1. Introduction
Polyphenols constitute a class of plant secondary metabolites characterized by great structural and functional diversity [
1]. They are found in various plant-derived sources, such as fruits (berries, watermelon, apples, grapes, etc.), vegetables (soybeans, onions, etc.), and cereals, as well as in beverages like coffee, red wine, and juices [
1,
2]. Polyphenolic compounds encompass phenols, flavonoids, phenolic acids, proanthocyanidins, tannins, lignans, coumarins, and stilbenes, which occur in different parts of plants, such as leaves, flowers, roots, and shoots [
3,
4]. They have attracted considerable scientific interest for their potent antioxidant activity, primarily because they can reduce oxidative stress and neutralize free radicals [
5], effects which, in turn, contribute to the prevention of chronic diseases such as cardiovascular disorders, neurodegenerative conditions, and certain types of cancer [
1,
6]. The food, pharmaceutical, and cosmetic industries have begun to utilize phenolic compounds more frequently in recent years because of their antibacterial, anti-inflammatory, and antioxidant characteristics [
3,
7]. The development and improvement of technologies for extracting and purifying these compounds from plant materials, foods, and food-industry by-products reflect the increasing interest in these compounds and their potential uses [
3].
Among plants rich in polyphenols is
C. monogyna (CM) Jacq., commonly known as the hawthorn [
8,
9]. Its scientific name derives from the Greek word “kràtaigos,” meaning “strength and robustness,” in reference to its hard, durable wood [
10]. Hawthorn is a deciduous shrub with distinctive white flowers and red berries [
8]. It is endemic to temperate regions of the Northern Hemisphere, including Europe, Asia, and North Africa, and has a long history of use in folk medicine for its cardioprotective and neuroprotective properties [
8,
9]. The fruits, leaves, and flowers of CM are rich in phenolic compounds, including flavonoids such as rutin, quercetin 3-
D-galactoside (hyperoside), and vitexin, as well as other constituents such as vitamin C, saponins, tannins, cardiotonic amines (e.g., phenylethylamine, tyramine), procyanidins, triterpenoid acids (e.g., ursolic acid), and purine derivatives (e.g., adenosine, guanine) [
4,
8,
11,
12,
13]. The bioactive profile of hawthorn has been linked to antioxidant, hypolipidemic, anti-inflammatory, and neuroprotective activities, underscoring its potential for applications in food, pharmaceutical, and cosmetic formulations [
8,
9,
14].
Extraction is a pivotal step in the chemical analysis of plant samples, necessary for sample preparation and the isolation of bioactive compounds from plant tissue [
15]. Various extraction techniques for bioactive compounds are reported in the literature, with solid–liquid extraction among the most common for isolating plant antioxidants [
16]. Today, extraction methods are broadly classified as conventional or non-conventional [
17]. In conventional approaches, bioactive substances are removed from plant material using traditional solvents (with optional heating) [
16]. The sample is first homogenized and then immersed in a single solvent or solvent mixture under continuous agitation, allowing the target compounds to diffuse and transfer into the solvent [
16,
18]. Traditional methods—including percolation, maceration, and Soxhlet extraction—rely on straightforward procedures to isolate specific constituents and produce crude extracts [
15]. These extracts may be used directly or formulated into herbal medicines, dietary supplements, and cosmetic ingredients [
15,
19,
20,
21]. However, conventional techniques suffer from drawbacks such as long extraction times, low yield, loss of nutrients, high solvent and energy consumption, and often require multiple extraction steps [
22,
23]. Moreover, heating can degrade or alter heat-sensitive phytochemicals [
22].
By contrast, modern extraction methods achieve higher yields and greater selectivity [
15]. These methods encompass ultrasound-assisted extraction, microwave-assisted extraction, pressurized liquid extraction, supercritical fluid extraction, pulsed electric field extraction, and enzymatic extraction [
15,
23]. The use of “green” solvents further transforms these approaches into environmentally friendly extraction techniques [
24]. In recent years, numerous laboratory-scale studies employing such green methods have yielded high-value extracts from plant materials [
25].
To overcome the limitations of both conventional and current green methods, advanced non-thermal technologies have been developed, among which PEF extraction stands out as a promising and efficient alternative. PEF applies short electric pulses that permeabilize cell membranes by creating pores [
26]. Its advantages over other techniques include ultrashort processing times (nanoseconds to milliseconds), enhanced extraction efficiency (greater membrane permeability), reduced energy consumption, preservation of cellular structure, and higher quality of the final extracts [
26].
Optimization of extraction parameters is essential to maximize yield while preserving the integrity of bioactive compounds. Response surface methodology (RSM)—a combined statistical and mathematical approach—provides an effective framework for modeling and optimizing extraction processes [
27]. RSM enables simultaneous evaluation of multiple independent variables to identify optimal conditions with fewer experimental runs than traditional methods.
The objective of this study is to identify optimal conditions for PEF-assisted polyphenol extraction from CM leaves utilizing RSM. Key extraction parameters, such as electric field intensity, pulse duration, solvent concentration, and extraction time, will be tuned by RSM. The Folin–Ciocalteu assay will quantify total polyphenol content (TPC), high-performance liquid chromatography coupled with a diode array detector (HPLC-DAD) will identify individual phenolic compounds, and DPPH radical scavenging activity and ferric-reducing antioxidant power (FRAP) assays will evaluate antioxidant activity.
3. Materials and Methods
3.1. Chemicals and Reagents
Ethanol (99.8%), Folin–Ciocalteu’s reagent, and gallic acid (97%) were obtained from Panreac Co. (Barcelona, Spain). Acetonitrile (99.9%) was purchased from Labkem (Barcelona, Spain). Hydrochloric acid (37%), 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ) (≥98%), and all polyphenolic standards for the HPLC determination (at least 97% purity or higher) were obtained from Sigma-Aldrich (Darmstadt, Germany). Formic acid (99.8%), sodium carbonate (anhydrous, 99.5%), and rutin (≥94%) were from Penta (Prague, Czech Republic). Iron (III) chloride hexahydrate (97%) was obtained from Merck (Darmstadt, Germany). A deionizing column that contains mixed-bed ion exchange resin, ensuring conductivity below 1 µS/cm, with a standard flow rate and operating pressure, was used to produce deionized water for all the experiments.
3.2. Plant Material
Dried CM leaves were purchased from a local market in Karditsa, Greece. Then, the CM leaves were sieved using an Analysette 3 PRO (Fritsch GmbH, Oberstein, Germany), and the average resulting particle size was 355 μm. Powder with a particle size below 400 μm was chosen for the experiments and stored in a freezer at −40 °C until further processing.
3.3. Experimental Design
The study employed the RSM with a custom design to optimize extraction conditions for TPC, FRAP, and DPPH antiradical activity. This approach was applied to the PEF extraction process for the dry leaves of CM. Six key independent variables were examined: electric field strength (
E, kV/cm) as
X1, pulse period (
Tpulse, μs) as
X2, pulse duration (
tpulse, μs) as
X3, ethanol concentration in water (
C, %
v/v) as
X4, liquid-to-solid ratio (
R, mL/g) as
X5, and extraction time (
t, min) as
X6, each tested at three levels—low (−1), medium (0), and high (+1)—as shown in
Table 7. To ensure reliability, 30 experimental runs were conducted, including two central points, with each experiment repeated three times and the average values recorded.
To improve the model’s predictive accuracy, stepwise regression was employed to eliminate unnecessary terms, thereby minimizing variance and refining the estimation process. This optimization resulted in a second-order polynomial equation (4) that defines the interactions among the six independent variables:
where the independent variables are denoted by
Xi and
Xj, and the predicted response variable is defined by
Yk. In the model, the intercept and regression coefficients
β0,
βi,
βii, and
βij represent the linear, quadratic, and interaction terms, respectively.
Two stainless steel chambers made by Val-Electronic of Athens, Greece, a mode/arbitrary waveform generator by UPG100 of ELV Elektronik AG of Leer, Germany, a digital oscilloscope by Rigol of Beaverton, Oregon, USA, and a high-voltage power generator by Leybold of LD Didactic GmbH (Hürth, Germany) were used to process the samples utilizing PEF. The stirring extraction (STE) technique was carried out using a stirring hotplate manufactured by Heidolph Instruments GmbH & Co. KG of Schwabach, Germany. The UAE treatment was conducted using an Elmasonic P70H ultrasonication bath, which was manufactured by Elma Schmidbauer GmbH of Singen, Germany. Upon completion of each extraction, samples were subjected to centrifugation for 10 min at 10,000× g using a NEYA 16R centrifuge (Remi Elektrotechnik Ltd., Palghar, India). Ultimately, supernatants were gathered and preserved at −40 °C.
3.4. Determinations
3.4.1. Total Polyphenol Content (TPC)
The TPC was evaluated through a photometric assay described in a previous study [
40]. The results were represented as milligrams of gallic acid equivalents (GAE) per gram of dry weight (dw), calculated using a calibration curve (10–100 mg/L of gallic acid; R
2 = 0.9996) in water. The samples’ absorbances were measured using a Shimadzu UV-1900i UV/Vis spectrophotometer, which is made in Kyoto, Japan. All analyses were carried out three times, and the average was taken to determine the outcomes.
3.4.2. Ferric-Reducing Antioxidant Power (FRAP) Assay
The procedure for determining the extracts’ antioxidant capacity using the widely-used electron-transfer method is detailed in a prior work [
40]. Finding out when the iron oxidation state went from +3 to +2 at 620 nm was the key to this technique. The data were presented as μmol of ascorbic acid equivalents (AAE) per gram of dry weight, and a calibration curve of ascorbic acid (50–500 μM in 0.05 M HCl, R
2 = 0.9997) was used. All analyses were carried out three times, and the average was taken to determine the outcomes.
3.4.3. DPPH Radical Scavenging Assay
A previously described assay [
30] for DPPH scavenging was employed. After mixing 25 μL of adequately diluted sample extract with 975 μL of DPPH solution (100 μmol/L in methanol), the absorbance at 515 nm was measured both immediately and 30 min later. The anti-radical activity of ascorbic acid (100–1000 μmol/L in methanol, R
2 = 0.9926) was measured using a calibration curve, and the results were presented as μmol of AAE per gram of dry weight. All analyses were carried out three times, and the average was taken to determine the outcomes.
3.4.4. Individual Polyphenols by HPLC-DAD
Based on our earlier research, we identified specific polyphenols from the CM leaves’ extracts using high-performance liquid chromatography coupled with a diode array detector (HPLC-DAD) [
30]. Shimadzu Europa GmbH, Duisburg, Germany, supplied the liquid chromatograph (type CBM-20A) and diode array detector (model SPD-M20A) used in this work. Between 200 and 800 nm is the detecting wavelength. Using a Phenomenex Luna C18(2) column (100 Å, 5 μm, 4.6 mm × 250 mm) from Phenomenex Inc. in Torrance, CA, USA, the compounds were injected with a volume of 20 μL and then separated at 40 °C. Both the acetonitrile (B) and water (A) mobile phases contained 0.5 percent formic acid. The gradient program began with a constant value of 10 min, progressed to 40% B after 10 min, 70% B after another 10 min, and finally 40% B. A steady 1 mL/min flow rate for the mobile phase was maintained. The chemicals were identified and then quantified using calibration curves (0–50 μg/mL) by comparing the absorbance spectra and retention times to those of purified standards.
3.5. Statistical Analysis
All statistical evaluations were conducted using JMP® Pro 16 software (SAS Institute Inc., Cary, NC, USA) supporting response surface methodology (RSM), regression modeling, and distribution analysis. Normality of data was assessed using the Kolmogorov–Smirnov test. To determine statistically significant differences among treatments, analysis of variance (ANOVA) was performed, followed by Tukey’s HSD multiple comparison test at a significance level of p < 0.05. Each extraction protocol was repeated at least twice, and all quantitative analyses were conducted in triplicate. Results are reported as mean ± standard deviation. Additionally, PLS, PCA, MCA, and Pareto plot analysis were applied to evaluate multivariate relationships and identify the most influential extraction parameters.