg/mL and an extraction efficiency of 96.74 µg/mL. By combining this formulation with ultrasound, the extraction parameters for carotenoids from pumpkin were optimized using the Response Surface Methodology (RSM) and the Artificial Neural Network (ANN). The highest β-carotene yield of 151.41 µg/mL was obtained at 50 °C, utilizing 60% ultrasonic power and a solvent-to-solid ratio of 7 mL/g. Additionally, the authors discussed the concept of switchable solvents, specifically by the addition of ammonium hydroxide, which transformed the NADESs from hydrophobic to hydrophilic. This alteration facilitated carotenoid precipitation, achieving a 90% recovery rate, with β-carotene as the primary compound recovered at 52.25% [27].

Menthol and other terpenoids, such as thymol and camphor, were also used as part of hydrophobic deep eutectic mixtures to extract carotenoids from carrot peels. In the initial studies, the combination of thymol and menthol showed 1.281–1.338 mg of β-carotene/mL of total carotenoids; however, it was surpassed by acetone, which measured 1.612 mg/mL. Then, a Box–Behnken design was used to optimize extraction parameters: a thymol/DL-menthol molar ratio of 1:4, a solvent-to-sample ratio of 10:1 (v/w), and an extraction time of 30 min. This method produced 3.545 mg/mL of total carotenoids, which includes 3.341 mg/mL of β-carotene and 0.049 mg/mL of lutein, surpassing acetone in terms of efficiency [28]. Additionally, the menthol–eucalyptol and lauric acid–octanoic acid HDESs obtained by Viñas-Ospino et al. [29] demonstrated significantly higher total carotenoid content in ultrasound-assisted extraction compared to conventional solvents such as hexane, acetonitrile, methanol, and ethanol.

According to the examples given above, hydrophobic deep eutectic solvents (HDESs) exhibit high affinity to non-polar carotenoids (such as β-carotene and lutein) because of their hydrophobic characteristics and components like fatty acids or terpenoids. Moreover, HDESs serve as a protective matrix, stabilizing carotenoids against oxidative degradation during extraction and storage. Additionally, they can switch between hydrophobic and hydrophilic states, simplifying carotenoid recovery and solvent reuse while minimizing energy and resource consumption [27,28,29].

The next group of bioactive compounds, anthocyanins, was also tested in this study. The selection of solvents significantly influenced anthocyanin recovery (expressed as mg of cyanidin-3-glucoside/g d.m.) from chokeberry pomace and red onion, as shown in Figure 3. The mixture of ethanol and hydrochloric acid demonstrated exceptional efficacy, yielding the highest anthocyanin content in both matrices: 0.212 mg/g for chokeberry pomace and 0.012 mg/g for red onion. For chokeberry pomace, NADES4 was identified as the second most effective solvent (0.104 mg/g), surpassing conventional solvents such as water (0.049 mg/g) and ethanol (0.050 mg/g). In contrast, NADES5 (0.032 mg/g) and NADES1 (0.034 mg/g) showed the lowest extraction capacities. In red onion, the most effective solvents were water (0.009 mg/g) and EtOH-HCl (0.012 mg/g). Most NADES formulations, such as NADES2 (0.005 mg/g) and NADES3 (0.004 mg/g), demonstrated limited effectiveness. Ethanol used alone (0.004 mg/g) also showed poor results, underscoring the essential role of HCl in improving anthocyanin extraction. As noted in a different study, solvent acidification is essential for maintaining the structural integrity of anthocyanins, as the flavylium cation—responsible for their bright red color—only exists in acidic environments (pH < 3), which ensures both color intensity and stability [4].

Similar to other compound groups, the use of green solvents, particularly NADESs, enhances the extraction efficiency of anthocyanins, as well as their stability, bioavailability, and antioxidant activity. According to the review conducted by Foroutani et al. [30], the eutectic mixture of choline chloride and oxalic acid is identified as the most effective solvent for extracting anthocyanins. The authors also suggest that organic acid-based eutectic solvents outperform other solvents in extracting polar anthocyanins. In this study, one of the NADES tested comprised citric acid and choline chloride, yielding results similar to ethanol, although still lower than those of acidified ethanol and NADES based on choline chloride combined with glycerol or urea.

Zannou et al. [31] also utilized a choline chloride–glycerol NADES to extract anthocyanins from borage (Echium amoenum) flowers. The results showed that this mixture was much more effective than conventional solvents like water, methanol, or ethanol. It achieved the highest levels of total anthocyanin, total phenolic, total flavonoid content, and antioxidant activity. Intriguingly, Zannou and Koca [32] found that using choline chloride–urea NADES to extract blackberry anthocyanins did not result in detectable anthocyanins in the final extract. This outcome is due to the degradation of anthocyanins at elevated pH levels (pH > 7) associated with this NADES, where the mixture has a pH of 9.99. The research showed that choline chloride–urea and choline chloride–glycerol NADES achieved greater values than water or ethanol when assessing chokeberry pomace, despite having high pH levels of 9.08 and a moderate acidity of 5.02, respectively.

The effectiveness of solvents for extracting total phenolic content was also evaluated and showed significant variation across different agro-food matrices, including spent coffee grounds (SCGs), hop cones (HCs), rowanberries (RBs), and honeysuckle flowers (LJFs), as illustrated in Figure 4. The 80% ethanol and NADES4 were found to be the most effective, though their efficiency varied across different matrices. For SCGs, the highest values were observed with water (20.96 mg/g), 80% EtOH (20.81 mg/g), and NADES1 (20.66 mg/g), while NADES2 (14.81 mg/g) demonstrated the least efficacy. However, all results were not statistically significant from one another.

In hop cones, 80% ethanol resulted in the highest total phenolic content (TPC) of 32.11 mg/g. NADES4 was moderately effective, providing 27.33 mg/g, while traditional water extraction produced lower yields of 20.43 mg/g. For rowanberries, both 80% ethanol (28.55 mg/g) and NADES4 (25.02 mg/g) were the most effective, with NADES2 (24.51 mg/g) also showing promise. Water extraction yielded a lower content at 16.14 mg/g, followed by NADES5 (15.97 mg/g) and NADES1 (13.98 mg/g).

The most interesting results were achieved for L. japonica flowers. This case showcased significant solvent-dependent variability, where NADES4 (46.76 mg/g), 80% ethanol (46.53 mg/g), and NADES1 (42.94 mg/g) produced comparable yields. Notably, all NADESs surpassed the yield from conventional water extraction (26.95 mg/g).

The choice of a variety of botanically diverse materials—spanning from flowers (L. japonica) to root vegetables (like beetroot and carrot) and fruit byproducts (such as chokeberry pomace)—aimed to systematically evaluate the effectiveness of NADES across different phytochemical categories. Despite being taxonomically diverse, these materials are significant natural sources of polyphenols, pigments, and antioxidants. This study assesses the performance of NADESs on polar compounds, including anthocyanins and betalains, as well as non-polar carotenoids found in carrots and sweet potatoes, illustrating their adjustable solvation characteristics. Notably, L. japonica flowers emerged as a priority for optimization due to their high phenolic yield in preliminary screening and industrial relevance as a traditional medicinal plant. This flexibility positions NADESs as adaptable solvent systems for processes involving varied botanical materials, reducing the need for multiple solvents. Future research may enhance NADES formulations for specific uses, yet the results highlight their potential as sustainable, versatile alternatives to traditional VOC-based methods.

3.2. Optimization of NADES4 Aqueous Dilution for Enhanced Polyphenol Recovery from L. japonica Flowers

Using L. japonica flowers as a model system for polyphenol extraction optimization, this study evaluated NADES4 (choline chloride-urea, 1:2 molar ratio) as an environmentally friendly alternative to 80% ethanol, a conventional solvent for phenolic recovery. Consequently, the polyphenol-rich material and the eutectic mixture composed of choline chloride and urea were utilized in the next phase of this research. Research indicates that the high viscosity of eutectic mixtures can significantly hinder the extraction of bioactive compounds [32,33]. Therefore, additional experiments were conducted to compare the effects of adding water to NADES4 for increasing the yield of phenolic compounds (Figure 5). The optimization of NADES4 dilution significantly influenced TPC recovery from L. japonica flowers. The 60% (w/w) NADES4 achieved the highest TPC (57.46 mg/g), surpassing both 80% ethanol by 23% and water by 113%. This peak performance at 60% dilution suggests a critical balance between solvent polarity and viscosity, where moderate water content enhances phenolic solubility while mitigating NADES4’s inherent viscosity limitations.

Additionally, the TPC of NADES4 extracts consistently increased at high dilutions (10–40%), reaching 53.30 mg/g at 40%, which surpassed the performance of ethanol. Conversely, at lower dilutions (70–100%), TPC sharply decreased (from 48.18 to 44.42 mg/g), suggesting that higher concentrations of NADES4 lead to reduced solvent efficiency, likely due to increased viscosity that hinders mass transfer.

This study phase was further enhanced by measuring antioxidant activity using the DPPH radical scavenging assay (Figure 6). At concentrations of 50% and 60% (w/w), NADES4 demonstrated the most significant antioxidant activities, measuring 153.69 and 151.37 μmol of Trolox/g, respectively. This performance surpasses that of 80% ethanol by almost 20% and water by about 150%. The optimal antioxidant activity observed at 50–60% dilution coincides with the highest TPC at 60% NADES4 (Figure 5), indicating a strong correlation between antioxidant activity and phenolic content.

Other authors have also improved the extraction of antioxidants from L. japonica using DESs. Deng et al. [33] optimized the composition of deep eutectic solvents (DESs), specifically by using tetramethylammonium chloride–ethylene glycol in a 1:3 molar ratio, along with the extraction conditions. The extracts obtained from our proposed method demonstrated enhanced antioxidant activity, yielding between 229.1 and 249.1 μmol of Trolox/g d.m. Additionally, they contained higher antioxidant levels, with total phenolic content ranging from 34.2 to 36.5 mg of GAE/g d.m., all achieved in just 5 min without using organic solvents. Moreover, the authors recognized that these results greatly exceed those achieved using the extraction method specified in the Chinese Pharmacopeia with methanol. Furthermore, Liu et al. [34] assessed the extraction of phenylpropanoids from L. macranthoides using 32 NADESs and also optimized the water content in the L-proline–DL-lactic acid-based NADES, identified as the best extraction solvent. Their analysis of water additions ranging from 0 to 40% revealed that 30% water was the optimal amount, enabling the extraction of approximately 90 mg/g.

3.3. Process Parameter Optimization Using Response Surface Methodology

Table 3. Experimental matrix for the three-variable Box–Behnken design with the total phenolic content (mg of chlorogenic acid/g d.m.) and antioxidant activity (µmol of Trolox/g d.m.) in L. japonica flower extracts.

Exp. No.	Temperature (°C)	Time (min)	Liquid/Solid Ratio (v/w)	TPC *	TPC (Approximated)	AA	AA (Approximated)
1	40	10	10	64.57 ± 6.22	64.57	278.49 ± 21.95	278.49
2	80	10	10	59.51 ± 5.66	59.51	280.60 ± 11.55	280.60
3	40	50	10	64.97 ± 6.34	64.97	284.93 ± 8.72	284.93
4	80	50	10	68.04 ± 5.37	68.04	326.43 ± 28.55	326.43
5	40	30	5	62.42 ± 4.41	62.42	204.90 ± 23.93	204.90
6	80	30	5	66.03 ± 1.43	66.03	248.81 ± 2.30	248.81
7	40	30	15	68.43 ± 3.96	68.43	348.62 ± 20.79	348.62
8	80	30	15	75.94 ± 2.94	75.94	418.29 ± 29.32	418.29
9	60	10	5	58.38 ± 2.17	58.38	240.76 ± 4.55	240.76
10	60	50	5	69.31 ± 5.70	69.31	248.37 ± 2.51	248.37
11	60	10	15	74.75 ± 2.66	74.75	451.00 ± 9.46	451.00
12	60	50	15	72.95 ± 6.66	72.95	345.27 ± 19.99	345.27
13	60	30	10	69.46 ± 4.82	69.96	328.53 ± 8.59	336.09
14	60	30	10	69.81 ± 7.18	69.96	338.20 ± 13.93	336.09
15	60	30	10	70.60 ± 2.29	69.96	341.54 ± 30.97	336.09

* TPC—total phenolic content (mg of chlorogenic acid/g d.m.), AA—antioxidant activity (µmol of Trolox/g d.m.).

presents the experimental matrix for the three-variable, three-level Box–Behnken design, which assesses how temperature (°C), time (min), and the liquid/solid ratio (v/w) influence TPC and AA in L. japonica flower extracts. The experimental matrix showed notable differences in TPC, ranging from 58.38 to 75.94 mg/g and antioxidant activity (AA) from 204.90 to 451.00 µmol of Trolox/g across the tested parameters. The highest TPC (75.94 mg/g) and AA (451.00 µmol/g) were obtained under specific conditions: TPC reached its maximum at 80 °C for 30 min with a 15 v/w ratio (Exp. 8), while AA peaked at 60 °C for 10 min with the same ratio (Exp. 11), emphasizing the critical importance of solvent accessibility during extended heating.

The Pareto charts (Figure 7) highlight the standardized effects of individual factors and their interactions on the TPC and antioxidant activity. The alignment between the experimental model and the observed total phenolic content (TPC) and antioxidant activity was assessed for the linear-quadratic model (Tables S1–S4, Figure S1) using the coefficient of determination, known as R². In this context, R² values of 0.9981 and 0.9985 were obtained, respectively. After adjustment, the R²_Adj values dropped to 0.9870 and 0.9900; nevertheless, these figures still indicated a strong fit of the model with the collected data. For total phenolic content, the liquid/solid ratio (linear term, 19.850) was identified as the key factor, having a significant positive influence on TPC. Following this, the quadratic effect of temperature (10.446) and the linear effect of incubation time (10.338) were observed. Importantly, interactions among these factors also had significant impacts. The antagonistic interaction between the linear effects of incubation time (2 L) and the liquid/solid ratio (3 L) (2 L vs. 3 L: −10.888) suggested that increasing both parameters simultaneously may lower TPC. Likewise, for antioxidant activity, the liquid/solid ratio (linear effect) emerged as the key factor boosting antioxidant activity in L. japonica extracts with 60% NADES4 (30.908). The temperature (quadratic effect, 8.489) revealed a non-linear optimal range, whereas the interaction between incubation time and solvent ratio adversely affected yields (−8.391).

The three-dimensional response surface and contour plots (Figure 8) elucidate the interactive effects of process variables on total phenolic content (TPC) in L. japonica flower extracts using 60% NADES4. Temperature showed non-linear effects, with an intermediate temperature of approximately 60 °C yielding better results than extremes like 80 °C (Exp. 2: 59.51 mg/g), indicating that thermal degradation occurs at higher temperatures. However, the combination of moderate time (35–45 min) and high temperature (65–75 °C) suggests a synergistic effect that enhances phenolic solubility and recovery under these conditions. Figure 8 further validates the findings illustrated by the Pareto charts (Figure 7), highlighting that the solvent-to-extracted material ratio is vital for the effective extraction of phenolic compounds. High liquid-to-solid ratios (15 v/w) combined with moderate time and temperature enhanced the TPC. In contrast, low ratios (5–10 v/w) decreased TPC to 58–63 mg/g, even at elevated temperatures, indicating limited solvent accessibility. Additionally, shorter processing times (<20 min) at high ratios resulted in a TPC reduction to below 63 mg/g.

Regarding the antioxidant activity (Figure 9), the peak AA (450–480 µmol/g) was reached at temperatures between 70 and 80 °C for 10 to 30 min and solvent ratios of 12 to 16 v/w. Lower ratios (5–8 v/w) or temperatures greater than 85 °C led to reduced AA levels (<220 µmol/g), again because of limited solvent penetration or thermal degradation. Utilizing a high solvent volume (12–15 v/w) with increased heat (70–80 °C) maximized yields; however, prolonging the duration beyond 50 min at low ratios reduced efficiency.

In conclusion, the contour plots shown in Figure 10 indicate that the optimal parameters for maximizing total phenolic content and high antioxidant activity are a solvent-to-flower ratio of 12–15 v/w, a temperature range of 55–75 °C, and a moderate incubation time of 10–30 min.

Fan et al. [35] explored betaine-based DESs for extracting chlorogenic acid from L. japonica. They identified a betaine–glycerol mixture as the optimal DES for enhancing CGA extraction through the Box–Behnken design. The study revealed favorable conditions: a DES with 40% water content, a 60 min extraction time at 90 °C, and a liquid-to-solid ratio of 30:1, resulting in a CGA yield of 37.28 ± 0.64 mg/g, which is 1.9 times greater than that obtained using ethanol. Both studies, including this one, demonstrate that incorporating 40% water reaches the target viscosity, and, additionally, for elevated phenolic values, a high liquid-to-solid ratio is essential. Furthermore, the authors considered that the extraction mechanism of CGA using betaine-based DESs indicated that systems with higher CGA extraction yields showed greater interaction energy and an increased number of hydrogen bonds.

A three-level, three-factor Box–Behnken design was also applied by Peng et al. [36], whereas the liquid/solid ratio, temperature, and time were optimized for the extraction of five phenolic acids from L. japonica. The authors confirmed that DESs are highly effective for extracting phenolic acids due to their ability to enhance solubility. A DES composed of 1,3-butanediol and choline chloride in a 6:1 molar ratio demonstrated superior extraction efficiency. The ideal conditions included 10% water content in the DES, a liquid-to-solid ratio of 9 mL/g, an extraction temperature of 60 °C, and a processing time of 20 min. Under these parameters, the yields for chlorogenic, caffeic, 3,4-dicaffeoylquinic, 3,5-dicaffeoylquinic, and 4,5-dicaffeoylquinic acids were 26.07, 0.15, 0.93, 23.67, and 8.85 mg/g, respectively.

The potential for reusing and recycling NADESs in industrial applications is essential for sustainable growth. Research indicates that NADESs can be regenerated using methods such as anti-solvent precipitation, membrane filtration, and liquid–liquid extraction. Of these, anti-solvent addition and density-based separation boast high recovery efficiency but are energy-intensive, whereas membrane filtration and extraction methods provide varied scalability at different costs [37]. Many NADES precursors, including choline chloride, citric acid, and glycerol, are GRAS-certified, satisfying food and pharmaceutical standards. The NADESs derived from these precursors show low-to-moderate cytotoxicity and minimal phytotoxicity, and several are noted for their biodegradability [38].

4. Conclusions

This study demonstrates the efficacy of the natural deep eutectic solvents (NADESs) as sustainable alternatives to conventional solvents for recovering diverse bioactive compounds. Notably, the choline chloride–urea system (NADES4) emerged as a particularly effective solvent, yielding L. japonica extracts with elevated polyphenol levels (75.94 mg/g) and robust antioxidant capacity (451.00 µmol of Trolox/g) when optimized for water content (60:40 NADES–water mixture) and process parameters (15:1 solvent ratio, 80 °C, 30 min). Apart from polyphenols, NADES2 exhibited superior performance in extracting betalains from beetroot (179.86 mg of betanin/100 g), while NADES4 proved to be adaptable to anthocyanin-rich matrices such as chokeberry pomace. Although NADESs demonstrated limited efficacy for carotenoids compared to acetone, their tunable properties (e.g., polarity modulation through water addition) underscore their potential for polar phytochemicals. These findings establish NADESs as versatile, eco-friendly solvents capable of replacing volatile organic compounds in industrial workflows. This study advances NADES research by utilizing a comprehensive formulation and optimization strategy, evaluating the effectiveness of NADESs with a variety of polar and non-polar phytochemicals and illustrating their versatility for diverse industrial feedstocks. In contrast to studies that only compare NADESs with a single solvent, this research compares them to both water and ethanol or acetone. These improvements position NADESs as practical solvents for recovering multiple phytochemicals, connecting green chemistry innovation with commercial scalability.