Multifactorial analysis of temperature, solute-to-solvent ratio, and ultrasound amplitude on the extraction of phenolic and antioxidant compounds from Aloysia citriodora Palau leaves

Collection of A. citriodora leaves

The leaves were collected from the cedrón plant (Aloysia citriodora) in August 2022, during the summer season, at the following coordinates: latitude 19.6912985, longitude −101.2054516, and an altitude of 1,909 m, in Morelia, Michoacán, Mexico. The age of the plant is estimated to be approximately 25 years. The specimen was taxonomically identified using the database of the Herbarium of the Universidad Michoacana de San Nicolás de Hidalgo. Taxonomic identification was based on a comprehensive analysis of the leaves, flowers, and fruits. The leaves were subjected to shade-drying at room temperature for a period of 15 days. Prior to the drying process, a meticulous examination of the plant material was conducted to guarantee its freedom from pests or indications of physiological stress. Subsequently, the samples were ground and sieved using a 0.3331 inch mesh.

Phenolic extracts: full factorial design

The conditions used to extract phenolic compounds in this study were based on previous studies that evaluated various ultrasound-assisted extraction parameters. According to these studies, the temperature range was from 30 to 90 °C, the solute-to-solvent ratio was from 1:5 to 1:100, the extraction time was from 10 min to 48 h, and the ultrasound amplitude was from 30% to 50% (Tena-Rojas et al., 2022; Tranquilino-Rodríguez & Martínez-Flores, 2023).

The extracts were obtained using a VCX 500 ultrasound-assisted extraction system (Sonics & Materials, Inc.) operating at 20 kHz with a 220-B type probe. A full factorial design was implemented to identify optimal conditions for the ultrasound process (Table 1). This design incorporated the following variables—ratio of dried A. citriodora leaves to solvent: 1:5, 1:10, and 1:15 (where 1:5 equals 1 g in 5 mL), temperature: 30 °C, 50 °C, and 75 °C and amplitude: 20%, 25%, and 35%. The extraction time was set at 10 min based on preliminary tests and in data found in the literature. For example, studies on Carica papaya Linn leaves reported that the optimal extraction time was 9 min (Puramshetti et al., 2025). Zerihun-Chala et al. (2024) also found that 10 min was optimal for extracting phenolic compounds from A. remota Benth leaves. Furthermore, numerous reports have demonstrated that temperatures exceeding 85 °C and extended exposure periods can result in the degradation of target compounds. This study evaluated solute/solvent ratios of 1:5, 1:10, and 1:15 (g/mL) for polyphenolic compound extraction to determine the most efficient ratio in terms of yield and phenolic content. This approach is supported by previous studies in this field. For instance, Rahmawati et al. (2024) reported that a ratio of 1:8 (g/mL) was optimal for the microwave-assisted extraction of polyphenols from banana peels, with concentrations reaching up to 354.02 mg GAE/g.

Additionally, Jeganathan et al. (2014) identified an optimal ratio of 1:8.7 g/mL for red grapes, which had a substantial impact on anthocyanins and flavonoids. Similarly, Shabri & Rohdiana (2016) found that a 1:15 (w/v) ratio produced the best polyphenol extraction from green tea with 70% acetone. The collective findings of these studies validate the selected range (1:5–1:15) as appropriate for evaluating extraction efficiency.

The amplitude percentage was determined based on the operational limits of the equipment, ensuring that it was not lower than 20% nor higher than 40%. The experimental workflow diagram is provided in the Material S4. The extracts were filtered using Whatman No. 1 paper and stored at −20 °C in amber glass containers. This method was employed to minimize light exposure and prevent oxidation. All extractions were carried out under controlled temperature conditions. The extracts were immediately cooled in an ice bath after extraction to prevent thermal degradation of phenolic compounds. The following response variables were employed to determine the most effective combination: quantification of total phenolic and flavonoid content and antioxidant activity of the extracts against ABTS^•+ (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)), DPPH^• (2,2-diphenyl-1-picrylhydrazyl), and hydroxyl radicals (OH^•).

Determination of phenolic compounds and antioxidant tests

Statistical analysis

The data were analyzed using a one-way analysis of variance (ANOVA) with JMP software (version 8.0). Tukey’s multiple comparison test was then performed to determine statistical differences in the parameters. Values of P ≤ 0.05 were considered statistically significant.

Sample identification

The specimen obtained was a deciduous shrub, measuring between two and six m in height, characterized by a rough texture and a lemon-like fragrance. The leaves are characterized by a blade measuring 30–110 × 7–25 mm, 3-verticillate trichomes, and stiff, antrorose trichomes with a diameter of 0.1–0.2 mm (see Fig. 1A). The calyx measures 2–3.5 mm, while the corolla measures 3.5–5.5 mm, with a white or purple coloration, as shown in Fig. 1B. Figure 1C illustrates an obovoid or oblong fruit, with mericarps measuring 1.3–1.8 × 0.5–0.6 mm, flat-convex, hairy at the apex, and brown in color. A comparison was made with specimens of the same species that had been deposited in the herbarium of the Universidad Michoacana de San Nicolás de Hidalgo in Morelia, Michoacán, Mexico. These specimens had folio numbers 27,450 and 23,355, respectively. The comparison was also made with digital herbarium images, such as SEINET. The specimen was formally designated as Aloysia citriodora Palau.

Ultrasound-assisted extraction

The extracts were obtained using the VCX 500 Ultrasonic Processor with the extraction parameters considered in the experimental arrangement (Table 1). A 3³-factorial design with three factors and three levels was employed, yielding 27 extracts. The extraction time was 10 min. The nomenclature of each extract was selected according to Table, in which T denotes the temperature variable, with three levels numbered from 1 to 3, corresponding to 1 = 30 °C, 2 = 50 °C, and 3 = 75 °C. The variable leaf and water concentration is indicated by the letter R, which has three levels, numbered from 1 to 3, corresponding to 1 = 1:5, 2 = 1:10, and 3 = 1:15 (g/mL). The variable amplitude is designated by the letter A, with three levels numbered from 1 to 3, corresponding to 1 = 20%, 2 = 25%, and 3 = 35%. For instance, T1R1A1 corresponds to a temperature of 30 °C, a solute/solvent ratio of 1:5, and an amplitude of 20%.

Total phenolic and total flavonoid contents

The total phenolic content is shown in Table 1 and expressed in mg equivalents of gallic acid (mg GAE/g of extract). The sample with the highest amount was T3R2A2, which had 132.84 ± 3.12 mg GAE/g. In a related study, Wernert et al. (2009) obtained 51.85 mg of tannic acid equivalents/g dry material from aqueous extracts of A. citriodora. Al-Qaraleh & Tarawneh (2016) evaluated methanolic extracts of A. triphylla plants and yielded 62.16 mg GAE/g of plant extract. In the present study, the pine bark extract (Oligopin^®) exhibited a concentration of 10.90 ± 0.41 mg GAE/g. As indicated by the results of the study, multiple treatments exhibited a higher level of total phenolic compounds in comparison to the Oligopin sample.

The total flavonoid content is presented in Table 1. The sample with the highest amount was T2R3A3, which had 98.96 ± 31.13 mg QE/g; this sample yielded a higher amount of total flavonoid compounds compared to the 7.01 mg quercetin equivalents/g dry leaves of aqueous extracts of A. citriodora observed by another author (Moeina, Zarshenasc & Etemadfarda, 2014). When comparing the results obtained by Athanasiadis et al. (2024), who used ultrasound-assisted extraction in A. citriodora L. and reported values of 40.66–150.96 mg GAE/g with exposure times ranging from 20 to 150 min, similar values to those in the present study are observed, although with a shorter exposure time.

DPPH and ABTS scavenging activity

Gallic acid and ascorbic acid were utilized as positive controls, yielding optimal performance in terms of inhibition and concentration (see Table 1). The values obtained were 2.35 ± 0.09 µg/mL and 8.25 ± 0.18 µg/mL, respectively, indicating 50% inhibition of the DPPH^• radical. The aqueous extracts that exhibited a lower concentration for radical inhibition were T3R3A2, T3R2A2, and T3R3A1, with values of 35.58 ± 0.50, 47.90 ± 0.39, and 42.64 ± 0.58 µg/mL, respectively, to inhibit 50% of the DPPH^• radical (Table 1). The relationship between temperature and solute-to-solvent ratio is significant in the context of molecular extraction, particularly concerning its capacity to inhibit DPPH radical activity. In their study, Cheurfa & Allem (2015) evaluated the hydroalcoholic extract of A. citriodora, finding significantly higher total phenolic and flavonoid contents compared to the aqueous extract. However, the aqueous extract demonstrated a higher IC₅₀ value (27.40 ± 0.1 mg/mL) for scavenging the DPPH^• radical, indicating a potential difference in antioxidant activity between the two extracts. In contrast, Al-Qaraleh & Tarawneh (2016) obtained an IC₅₀ concentration of 0.8 mg/mL in A. triphylla leaf extracts, requiring a higher concentration to eliminate 50% of the radicals compared to the extracts obtained in this study using ultrasound-assisted extraction, which required a lower concentration of 35.58 µg/mL to achieve the same 50% radical elimination.

An ABTS^•+ radical inhibition curve was plotted using gallic acid and trolox as standards, yielding 50% inhibition of the ABTS radical at concentrations of 0.91 ± 0.02 and 0.12 ± 0.07 µg/mL, respectively (see Table 1 for details). Treatment T3R3A2 exhibited a lower concentration, with an inhibition rate of 9.1 ± 0.08 µg/mL, corresponding to 50% inhibition of the ABTS^•+ radical. A study by Hernández et al. (2015) examined the antioxidant properties of 20 plants utilized in Mexican traditional medicine. The study reported that the plants gobernadora, “hierba dulce”, and “nogal” exhibited the highest ABTS^•+ radical scavenging capacity, with IC₅₀ values of 1.67, 2.68, and 2.70 µg/mL, respectively. Conversely, higher concentrations were required to achieve IC₅₀ values of 6.90 and 11.18 µg/mL in rosemary and chamomile plants, respectively. The trolox equivalent activity (TEAC) of the “T3R3A2” extract was obtained with 343.52 ± 3.09 (mg Trolox/100 g leaves). The oral administration of verbascoside, a molecule identified in Aloysia citriodora, at a maximum dose of 3 mg/day as a dietary supplement in rabbits, demonstrated a protective effect on ocular tissue and fluids associated with oxidative stress. This protective effect was assessed by thiobarbituric acid reactive substances (TBARS) and the Trolox equivalent antioxidant capacity (TEAC) assay (Mosca et al., 2013). These results emphasize the necessity of future research on the extracts obtained in this study.

Hydroxyl radical scavenging activity

The hydroxyl radical inhibition curve was determined using Oligopin (Nutri-Dyn, Maple Plain, MN, USA), which is recognized for its antioxidant power derived from natural pine bark extracts, as the standard. The results of this study are presented in Table 1. Oligopin, at a concentration of 1.04 mg/mL, demonstrated the most effective performance in terms of inhibition and the concentration utilized. For the aqueous extract, which achieved a lower concentration, T3R3A2 demonstrated the most effective result, with 0.72 ± 0.007 mg/mL inhibiting 50% of the radical.

In a study by Valencia-Avilés et al. (2018), the inhibitory effect on the hydroxyl radical was investigated using oak extracts. The study found that the inhibition occurred at a concentration of 1,271 ± 72 µg/mL of Oligopin. In this study, the value obtained for Oligopin was 1.044 ± 0.04 mg/mL, which is similar to the values obtained for extract T3R3A2. According to the values obtained for the Trolox equivalent activity, there was higher activity at 75 °C and an amplitude of 35%, as well as a ratio of solvent to sample of 1:15. Research conducted by Mehmood et al. (2024) demonstrated that the methanolic extract of fresh P. roxburghii leaves exhibited the highest DPPH^• and OH radical scavenging activity, achieving 80% radical elimination with 200 µL of extract. Conversely, studies documented by Olakanmi et al. (2024) on the extraction with ethyl acetate from Croton Albizia ferruginea leaves yielded an IC₅₀ of 350 µg/mL for hydroxyl radical elimination, signifying elevated antioxidant activity in comparison to the present study, in which a value of 1.04 mg/mL was obtained. This difference could be attributed to the presence of lipophilic compounds, which may be present in the ethyl acetate extraction performed by Olakanmi et al. (2024). This phenomenon has also been observed in studies reported by Camarena-Tello et al. (2018). The researchers found an IC₅₀ in Psidium guajava leaf extracts of 1,103.42 µg/mL using chloroform. However, an IC₅₀ of 1,726.51 µg/mL was found using water for extraction, indicating that chloroform extraction is more efficient.

In addition to their demonstrated antioxidant value, the phenolic extracts of Aloysia citriodora show great potential for industrial applications. Their use in pharmaceutical and nutraceutical formulations has been documented, owing to their capacity to modulate oxidative, inflammatory, and microbial processes (Athanasiadis et al., 2024). In the field of food science, these extracts have demonstrated prebiotic and antimicrobial properties, which are advantageous for the development of functional foods and natural preservatives. These extracts have been shown to stimulate the growth of beneficial bacteria, such as Lactobacillus rhamnosus, and to suppress common pathogens (Gkalpinos et al., 2023). From a technological standpoint, ultrasound-assisted extraction (UAE) is not only efficient and reproducible but also easily scalable to an industrial level, with lower solvent and energy consumption, making it a more environmentally sustainable option compared to traditional methods (Hashemi et al., 2017). In a study conducted by Buchwald-Werner et al. (2018), the intake of 400 mg/day of A. citriodora extract (Recoverben^®) led to a significant reduction in muscle strength loss in healthy, moderately active adults. These elements reinforce the applicability of the optimized extract in key sectors of the bioindustry. Although there are currently no studies of scale-up of ultrasound extraction using Aloysia citriodora leaves, numerous reports are available for other species relevant to food studies. Various studies support the scalability of UAE for phenolic compounds. Rodríguez et al. (2022) successfully scaled UAE to pilot scale (120 kg) for extracting polyphenols from olive pomace, achieving high yields (3.0 g GAE/L) under controlled conditions. Panić et al. (2024) demonstrated that UAE with natural deep eutectic solvents produces stable bioactive extracts suitable for functional foods and cosmetics. These findings support the industrial feasibility of applying UAE for obtaining Aloysia citriodora extracts.

Multifactorial analysis

The following multifactorial analysis of variance (ANOVA) was conducted to determine whether there was a relationship between response variable Y and certain categorical variables, which were the factors. The response variables included the phenolic compound content (mg GAE/g), flavonoid compound content (mg QE/g), and the IC₅₀ of DPPH^•, ABTS^•+, and hydroxyl radicals. The results of this study are shown in Table 2.

A three-way analysis of variance (ANOVA) was conducted on the data obtained from the 3³ experimental arrangements. A subsequent statistical analysis employing one-way analysis of variance (ANOVA) revealed discrepancies in the levels of phenolic compounds, flavonoid compounds, and the IC₅₀ values of hydroxyl, ABTS^•+, and DPPH^• radicals. The statistical significance of these findings was determined to be below the α = 0.05 level of significance, as shown in Table 2. Consequently, the null hypothesis was refuted, and the alternative hypothesis proposing a significant discrepancy in the extraction of flavonoids from A. citriodora by the ultrasound extraction method, contingent on temperature, solute/solvent ratio, and amplitude, was accepted.

As is shown in Table 2, an increase in temperature resulted in a corresponding increase in the extraction of total phenolic content (mg GAE/g). The temperature of 75 °C was found to yield the highest extraction yield. It was observed that a higher total flavonoid content (quercetin equivalents) was obtained at 50 °C, which corresponded to the ability to inhibit the hydroxyl radical at the same temperature. An estimated concentration of 1.65 mg/mL was required to eliminate 50% of the radical. In a similar manner, the other radicals exhibited superior IC₅₀ values at 50 °C. A correlation was identified between quercetin equivalents (QE) and the capacity to inhibit the radical. It was determined that an increase in the QE mg QE/g content of the extract was associated with a decrease in the amount of mg/mL of extract necessary to eliminate the hydroxyl radical. Several studies have demonstrated that extraction temperature exerts a substantial influence on the recovery of phenolic compounds. As reported by Ozdemir et al. (2023), an increase in total phenolic content up to 70 °C was observed. This increase was attributed to greater cell disruption induced by ultrasonic cavitation. Zampar et al. (2022) examined the polyphenol content of compounds derived from pineapple byproducts. They found a maximum content of 405 mg GAE/100 g of compound at 60 °C. This result suggests that higher temperatures enhance the solubility and diffusion of phenolic compounds. In the study by Oliveira et al. (2022), the researchers evaluated the effectiveness of ultrasound-assisted extraction and pressurized liquid extraction in Citrus latifolia Tan. It has been observed that elevated temperatures can enhance process efficiency, provided that the thermal degradation threshold is not exceeded. In the study, the highest yields of total phenolic compounds were obtained at 110 °C (p < 0.05), with a value of 17.66 mg GAE/g of dry weight. Leliana et al. (2022) have indicated that the optimal conditions for metabolite extraction from young coconut mesocarp using UAE entail a temperature of 70 °C and an extraction time of 5 min. In accordance with the data presented above, it is reaffirmed and emphasized that, for the present study, the optimal temperature for polyphenol extraction was 75 °C.

As illustrated in Fig. 2, the r values (least squares means) were derived from the multiple linear regression equation. The highest r values (total phenolic content of 92.70 mg GAE/g) were consistently obtained at a high temperature (75 °C), while the lowest (3.35 mg GAE/g) was obtained at a low temperature (30 °C). The contribution of the three factors (T, R, and A) was significant in yielding the highest r value (total phenolic content of 92.70 mg GAE/g). According to the findings of several studies, the optimal temperature range for extraction methods aimed at achieving the highest recovery of polyphenols is between 60 and 80 °C. In this study, it was observed that the highest antioxidant activity occurred within this temperature range (Gironi & Piemonte, 2011; Antony & Farid, 2022).

As shown in Table 2, the solute-to-solvent ratio exerts a significant influence on the extraction of phenolic and flavonoid compounds. The ratio of 1:10 (g/mL) yielded the highest phenolic content (gallic acid equivalents extraction), while a ratio of 1:15 yielded a higher flavonoid content (quercetin equivalents). The 1:10 ratio resulted in a lower concentration, with the capacity to eliminate 50% of the radicals. The calculated r values from the multiple linear regression equation demonstrate the significance of the three factors (T, R, and A), yielding the highest r value (flavonoid content of 94.583 mg QE/g) (Fig. 3). This value was obtained at the highest temperature and ratio, with the amplitude having no significant influence and being at the lowest level to reach the highest r. Dzah et al. (2020) found that a low solvent–material ratio increased the dispersion and absorption of acoustic energy and attenuated the intensity of ultrasonic power, and vice versa. This effect influences the attraction of phenolic and antioxidant compounds. This phenomenon was observed in the present study, in which the extraction of phenolic compounds was found to be lower at higher organic matter concentrations. As stated by Jovanović et al. (2021), the presence of a higher amount of extraction solvent (1:30 ratio) has been shown to prevent medium saturation and improve polyphenol content. In contrast, an excessive amount of plant material (at a ratio of 1:10) has been shown to increase viscosity, which in turn inhibits the expansion of ultrasonic waves and the diffusion of polyphenols Jovanović et al. (2021). This finding is consistent with the individual analysis, as shown in Table 3, which demonstrates a significant difference in higher flavonoid content at a 1:10 ratio compared to the other dilutions examined in this study.

Table 2 shows a positive correlation exists between the amplitude and the extraction of total phenolic and total flavonoid compounds. The highest amplitude (35%) yielded the highest flavonoid content (equivalent to quercetin extraction). Regarding radicals, there were no differences between amplitudes of 25 and 35%; differences were observed only between the low amplitude (20%), which resulted in a higher concentration to eliminate 50% of the radicals. In a recent study, Babotă et al. (2022) reported an ultrasound amplitude of 34.8% for the extraction of phenolic compounds from Thymus comosus Heuff. In the present study, the optimal amplitude for the extraction of individual flavonoids was determined to be 35%. In a series of studies conducted within our research group (Tena-Rojas et al., 2022; Tranquilino-Rodríguez & Martínez-Flores, 2023; Garnica-Romo, Sanchez-Pahua & Martinez-Flores, 2024), the extraction of phenolic compounds from P. guajava, Moringa oleifera, and A. citriodora leaves was examined. This extraction was performed using an amplitude ranging from 30% to 50% during ultrasound-assisted extraction. The findings of these studies indicated optimal extraction yields of total phenolic compounds, total flavonoid compounds, and antioxidant capacity.

As illustrated in Fig. 4, the calculated r values from the multiple linear regression equation demonstrate that the contribution of the three factors (T, R, and A) was significant in obtaining a low r value (0.86 mg/mL). The value for the hydroxyl radical was obtained under conditions of elevated temperature and solute-to-solvent ratio. The amplitude also had an influence, being at the lowest level to obtain the r value. The graphs depicting DPPH^• and ABTS^•+ are not presented due to their lack of statistical significance in comparison to the hydroxyl radical data.

Table 3 presents the root mean square error (RMSE) and the coefficient of determination (R²). The model’s predictions regarding the hydroxyl radical and the phenolic content demonstrated a satisfactory degree of accuracy, as evidenced by the high degree of fit observed. The F-ratio values were found to be elevated, as evidenced by the high values observed in the case of ABTS^•+ (648.8), suggesting that the model exerted a significant influence on this particular response variable. The findings of the multifactorial analysis indicated that treatment T3R3A2, which corresponded to a temperature of 75 °C, an amplitude of 25%, and a solute/solvent ratio of 1:15, exhibited the optimal capacity to inhibit hydroxyl and ABTS radicals. Furthermore, the study demonstrated the highest total phenolic compound content and a significant flavonoid compound content, suggesting its potential as a promising therapeutic agent. In order to enhance the statistical interpretation and facilitate a more profound comprehension of the impact of independent variables (temperature, solute/solvent ratio, and ultrasonic amplitude), readers are advised to refer to the Supplemental Information, in which the prediction equations and fitting models for each response variable are provided, including flavonoid content, total polyphenol content, and free radical scavenging capacity (ABTS, DPPH, and hydroxyl radical).

Several studies have demonstrated the efficacy of green extraction technologies, such as microwave-assisted extraction (MAE), ultrasound-assisted extraction (UAE), and supercritical fluid extraction (SFE), in the recovery of phenolic compounds from plant matrices, including aromatic species such as Aloysia citriodora. These techniques have been shown to enhance extraction yield and reduce processing times, while also reducing or eliminating the use of toxic solvents. This aligns with principles of sustainability (Ameer, Shahbaz & Kwon, 2017). In this context, Leyva-Jiménez et al. (2019) evaluated the use of microwave-assisted extraction in Aloysia citriodora leaves, observing that the extraction time influenced the recovery of phenylpropanoids, with shorter times resulting in higher efficiency. In contrast, studies by Garnica-Romo, Sanchez-Pahua & Martinez-Flores (2024) on Aloysia citriodora leaves identified a robust correlation between the IC₅₀ values for DPPH^• and ABTS^•+ radical scavenging activity and the total phenolic content in the extracts. These studies also confirmed that ultrasound-assisted extraction produced higher yields in less time compared to maceration. Furthermore, Parodi et al. (2013) identified the presence of cinerolone, a type of ketoalcohol, in Aloysia citriodora leaf extracts obtained through supercritical fluid extraction. Finally, Athanasiadis et al. (2024) demonstrated that optimizing the extraction process of A. citriodora using 50% ethanol, high temperatures (80 °C), and technologies such as ultrasonication and pulsed electric fields (PEF) significantly enhances the recovery of bioactive compounds. These extracts, which are abundant in total phenols and antioxidants, demonstrate considerable promise for utilization in pharmaceutical and food applications. Additionally, the equations used to estimate the values shown in Figs. 2–4 are included in the Supplementary Material, along with contour plots that visualize the interactions between factors such as temperature, solute/solvent ratio, and amplitude.