The Optimisation of Ultrasound-Assisted Extraction for the Polyphenols Content and Antioxidant Activity on Sanguisorba officinalis L. Aerial Parts Using Response Surface Methodology

Abstract

The aim of this study was to optimise ultrasound-assisted extraction (UAE) of the herb Sanguisorba officinalis L. in terms of the antioxidant activity (DPPH and FRAP method) and total polyphenol content (TPC). Optimisation was performed using the response surface methodology (RSM) with a third-degree (3³) Central Composite Design (CCD) approach. The RSM was applied to obtain the optimal combination of (1) raw material content (2.25–7.5 g raw material/100 mL of solvent), (2) ethanol concentration (20–60% v/v), and (3) extraction time (1–15 min). The optimal conditions for the extraction of polyphenols and antioxidant potential were a raw material content of 7.5 g/100 mL of solvent (solid/solvent ratio 13.3 mL/g), an ethanol concentration of 47% v/v, and an extraction time of 10 min. At these optimal extraction parameters, the maximum extraction of polyphenols and antioxidant activity obtained experimentally was found to be very close to its predicted value and was 12.9 mmol Trolox/L (DPPH method), 19.4 mmol FeSO₄/L (FRAP method), and 2.1 g GA/L (TPC). The mathematical model developed was found to fit with the experimental data on the antioxidant potential and polyphenol extraction. The n-octanol/water partition coefficient of the optimised extract was used to determine their lipophilicity. Our studies have shown that the optimised extract is highly hydrophilic (log P < 0). Optimal parameters can be used for the industrial extraction of the S. officinalis herb for the needs of, among others, the pharmaceutical or cosmetic industry.

1. Introduction

Current ecological trends focus on reducing the negative environmental impact of extraction processes. Therefore, modern extraction methods are developed by the “green chemistry” concept. One of the modern techniques is ultrasound-assisted extraction (UAE), characterised by high efficiency, repeatability, and a shorter duration compared to conventional methods, which ultimately reduces process costs. Moreover, extracts obtained by this method are characterised by high purity [1,2]. The cavitation phenomenon is key in this process. The mechanical effect of ultrasound-induced acoustic cavitation enhances the surface area of the contact between solvents and plant materials, as well as the permeability of cell walls. There are two ways to carry out this process: directly with an ultrasonic probe and indirectly with an ultrasonic bath [3].

The efficiency of the extraction of biologically active compounds from plant material depends on many factors, including the method of preparation and concentration of the material used for the extraction process, the selection of solvent and isolation techniques for active compounds, as well as the choice of extraction parameters (time or temperature, among others). Therefore, the optimisation of the process parameters is crucial to obtain the highest possible content of the desired active compounds in the extract [4]. One of the most frequently used methods for optimising plant material extraction processes is the response surface methodology (RSM). This optimisation tool allows for the identification of interdependencies between variables used in research, including those related to plant material extraction or food technology. The RSM develops an experimental design that integrates independent variables and uses input data from the experiment to generate a series of equations that can determine a theoretical value for the output. The outputs result from a well-designed regression analysis grounded in controlled independent variable values. As a result, based on the new values of the independent variables, the dependent variable can be predicted [5]. The use of mathematical modelling allows for a reduction in the number of experimental samples compared to the standard one-dimensional approach, which has a direct impact on reducing the final cost of the process, among other things, by limiting the consumption of energy or reagents [1,6]. In addition to increasing the efficiency of laboratory testing, parameter estimation can help identify variables that significantly influence the model. To design an experiment in RSM, we can use, for example, the Box Behnken Design (BBD) and the Central Composite Design (CCD) [5]. The experimental points in the BBD are located on a hypersphere equidistant from the central point. In the CCD, we distinguish points of factorial design, axial points, and the central point [7]. The RSM-CCD model is successfully used in the optimisation of plant raw material extraction processes [8,9,10,11].

Sanguisorba officinalis L., called great burnet, is a plant belonging to the Rosaceae family. The most common raw material used in medicine is a root (Sanguisorbae radix). Because the aerial parts of S. officinalis are also a valuable source of biologically active compounds, the leaves, flowers, and stalks are increasingly assessed for their health-promoting properties [12,13,14,15]. The antioxidant effect of this plant is often emphasised, for which the phenolic compounds are mainly responsible [13,14,16,17]. Antioxidants are one of the methods of protecting the body against the negative effects of oxidative stress. This phenomenon causes damage to key structures of the body, such as proteins, nucleic acids, and membrane lipids, which contributes to the development of many diseases and accelerates the ageing process [18]. Tocai et al. (2022) assessed the content of polyphenolic compounds in flowers, leaves, roots, and stalks. They determined the content of hydrolysable tannins, sanguiins, sanguisorbic acids, phenolic acids, anthocyanins, flavonols, catechins, and proanthocyanins. The total content of the tested compounds decreased in the following order: flowers > leaves > roots > stalks (14,445, 9963, 8687, and 4606 mg/100 g d.w., respectively) [12]. Gawron-Gzella et al. (2016) also confirmed the presence of polyphenolic compounds in S. officinalis [14]. Numerous studies have confirmed the antioxidant activity of extracts from the aboveground and underground parts of S. officinalis. Muzykiewicz-Szymańska et al. (2024) confirmed the antioxidant potential and polyphenol content in ethanolic extracts of Sanguisorbae herba prepared in 70% and 40% ethanol. Extracts in both ethanol concentrations showed similar antioxidant potential and total polyphenol content [4]. Tocai et al. (2021) indicated that ethanolic extracts from great burnet roots and leaves are a much more valuable source of antioxidants than flowers and stems [13]. Gawron-Gzella et al. (2016) showed that the great burnet methanolic extract was characterised by a higher antioxidant potential, assessed using the DPPH technique and the content of flavonoids, while the aqueous extract was characterised by a higher content of total polyphenols and phenolic acids [14]. Do et al. (2005) confirmed the antioxidant activity of ethanolic extracts of Sanguisorbae radix. The radical scavenging activity (RSA%) of the obtained extracts, assessed using the DPPH technique, was 32.92%, while the polyphenol content was 368.25 mg/g [19]. In the cited studies, the authors used various solvents and extraction methods as well as in vitro spectrophotometric techniques to assess the antioxidant potential. However, the extraction process was optimised in none of the presented publications. To our knowledge, there are no reports in the available literature on optimising the ultrasound-assisted extraction process of the great burnet herb in terms of antioxidant activity and total polyphenol content.

The aim of the study was to optimise the extraction process of the S. officinalis herb using the response surface methodology. We used a third-degree (3³) Central Composite Design (CCD). The raw material content, alcohol concentration, and extraction time were the independent variables used. Their influence on the antioxidant activity, assessed by the DPPH and FRAP methods, and the total polyphenol content (TPC), estimated by the Folin–Ciocalteu technique, was determined. We found no publications in the available literature discussing the optimisation of the S. officinalis herb extraction process in terms of the studied properties. The analysis of the obtained results revealed that the optimal extraction parameters for the S. officinalis herb in the ultrasound-assisted process, in terms of the antioxidant activity and TPC, are a raw material content of 7.5 g/100 mL of solvent, an ethanol concentration of 47% v/v, and an extraction time of 10 min. Lipophilicity tests showed that the optimised extract is highly hydrophilic (log P = −0.307 ± 0.001).

The obtained extract, characterised by high antioxidant activity and polyphenol content, can be used in various industrial sectors, such as the pharmaceutical and cosmetic industries. Due to a number of other biological activities of the S. officinalis herb (such as the antihemorrhagic, antibacterial, or wound-healing properties [4]), the use of this extract in preparations intended for topical application to the skin seems particularly interesting.

2. Materials and Methods

2.1. Reagents

2,2-diphenyl-1-picrylhydrazyl (DPPH), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), Folin–Ciocalteu reagent, gallic acid, and 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) were obtained from Merck (Darmstadt, Germany). Hydrochloric acid, n-octanol, ethanol, methanol, 99.5% acetic acid, iron(III) chloride, iron(II) sulfate, sodium acetate anhydrous, and sodium carbonate were obtained from Chempur, Piekary Śląskie (Poland). All reagents were analytical grade.

2.2. Extracts Preparation

Before the ultrasound-assisted extraction process (UAE), the dried and cut S. offcinalis herb (EkoHerba, Hajnówka, Poland) was ground in a laboratory grinder and sieved through laboratory sieves (MULTISERW-Morek, Brzeźnica, Poland) with a diameter of 0.25 to 0.63 mm. The raw material fraction collected in the receiver (<0.25 mm) was used for extraction. The appropriate amount of plant material was weighed on an analytical balance (RADWAG AS220.R2 PLUS, Radom, Poland) into plastic Falcon test tubes with a capacity of 50 mL, to obtain the concentration of plant material in the extracts equal to 7.5 g/100 mL of solvent, 5.75 g/100 mL of solvent, 4.0 g/100 mL of solvent, and 2.25 g/100 mL of solvent. Aqueous solutions of ethanol with concentrations of 20, 40, and 60% v/v were used as solvents. The samples were shaken intensively and then extracted in an ultrasonic bath (40 kHz) with a thermostat (FSF-031S, Chemland, Stargard, Poland) at a controlled temperature of 40 ± 1 °C. From each sample, 0.5 mL of extract was taken at the time intervals of 1, 3, 5, 10, 15, and 20 min and stored at +6 °C until analysis.

2.3. Experimental Design

A 3rd degree (3³) Central Composite Design (CCD) was used to mathematically optimise the extraction process of S. officinalis and the response surface methodology (RSM). The main functions describing the extraction process were as follows: (i) antioxidant activity (AA-DPPH) assessed by the DPPH method; (ii) antioxidant activity (AA-FRAP) assessed by the FRAP method; and (iii) total polyphenol content (TPC-FC) assessed by the Folin–Ciocalteu method. The experimental plan and contour drawings for the three studied functions were made using the computer program Statistica 13.3 PL software 7 (StatSoft, Krakow, Poland), followed by an analysis of the plotted drawings. The coefficients of the regression equations (a₁–a₉) for the standardised input quantities, the (1) raw material content, (2) ethanol concentration, and (3) extraction time, were determined by the least squares method using matrix calculus (

Table 1. Factors of the designed experiment.

No. Experiment	Raw Material Content		Ethanol Concentration		Extraction Time
	[g/100 mL of Solvent]		[% v/v]		[min]
	XR 1	XK 2	XR 1	XK 2	XR 1	XK 2
1	7.5	1	20	−1	1	−1
2	7.5	1	20	−1	5.5	0
3	7.5	1	20	−1	10	1
4	7.5	1	40	0	1	−1
5	7.5	1	40	0	5.5	0
6	7.5	1	40	0	10	1
7	7.5	1	60	1	1	−1
8	7.5	1	60	1	5.5	0
9	7.5	1	60	1	10	1
10	5.75	0	20	−1	1	−1
11	5.75	0	20	−1	5.5	0
12	5.75	0	20	−1	10	1
13	5.75	0	40	0	1	−1
14	5.75	0	40	0	5.5	0
15	5.75	0	40	0	10	1
16	5.75	0	60	1	1	−1
17	5.75	0	60	1	5.5	0
18	5.75	0	60	1	10	1
19	4	−1	20	−1	1	−1
20	4	−1	20	−1	5.5	0
21	4	−1	20	−1	10	1
22	4	−1	40	0	1	−1
23	4	−1	40	0	5.5	0
24	4	−1	40	0	10	1
25	4	−1	60	1	1	−1
26	4	−1	60	1	5.5	0
27	4	−1	60	1	10	1

¹ real value, ² coded value.

and

Table 2. The most relevant response surface methodology (RSM) optimisation statistics.

	AA-DPPH	AA-FRAP	TPC-F-C
a0	−4.06156	17.03347	0.271311
a1	2.05059	−7.53573	−0.149816
a2	−0.12100	0.75467	0.026748
a3	0.22767	0.53006	0.044819
a4	−0.00326	−0.00631	−0.000440
a5	0.32177	0.40814	0.065938
a6	−0.05023	−0.06310	−0.009276
a7	0.00028	−0.01391	−0.001604
a8	0.04664	0.09204	0.014078
a9	0.00665	0.01223	0.000189
R2	0.872	0.709	0.860
AdjR2	0.804	0.555	0.786
SD	0.75	2.4	0.14
PRESS	15.4	150	0.529

).

Table 1 shows the experiment planning matrix, with coded values (X_K) calculated using the following Formula (1):

$$X_{Ki}={\displaystyle \frac{2·X_i-X_{max}-X_{min}}{X_{max}-X_{min}}}$$

(1)

where

X_i—the raw material content, ethanol concentration and extraction time, respectively;

X_max, X_min—the minimum and maximum values in the optimisation area.

In addition, an approximating function was determined for the three tested functions describing the extraction process, after which their adequacy was checked using an ANOVA test. The correlation coefficients (R²), the adjusted R-square (AdjR²), the standard deviation (SD), and the PRESS value were also calculated.

The effect of the normalised independent factors (X_i, X_j) of the extraction process of S. officinalis on the value of the response function (Y_i) is presented by means of a second-degree algebraic polynomial (2):

$$Y_i=a_0+{\displaystyle \sum }_{i=1}^3{a}_{i1}·X_i+{\displaystyle \sum }_{i=1}^3{a}_{i2}·X_i^2+{\displaystyle \sum }_{i=1}^2{\displaystyle \sum }_{j=i+1}^3{a}_{ij}·X_i·X_j$$

(2)

where

X_i—the raw material content, ethanol concentration, and extraction time, respectively;

a_i—regression coefficients;

Y_i—the relevant dependent variable, i.e., the antioxidant activity (AA-DPPH and AA-FRAP) and total polyphenol content (TPC-FC).

The waveforms of each function were plotted with simultaneous changes in two process parameters, while the other parameters were constant and took on values consistent with the specified process optimum (Figure 1A–C and Figure 2A–C). In order to obtain the most adequate mathematical description of the extraction process, three technological parameters were taken into account, whose ranges of variation were defined as follows: X₁ (the raw material content): 2.25–7.5 g/100 mL of solvent; X₂ (the ethanol concentration): 20–60% v/v; X₃ (the extraction time): 1–15 min. Actual input values were used for the calculations (

Table 3. Values obtained and calculated for the corresponding experiment number.

No. Experiment	Raw Material Content			Ethanol Concentration			Extraction Time
	Observed Values	Expected Values	Residual Values	Observed Values	Expected Values	Residual Values	Observed Values	Expected Values	Residual Values
	mmol Trolox/L			mmol FeSO4/L			g GA/L
1	8.99	8.56	0.44	10.57	10.24	0.33	1.34	1.30	0.04
2	11.53	10.71	0.82	16.59	14.44	2.16	1.91	1.82	0.09
3	11.39	10.83	0.57	16.72	16.08	0.64	1.93	1.96	−0.02
4	8.09	9.38	−1.28	8.97	11.43	−2.46	1.25	1.43	−0.18
5	11.92	12.13	−0.20	17.53	16.73	0.81	2.01	1.96	0.04
6	12.62	12.84	−0.22	19.24	19.47	−0.23	2.07	2.12	−0.05
7	6.54	7.59	−1.04	4.38	7.58	−3.20	0.95	1.21	−0.26
8	11.84	10.94	0.90	16.67	13.98	2.69	2.02	1.76	0.26
9	12.28	12.25	0.03	17.09	17.82	−0.73	2.02	1.94	0.08
10	7.96	7.68	0.27	8.00	6.25	1.75	1.14	0.97	0.17
11	8.16	9.47	−1.31	8.13	9.73	−1.60	1.33	1.38	−0.05
12	8.09	9.22	−1.12	8.10	10.64	−2.54	1.30	1.41	−0.10
13	9.23	8.49	0.74	9.39	7.93	1.45	1.25	1.16	0.10
14	10.83	10.87	−0.04	6.03	12.50	−6.47	1.50	1.58	−0.08
15	10.97	11.22	−0.25	15.60	14.52	1.07	1.52	1.63	−0.11
16	8.42	6.69	1.72	9.33	4.57	4.77	1.21	0.99	0.22
17	10.08	9.67	0.41	12.40	10.24	2.16	1.40	1.43	−0.03
18	10.20	10.62	−0.42	12.77	13.36	−0.59	1.40	1.50	−0.10
19	6.19	6.07	0.12	6.40	6.89	−0.49	0.74	0.81	−0.07
20	7.17	7.48	−0.31	9.31	9.64	−0.33	0.97	1.10	−0.13
21	7.39	6.87	0.53	9.91	9.83	0.08	1.10	1.02	0.08
22	6.34	6.87	−0.53	8.00	9.06	−1.06	0.94	1.05	−0.11
23	9.56	8.88	0.67	15.91	12.90	3.01	1.42	1.37	0.06
24	9.98	8.86	1.12	18.08	14.20	3.89	1.65	1.30	0.34
25	4.62	5.06	−0.44	5.09	6.18	−1.09	1.04	0.94	0.10
26	6.73	7.67	−0.94	8.71	11.13	−2.42	1.12	1.27	−0.15
27	8.02	8.25	−0.23	11.93	13.52	−1.59	1.12	1.23	−0.11

).

PRESS values were determined from the following Equation (3):

$$PRESS={\displaystyle \sum }_{i=1}^n{\left(Y_i-Y_{ie}\right)}^2$$

(3)

where

Y_i—predicted value;

Y_ie—experimental value.

Table 2 shows the most relevant statistical parameters for optimising the extraction process using the response surface methodology (RSM): the determined coefficients of Equation (2), a₀, a₁, a₂, a₃, a₄, a₅, a₆, a₇, a₈, and a₉, for the function ranges tested, the correlation coefficients (R²), the adjusted R-square (adjR²), the standard deviation (SD), and the PRESS value.

The reliability of the response surface methodology (RSM) for the optimised plant extract was confirmed experimentally. For this purpose, two independent test samples (extracts) were prepared, and each of them was tested in triplicate. The results were expressed as an arithmetic mean of six measurements (n = 6).

2.4. Antioxidant Activity and Total Polyphenol Content Evaluation

The antioxidant activity of the extracts was assessed using the DPPH and FRAP methods, and the total polyphenol content was assessed using the Folin–Ciocalteu technique. The analyses were performed on a Hitachi U-5100 UV-Vis spectrophotometer (Tokyo, Japan) according to the methodology described by Jakubczyk et al. [20]. In the DPPH technique, 150 μL of the tested extract was mixed with 2850 μL of a DPPH radical (0.3 mM) solution prepared in concentrated ethanol. The DPPH working solution was diluted with ethanol until the absorbance at 517 nm was 1.00 ± 0.02. Concentrated ethanol was used as a blank zero. The incubation time of the samples in the dark was 10 min. Results based on the standard curve (y = −0.9795x + 0.9342, R² = 0.977, standard concentration 0–0.9 mmol/L) are expressed as Trolox equivalents [mmol Trolox/L]. The working solution for the FRAP method was made up of 10 volumes of acetate buffer (pH 3.6), 1 volume of 10 mM TPTZ (dissolved in 40 mM HCl), and 1 volume of 20 mM FeCl₃ (aqueous solution). A total of 2320 µL of this solution was mixed with 80 µL of the tested extract, and then the absorbance at 593 nm was measured after 15 min of incubation. Results based on the standard curve (y = 0.6362x + 0.0929, R² = 0.998, standard concentration 0–2.25 mmol/L) are expressed as FeSO₄ equivalents [mmol FeSO₄/L]. The Folin–Ciocalteu technique was used to evaluate the total polyphenol content. A total of 150 μL of the tested extract, 150 μL of Folin–Ciocalteu reagent (tenfold diluted with water), 1350 μL of water, and 1350 μL of 0.01 M sodium carbonate solution were mixed. After 15 min of incubation at room temperature, the sample absorbance was measured at 765 nm. Results based on the standard curve (y = 0.9747x − 0.0032, R² = 0.998, standard concentration 0–0.6 g/L) are expressed as gallic acid equivalents [g GA/L]. All results are the arithmetic mean of three independent samples ± SD (standard deviation).

2.5. Partition Coefficient

To determine the lipophilicity of the optimised extract of S. officinalis (involving the determination of its partition coefficient between two immiscible liquids), n-octanol and water were used, which model the properties of cell structures well [21,22]. In order to determine the lipophilicity of the extract, the n-octanol/water partition coefficient (P) values were investigated. The partition coefficient was expressed as the ratio of the logarithms of the substance concentrations in the two phases.

The logarithm of the partition coefficient (log P) n-octanol/water was determined by a spectrophotometric method by carrying out analyses with the Thermo Scientific GENESYS 50 apparatus (Thermo Fisher Scientific, Norristown, PA, USA). The log P was determined for the sum of active compounds having absorbances in the wavelength range of 190–450 nm. Based on the mass balance and the assumption that Lambert–Beer’s law is satisfied in the studied wavelength range (190–450 nm), the concentration ratio of the compounds in the extract was determined, as illustrated by Equation (4):

$$P={\displaystyle \frac{C_o}{C_w}}={\displaystyle \frac{C^0-C_w}{C_w}}={\displaystyle \frac{S^0-S}{S}}={\displaystyle \frac{{{\displaystyle \int }}_{{\mathsf{\Lambda }}_1}^{{\mathsf{\Lambda }}_2}{A}^{0}d\mathsf{\Lambda }-{{\displaystyle \int }}_{{\mathsf{\Lambda }}_1}^{{\mathsf{\Lambda }}_2}Ad\mathsf{\Lambda }}{{{\displaystyle \int }}_{{\mathsf{\Lambda }}_1}^{{\mathsf{\Lambda }}_2}Ad\mathsf{\Lambda }}}$$

(4)

where

C—the concentration of the sum of the compounds in the n-octanol layer and in the aqueous layer;

S—the area occupied by the compound under the UV-Vis spectrum;

A—the absorbance;

0—the superscript in the extract before extraction;

Λ—the wavelength.

A total of 10 mL of n-octanol saturated with water was mixed with 10 mL of deionised water saturated with 1-octanol (1:1 ratio) containing an optimised extract of S. officinalis in the amount of 100 µL. The mixture was then shaken on a SK-L330-PRO linear shaker (ChemLand, Stargard, Poland) for an appropriate time (i.e., until equilibrium was reached) at a constant temperature of 25 °C, which was controlled by a circulation thermostat, and then it was centrifuged at the same temperature for 10 min at 7500 rpm for better separation of the layers and analysed by UV–Vis spectroscopy. The sum of the concentrations of the active substances present in the analysed extract was determined by spectrophotometry in the wavelength range of 190–450 nm. A blank was carried out under the same conditions using n-octanol and water in a 1:1 ratio, and the procedure was as above [21,22].

2.6. Statistical Analysis

During the optimisation of the extraction process of S. offcinalis, a 3-degree (3³) Central Composite Design (CCD) was used. The experimental plan and contour drawings were made using the computer program Statistica 13.3 PL software 7 (StatSoft, Kraków, Poland). A one-way analysis of variance (ANOVA) was used to statistically analyse the optimisation of the extraction process. The adequacy of the function tested was checked using an ANOVA test. Results are displayed as the mean ± standard deviation (SD). Statistical computations were performed using Statistica 13.3 PL software 7 (StatSoft, Kraków, Poland).

3. Results

3.1. Response Surface Methodology for Extraction Process of S. offcinalis

Table 2 shows the extraction process’s most relevant response surface methodology (RSM) optimisation statistical parameters.

On the basis of the statistical coefficients that were determined (Table 2), it can be concluded that, for each output parameter (i.e., response functions AA-DPPH, AA-FRAP, and TPC-F-C), the response surface described by the second-order equation correlates very well with the experimental data, which made it possible to determine the optimal areas of the S. offcinalis extraction process (Table 2).

In Table 3, the observed and expected (calculated) values and the corresponding residual values for the respective experiment number are presented.

Figures S1–S3 present scatter plots of the observed and expected values (Table 3) during the interaction of the studied extraction process parameters for S. offcinalis. The scatter plots indicate a good correlation between the response surface methodology and the experimental values (Figures S1–S3).

Table 4. The input variable values for which the functions described take extreme values.

Input Variables	Process Functions Examined
	AA-DPPH	AA-FRAP	TPC-F-C
Raw material content [g/100 mL of solvent]	11	4.7	2.6
Ethanol concentration [% v/v]	47	48	47
Extraction time [min]	11	11	6

shows the values of the input variables for which the tested functions describing the extraction process (AA-DPPH, AA-FRAP, and TPC-F-C) take extreme values.

Figure 1A–C shows the changes in the antioxidant activity (AA-DPPH and AA-FRAP) and total polyphenol content (TPC-F-C) during the interaction of the studied process parameters. Studies on the influence of ethanol concentration and plant material content on the extraction of S. offcinalis were carried out at an extraction time of 10 min, considered to be the most favourable, and with the other parameters used as starting points at the beginning of the study.

Figure 1A, which illustrates the relationship between the plant material content and ethanol concentration, demonstrates that the highest plant material content (≥7.5 g/100 mL of solvent) and ethanol concentration (47% v/v) yield the highest values of antioxidant activity (AA-DPPH) by the DPPH method (>12 mmol Trolox/L). A reduction in the plant material content (below 7.5 g/100 mL of solvent) at an ethanol concentration of 47% v/v results in a decrease in the value of the function tested: below 12 mmol Trolox/L (Figure 1A).

Figure 1B, showing the relationship between the plant material content and ethanol concentration and antioxidant activity (AA-FRAP) using the FRAP method, shows that during a 10 min extraction run using 4.7 g/100 mL of solvent of plant material and an ethanol concentration of 48% v/v, the extremum of the function under study is reached as a minimum. In the area of input conditions, the FRAP antioxidant activity function describing the extraction process increases monotonically with respect to the raw material content (from <8 to >20 mmol FeSO₄/L) for a constant extraction time of 10 min. During a 10 min extraction run, the highest value of the function under study (i.e., >20 mmol FeSO₄/L) is achieved using the maximum amount of plant material (i.e., ≥7.5 g/100 mL of solvent), while the use of lower amounts of plant material (i.e., less than 7.5 g/100 mL of solvent) results in a decrease in AA-FRAP (Figure 1B).

Figure 1C illustrates the relationship between the plant material content and ethanol concentration, showing that the TPC-F-C function reaches a local minimum during a 10 min extraction with 2.6 g/100 mL of solvent of plant material and a 47% v/v ethanol concentration. In the area of input conditions, the total polyphenol content (TPC-FC) function describing the extraction process increases monotonically with respect to the raw material content (from <0.9 to >2.2 g GA/L) for a constant extraction time of 10 min. During a 10 min extraction, the highest total phenolic compound content in the extract (i.e., >2.2 g GA/L) is obtained using the highest content of plant material (i.e., ≥7.5 g/100 mL of solvent) —Figure 1C.

For the tested functions describing the extraction process, the antioxidant activity (AA-FRAP) and total polyphenolic content (TPC-FC), the fitted surface has a local minimum for the plant’s raw material content of 4.7 and 2.6 g/100 mL of solvent, respectively. In the input conditions area, the function describing the extraction process is monotonically increasing with respect to the raw material content, as shown in Figure 1B,C for a fixed extraction time of 10 min. The fixed extraction time is close to the time for which the function describing the process takes extreme values (Table 4). The choice of the extraction running time was dictated by the maximum recovery of active substances with the antioxidant potential and economic considerations of running the process.

The optimum extraction parameters were identified, i.e., those at which the maximum values of the main functions describing the process were obtained (

Table 5. Optimal parameters for the extraction process of S. offcinalis in the presence of ethanol as a solvent and the corresponding values of the main functions of the process.

Optimal Parameters of the Extraction Process	Unit
Raw material content	g/100 mL of solvent	7.5
Ethanol concentration	% v/v	47
Extraction time	min	10
Functions of the extraction process
AA-DPPH	mmol Trolox/L	12.9
AA-FRAP	mmol FeSO4/L	19.4
TPC-F-C	g GA/L	2.1

).

The changes in antioxidant activity (AA-DPPH and AA-FRAP) and total polyphenol content (TPC-F-C) are shown in Figure 2A–C. Studies on the effects of the ethanol concentration and extraction time on the extraction process of S. offcinalis were conducted at a plant material content of 7.5 g/100 mL of solvent (considered to be the most favourable and corresponding to the maximum concentration resulting from the technical conditions) and with the other parameters adopted as starting parameters at the beginning of the study.

Figure 2A, showing the relationship between the ethanol concentration and extraction time, shows that the function of AA-DPPH reaches its highest value (above 11 mmol Trolox/L) using an ethanol concentration of 47% v/v and an extraction time in the range of 7–11 min, with a plant material content of 7.5 g/100 mL of solvent. The observed antioxidant activity of the extracts depended on the extraction time: increasing (i.e., above 11 min) or decreasing (i.e., below 7 min) the extraction time results in a decrease in the value of the test function (<11 mmol Trolox/L)—Figure 2A.

Figure 2B, showing the relationship between the ethanol concentration and extraction time, shows that for an ethanol concentration of 48% v/v and an extraction time ranging from about 9 to about 14 min (at a plant material content of 7.5 g/100 mL of solvent), the highest antioxidant activity by the FRAP method is achieved (i.e., above 14 mmol FeSO₄/L). Reducing the extraction time (below 9 min) or increasing the extraction time (above 14 min) results in lower AA-FRAP values (below 14 mmol FeSO₄/L. Extending the extraction time positively affects the antioxidant activity of the extract only during a certain time interval, i.e., until the equilibrium state of concentrations of extracted components inside and outside the plant material is established. After this time, prolonging the extraction (more than 14 min) does not affect the activity of FRAP, causing secondary changes in the composition of the extract—Figure 2B.

Figure 2C, showing the relationship between the ethanol concentration and extraction time (at a constant plant material content of 7.5 g/100 mL of solvent), shows that using an ethanol concentration of 47% v/v and an extraction time ranging from about 8 to about 10 min, the highest contents of phenolic compounds (above 1.6 g GA/L) are achieved. The choice of an ethanol concentration of 47% v/v was dictated by the fact that the aqueous ethanol solvent is more polar and effective in the extraction of polyphenols compared to pure ethanol (Figure 2C).

For all the functions investigated, the local maximum of the function corresponds to values of the alcohol concentration and extraction time consistent with those in Table 4.

Since the extraction process of S. offcinalis is carried out simultaneously towards three input values, such as the raw material content, ethanol concentration, and extraction time, the optimal values for AA-DPPH, AA-FRAP, and TPC-F-C correspond to an averaged input value of 9.3 min (i.e., 10 min) and an ethanol concentration of 47% v/v—Table 5. Given the input variables, it should be noted that the optimisation of each response function under investigation is a balancing act between properly setting the individual extraction parameters in such a way as to obtain the greatest amount of the desired components in the extract, affecting the values of the process-describing functions under investigation.

The results obtained and the fitted plots for the assumptions indicate a clear area of optimal conditions for the extraction process. The plan used and the RSM optimisation method describe the extraction process well. This applies to each output parameter (i.e., response functions such as AA-DPPH, AA-FRAP, and TPC-F-C). The results indicate that the application of the response surface methodology allows optimal areas to be found quickly.

Table 5 shows the optimal parameters determined for the extraction process of S. offcinalis in the presence of ethanol as a solvent and the corresponding values of the main process functions.

The results obtained from the optimisation of the extraction process of S. offcinalis show that conducting a 10 min extraction in the presence of 47% v/v ethanol, using the highest content of plant material (i.e., 7.5 g/100 mL of solvent), yields the highest values of the studied functions: AA-DPPH 12.9 mmol Trolox/L, AA-FRAP 19.4 mmol FeSO₄/L, and TPC-F-C 2.1 g GA/L (Table 5).

The reliability of the response surface methodology for the extract obtained from S. offcinalis was confirmed experimentally. The obtained values for the main functions of the extraction process are shown in

Table 6. Values of the main functions of the extraction process of S. offcinalis based on RSM analysis.

Functions of the Extraction Process	Unit	Value
AA-DPPH	mmol Trolox/L	12.7 ± 0.9
AA-FRAP	mmol FeSO4/L	20.8 ± 0.9
TPC-F-C	g GA/L	2.4 ± 0.1

n = 6, optimal parameters of the extraction process: raw material content, 7.5 g/100 mL of solvent; ethanol concentration, 47% v/v; extraction time, 10 min.

.

The values of the main process functions (AA-DPPH = 12.7 mmol Trolox/L ± 0.9, AA-FRAP = 20.8 mmol FeSO₄/L ± 0.9, and TPC-F-C 2.4 g GA/L ± 0.1) obtained from the validation studies performed (Table 6) coincide with the values calculated using the equation describing the response surface.

3.2. Measurement of Lipophilicity

The results showed that the value of the partition coefficient determined by the shake-flask method for the optimised extract obtained from S. offcinalis was −0.307 ± 0.001 (Figure 3).

Lipophilicity studies have shown that the optimised extract from S. offcinalis is highly hydrophilic, for which the log P partition coefficient is −0.307 ± 0.001, which is due, among other things, to the presence of phenolic acids (i.e., gallic acid, 3,4-dihydroxybenzoic acid, 2,5-dihydroxybenzoic acid, vanillic acid, 2,3-dihydroxybenzoic acid, caffeic acid, and chlorogenic acid) in the extract [4]—Figure 3.

4. Discussion

The optimal extraction parameters of the tested plant material in terms of the antioxidant activity and TPC were 7.5 g raw material/100 mL of solvent, 47% v/v ethanol as a solvent, and a processing time of 10 min. The presented studies used indirect ultrasound-assisted extraction performed using an ultrasonic bath with a frequency of 40 kHz. González-Centeno et al. (2014) optimised the aqueous ultrasonic extraction of grape pomace using the RSM in terms of antioxidant capacity as well as total phenolic and flavonol content. The acoustic frequency (40, 80, 120 kHz), ultrasonic power density (50, 100, 150 W/L), and extraction time (5, 15, 25 min) were used as independent variables. They set the optimal process conditions at 40 kHz, 150 W/L, and 25 min, respectively [23]. Vetal et al. (2013) optimised the parameters of UAE for ursolic acid from Ocimum sanctum. Similar to our studies, they analysed the extraction time and solid-to-solvent ratio as well as other parameters such as solvent selection, process temperature, and ultrasound power and frequency. They obtained the maximum extraction efficiency of ursolic acid at 12 min, a solid/solvent ratio of 1:30, 45 °C, and 25 kHz. However, by analysing the effect of the ultrasound frequency (25 and 40 kHz), they showed higher extraction efficiency when using the frequency of 40 than 25 kHz [24]. Lower ultrasound frequencies in the range of 18–40 kHz seem to be effective in extracting biologically active compounds from plant materials, including those with antioxidant potential, by disrupting cell walls, resulting in increased solvent penetration and mass transfer rates [23]. Therefore, in our own research, an ultrasound frequency of 40 kHz was used in the process of extraction of biologically active compounds from the herb S. officinalis, and in order to increase the extraction efficiency, powdered plant raw material was used. In our research, we used powdered S. officinalis herb (<0.25 mm). Alsaud and Farid (2020) showed that, as the particle size of the plant material decreases, the extraction efficiency increases, resulting in a higher content of biologically active compounds extracted from the plant material. Furthermore, they observed that, in the case of the extraction of powdered material, most of the extraction was obtained directly after the contact of the raw material with the solvent. They suggest that, when grinding into a fine powder, bioactive compounds are released; therefore, a short-term contact of the raw material with a solvent is sufficient to recover most of the active substances [25]. Becker et al. (2016) analysed the effect of grinding and sieving of plant raw materials, Hypericum perforatum and Achillea millefolium, on antioxidant activity and total polyphenol content. In their studies, they used raw materials with particle sizes of 20 to 500 μm. They observed the highest antioxidant activity for the 100–180 μm fraction, while the 180–315 μm fraction contained the most polyphenols. Based on the obtained results, they concluded that the optimal raw material fractions in terms of the tested properties were those with a diameter of 100–315 μm [26]. These researchers confirmed their observations in subsequent studies on Hieracium pilosella. The raw material fractions 180–315 μm and 315–500 μm exhibited the highest antioxidant activity and content of the selected phenolic compounds [27]. Deli et al. (2019) showed the highest antioxidant activity and total polyphenol content in granulometric classes of 0–180 μm for Boscia senegalensis, 180–212 μm for Dichrostachys glomerata, and 212–315 μm for Hibiscus sabdariffa. The authors postulate that the technology of grinding and controlled sieving is an effective tool for increasing the antioxidant activity and polyphenol content in plant products [28]. The validity of using ground S. officinalis herb in our own research is confirmed in the previously published study by Muzykiewicz-Szymańska et al. (2024). This study presents the results of the extraction of unground dried great burnet herb, 0.5–1 cm in size. Analogously to the presented study, the antioxidant activity and total polyphenol content of ethanolic extracts prepared in 40%, 70%, and 96% v/v aqueous alcohol solutions were examined. Depending on the solvent concentration used, they obtained antioxidant activity, assessed using the DPPH technique, ranging from 127 to 186 mg Trolox/L. The TPC was in the range of 128–203 mg GA/L [4]. Comparing the results presented by us with the observations made in the cited study, it can be seen that using ground raw material for extraction led to more than a 15-fold increase in antioxidant activity and more than a 10-fold increase in the content of polyphenols in the extracts, even though the processing time was reduced from 60 min to 10 min. These results clearly indicate the beneficial effect of grinding and sieving the plant material used to extract compounds with antioxidant potential from various plants, including the Sanguisorbae herba. The extraction process was carried out at a constant, controlled temperature of 40 ± 1 °C. Many authors indicate that an increase in the process temperature may have a beneficial effect on the extraction efficiency of biologically active compounds from plant material due to an increase in the solubility of dissolved substances as well as the diffusion coefficient. An increase in temperature may also reduce the solvent’s viscosity and surface tension, soften plant tissues, and improve the release of antioxidants into the solvent. However, higher temperatures may cause the denaturation of phenolic compounds, a change in their stability due to chemical or enzymatic degradation, and losses due to thermal decomposition [29,30,31]. The optimum temperature of 40 °C for the extraction of polyphenols is reported in the literature [32].

The optimal concentration of S. offcinalis herb in the extract was 7.5 g/100 mL of solvent (the solvent-to-material ratio was 13.3 mL/g). In the case of the antioxidant activity assessed by the DPPH method, it was noted that reducing the concentration of the raw material below 7.5 g/ 100 mL of solvent using the optimal ethanol concentration (47% v/v) caused a decrease in the value of the tested function below 12.9 mmol Trolox/L. The results from the FRAP method showed a similar relationship. Using the optimal extraction time (10 min), the highest activity (i.e., 19.4 mmol FeSO₄/L) was achieved with the highest concentration of the raw material in the sample, while its reduction resulted in a decrease in the tested potential. We made similar observations regarding the total polyphenol content. We obtained the highest concentration of polyphenols in the extract (2.1 g GA/L) during 10 min UAE, using a raw material concentration of 7.5 g/100 mL of solvent. The observations of higher polyphenol content and antioxidant activity at the highest concentration of plant material in the sample are consistent with mass transfer principles. The higher the solvent-to-material ratio, the higher the total amount of solids obtained, regardless of the solvent used [30]. Adil et al. (2007) obtained similar solid-solvent ratio values as in our study. The authors optimised the sour cherry pomace (solid)/ethanol (solvent) ratio within the range of 0.05, 0.15, and 0.25 g/mL. The optimal ratio in terms of antioxidant activity was determined to be 0.06–0.07 g/mL (6–7 g of plant raw material/100 mL of solvent) [33]. Topuz et al. (2016) optimised the solvent: the seaweed (Laurencia obtuse) ratio was in the range of 10:1–30:1 for the antioxidant activity and TPC. In their study, the optimal ratio for the tested parameter was 30:1 [34].

Our optimisation showed that, regardless of the response function, the optimal ethanol concentration is 47% v/v. Ethanol is an effective solvent for the extraction of compounds with antioxidant potential, such as polyphenols and flavonoids, as well as tannins, terpenoids, and alkaloids. Water is used, among other substances, for the extraction of tannins, anthocyanins, terpenoids, and saponins [35]. Ghasemi et al. (2024) also optimised the ethanol concentration (in the range of 25–75% v/v) used for the extraction of Rheum ribes flowers in terms of phytochemical (the total polyphenol and flavonoid content) and antioxidant activity (the FRAP method). In their study, the optimal ethanol concentration was 53.14% [36]. Hammi et al. (2015), in their study on the optimization of UEA of Zizyphus lotus fruits for antioxidants, found an optimal ethanol concentration of 50%. They optimised the ethanol concentration in the range of 0 to 100% [37]. A significant effect of the ethanol concentration on the antioxidant potential, including the total phenolic content, is postulated by Waszkowiak and Gliszczyńska-Świgło (2016). In their studies, extraction with a binary solvent (alcohol and water) was more efficient in obtaining phenolic compounds compared to the process using pure ethanol. This is probably related to the polarity of both solvents [38]. Ethanol and water mixtures are more polar than absolute ethanol, which may positively influence the extraction efficiency of phenolic compounds from plant material [39]. Spigno et al. (2007) made interesting observations about the content of phenolic compounds in grape marc extracts, depending on the ethanol concentration. Researchers observed an increase in the yield of phenols at a water content in ethanol from 10% to 30% and their constant levels at a water content of 30% to 60%. When the water content exceeded 50%, the phenol concentration decreased [30]. Byun et al. (2021) compared the antioxidant activity (TAC, FRAP, DPPH, ABTS, and SRSA) as well as the polyphenol and flavonoid content of S. officinalis root extracts prepared in water and 20, 40, 80, and 100% ethanol. The lowest content of polyphenols (TPC) and flavonoids (TFC) was noted in aqueous extracts, while the highest was in 40% (TPC) and 80% (TPC and TFC) ethanol. The concentration of these compounds was almost twice as high in extracts prepared in the above-mentioned alcohol concentrations, compared to aqueous extracts. A similar relationship between the lowest antioxidant activity in water extracts and in ethanol extracts was observed in the case of the TAC, FRAP, and SRAS methods, in which the highest results were observed after using 80% (TAC, SRAS) and 40% (FRAP) alcohol. The highest activity assessed by the DPPH technique was noted for the extract in pure ethanol and the lowest for the sample in 40% solvent. The opposite relationship was observed in the ABTS method. These observations led the authors to choose 80% ethanol as the most efficient solvent for extracting compounds with antioxidant potential. However, it is crucial to note that these studies did not optimise the ethanol concentration as a solvent [40]. In the aforementioned work by Muzykiewicz-Szymańska et al. (2024), compounds with antioxidant potential were extracted more efficiently after using ethanol at concentrations of 40% and 70% v/v compared to concentrated ethanol [4].

The extraction time was the third independent variable in the optimization process. As already mentioned, one of the main advantages of UAE is the reduction in the processing time while maintaining or even increasing the efficiency compared to conventional techniques [41]. In the case of the DPPH method, it was noticed that extending the extraction time above 11 min and shortening it below 7 min resulted in a decrease in the antioxidant activity of the extract below 11 mmol Trolox/L. We also observed a decrease in activity (assessed using the FRAP method) below 14 mmol FeSO₄/L when we shortened the processing time to under 9 min and extended it to above 14 min. The highest polyphenol content (>1.6 g GA/L) was observed during 8–10 min of UAE. Therefore, the optimal extraction time of S. officinalis herb in terms of the antioxidant activity and total polyphenol content was 10 min. As noted, extending the extraction time above certain values may have an adverse effect on plant extracts’ antioxidant potential or polyphenol content. This could be due, among other things, to the reaction of phenols with other plant components, which makes their extraction difficult. In such a situation, extending the process duration will result in greater losses of polyphenols [29]. Yim et al. (2012) noted a linear decrease in the FRAP activity during the extraction of edible wild mushrooms (Pleurotus porrigens) from 240 to 420 min [29], while Thoo et al. (2010) noted a decrease in polyphenol content as a result of prolonging the extraction time of Morinda citrifolia. In their study, an extraction time of 20–120 min significantly affected the antioxidant activity (using the DPPH and ABTS methods) and the TPC but not the total flavonoid content (TFC). The authors explain the difference in the optimal extraction times for the TPC and TFC by the different degrees of phenol polymerisation, their solubility, and their interactions with other components, which leads to the difference in time needed to reach an equilibrium between the solution in the solid matrix and in the bulk solution [42]. Therefore, optimising the extraction time of a specific plant raw material in terms of appropriate activity or the content of a specific group of biologically active compounds is a key element in planning the extraction process. Many studies have focused on optimising the extraction time using the response surface methodology [43,44,45,46,47,48]. It should be remembered that an additional parameter accelerating the process of the extraction of compounds with antioxidant potential was the grinding of the raw material and the use of the powdered fraction. Szydłowska-Czerniak and Tułodziecka (2015) optimised the time (in the range of 6–18 min) of the UAE of antioxidants from two rapeseed varieties. Before the extraction process, the tested plant material was ground, and a 0.5 mm fraction was used for the process. The optimal extraction time for the tested plant material varieties was 13.8 min [49]. Xu et al. (2016) optimised the time (range 0–30 min) of UAE on the antioxidant activity of Jatropha integerrima flowers using the RSM. For extraction, they used a ground fraction of the raw material smaller than 0.3 mm. The optimal process time was 7 min [50]. Imtiaz et al. (2023) optimised UAE for the TPC, TFC, and antioxidant activity of Thuja orientalis leaf powder. The optimal UAE time was 2 min. In comparison, the optimal microwave-assisted extraction (MAE) time was even shorter, at just 90 s [51]. Wang et al. (2013) optimised the extraction of Sanguisorbae radix for flavonoid content. Test samples were prepared by heating, refluxing, and extraction. The optimal time for the extraction technique they used was 90 min [46]. Shehata et al. (2021) found that the optimal UAE time (range 10–80 min) for ground freeze-dried Citrus sinensis peels for antioxidant activity, TPC, and TFC was 44 min [52]. The cited results confirm the importance of optimising the plant material’s extraction time in terms of the expected activity or content of a specific group of biologically active compounds as well as the importance of preparing the plant material for the extraction process. Therefore, the use of mathematical modelling, including the response surface methodology, allows for the optimisation of process parameters with a significantly smaller number of tests, which directly translates into a lower energy and solvent consumption.

We tested the optimised S. officinalis herb extract for lipophilicity by calculating its n-octanol/water partition coefficient. The log P partition coefficient was valued at −0.307 ± 0.001, which indicates the strongly hydrophilic character of the extract. This may be related to the content of phenolic acids in the tested sample, which was identified in previous studies by Muzykiewicz-Szymańska et al. (2024). In this study, the presence of several phenolic acids (i.e., gallic acid, 3,4-dihydroxybenzoic acid, 2,5-dihydroxybenzoic acid, vanillic acid, 2,3-dihydroxybenzoic acid, caffeic acid, and chlorogenic acid) was identified in the ethanol extract of Sanguisorbae herba, with a total concentration of approximately 1.3 g/L. Particularly high concentrations (224 ± 10 mg/L) were found for gallic acid [4]. Kucharski et al. (2022) noted the high hydrophilicity of preparations from leaves of plants from the Rubus genus (R. idaeus and R. fruticosus), estimated using log P. The log P values for the preparations were −0.18 and −0.19, respectively. The authors postulate that the high hydrophilicity of the preparations is most likely related to the high content of polyphenols, which in turn affect the antioxidant potential of the tested samples [21].

The conducted analysis provides us with information on the optimal parameters of the ultrasound-assisted extraction of the S. officinalis herb. Despite possessing numerous health-promoting properties, the S. officinalis herb remains an underestimated plant raw material in industrial applications. It should be remembered that the optimisation of process parameters is the first step leading to the industrial use of the extract. The next step should be to standardise the extract in terms of the contents of the key active ingredients. This fact constitutes a limitation of the performed analyses. As already mentioned, the high antioxidant potential and the content of health-promoting polyphenols, combined with other biological activities of Sanguisorbae herba, create the potential to use this herbal raw material in pharmaceutical and cosmetic preparations for topical use. However, for this purpose, it is necessary to conduct in-depth analyses of the effect of the obtained extract on the skin, including an assessment of its cytotoxic potential, in order to confirm the safety of its use. These analyses determine the direction of further research.

5. Conclusions

The optimisation of the extraction process for the antioxidant activity and total polyphenol content of S. officinalis herb extract was successfully carried out using the response surface methodology. The results obtained and the fitted plots for the assumptions made indicate a clear area with optimal conditions for the extraction process studied. The optimised condition was validated and found to be fitted with the experimental values. The analyses showed that extraction parameters, such as the process time, ethanol concentration, or plant material content in the sample, have a significant impact on the tested extract potential. The high antioxidant potential and content of the health-promoting polyphenols in the optimised extract allow it to be used in a variety of industries, including pharmaceuticals and cosmetics. However, further research is needed on the effect of the obtained extract on the skin, including an assessment of the safety of its use in topical preparations.