Introduction

Saponins are bioorganic compounds that have at least one glycosidic bond (C–O-sugar bond) at the C-3 carbon between the aglycone and the saccharide chain. They are found in higher plants and some marine animals (Ashour et al. 2019; Yang et al. 2014). They are natural surfactants that are ecological, non-toxic and biodegradable. Isolated saponins are solids of an amorphous nature, having a relatively high molecular weight and the aglycon consists of 27–30 carbon atoms. Saponins differ from other glycosides mainly due to their surface activity and thus their ability to form colloidal solutions in an aqueous environment (Ashour et al. 2019; Sparg et al. 2004). Saponins can be divided into two main groups based on the chemical structure of their aglycone skeleton. The first group consists of steroid saponins. The second group is characterized by triterpene saponins, which are the most common and occur mainly in dicots plants. Some authors in their articles also distinguish a third group called steroid amines, in many articles also classified as steroid alkaloids. Based on the different number of saccharide chains from one to three, bound to the aglycone core, we distinguish between monodesmosidic, bidesmosidic and tridesmosidic saponins (Fig. 1) (Moses et al. 2014).

Fig. 1
figure 1

Overview of structural diversity in saponin aglycones (Moses et al. 2014)

During recent years, research has revealed both the chemical and biological aspects of triterpene saponins. The insecticidal, anthelmintic, antibacterial, antifungal and antiviral effects have been reported. The in-vitro hemolytic activity of saponins is also reported. The hemolytic property of saponins is due to their ability to solubilize the plasma membrane of erythrocytes (Ashour et al. 2019; Yang et al. 2014; Sparg et al. 2004; Moses et al. 2014; Dinda et al. 2010).

Hedera helix, commonly known as Ivy, is a higher plant that contains relatively large amounts of triterpene saponins of diverse structure. The content of saponins in this plant can reach up to 6%. Such saponins include, for example, bidesmosidic hederasaponin B, C, D, helixoside A, B and monodesmosidic alpha-hederin. The studied saponins are mainly found in the leaves and fruits of the Hedera plant. Triterpene saponins isolated from Hedera have many interesting biological properties, such as leishmanicidal, molluscicidal, cytotoxic, antitumor and antiviral activity. One of the most frequent uses of saponin-rich Ivy extract, which is commonly encountered in the treatment of cough and bronchitis, is the use of its expectorant, antitussive and bronchodilator effects (Majester-Savornin et al. 1991; Bedir et al. 2000; Elias et al. 1991; Hostettmann 1980; Tatia et al. 2019; Song et al. 2015, 2014; Gumushan-Aktas and Altun 2016).

Saponin extraction techniques have gradually evolved from simple macerations and Soxhlet extractions to microwave and ultrasound-assisted extractions. Simple methods used relatively large amounts of solvents, elevated temperatures and were time-consuming. Modern extraction methods take exponentially less time with ultrasonic membrane cavitation and microwave increased intracellular temperature interacting with polar water molecules inside the plant cells (Moghimipour and Handali 2015; Gavrila et al. 2022). The purification of isolated saponins is a very important topic, since the structures of saponins often differ only to a negligible extent and their separation by conventional chromatographic methods is often multi-step, incorporating normal phases, reverse phases and also modern high-performance liquid chromatography (HPLC) methods. HPLC methods are often economically demanding with a lot of consumables and the need for highly pure solvents (Ashour et al. 2019; Yang et al. 2014; Bedir et al. 2000; Bezruk et al. 2020; Havlíková et al. 2015; Hussien and Awad 2017).

The surface activity of saponins is based on their amphiphilic nature, which is related to the lipophilic aglycone and the hydrophilic nature of the saccharide units. These amphiphilic molecules gather on the surface at low concentrations, and after the critical micelle concentration (cmc) is exceeded, when the surface is saturated with surfactant molecules, aggregates called micelles are formed. At cmc, many physicochemical properties such as surface tension, conductivity, zeta potential or light scattering show an abrupt change and inflection point in the curve when plotted against surfactant concentration. Knowledge of these physicochemical properties is important because saponins are also used in technology as solubilizers, emulsifiers, foaming stabilizers, cleaning and wetting agents (Rai et al. 2021; Böttger et al. 2012; Verza et al. 2012; Chapagain and Wiesman 2006; Timko et al. 2020).

While triterpene saponins have been extensively studied in numerous plant species (Ashour et al. 2019; Yang et al. 2014; Sparg et al. 2004; Moses et al. 2014; Dinda et al. 2010; Moghimipour and Handali 2015; Rai et al. 2021, 2023; Luo et al. 2006; Negi and Negi 2013; Sabri and Moulai-Mostefa 2020), Hedera helix remains relatively unexplored in this regard. One of the key motivations of our research is to expand the existing knowledge about these bioactive compounds. By carrying out this study, including isolation and purification and physicochemical properties evaluations, we aim to obtain valuable data that can contribute to a better understanding and better accessibility of these natural compounds.

Experimental

Materials and methods

Chemical compounds used in extraction, purification and synthesis were obtained commercially (acetic acid, formic acid, hydrochloric acid—37% fuming, potassium hydroxide, sodium hydroxide, sulphuric acid—p.a., Centralchem, Slovakia; p-anisaldehyde—97.5%, Sigma-Aldrich, China). Solvents (acetonitrile—HPLC gradient grade, butanol, chlorophorm, diethyl ether, methanol, petrolehter—p.a., Centralchem, Slovakia; deuterated methanol, 99.8% D for NMR, VWR) of p.a. grade were purified by redistillation to increase purity. Superpure deionized water was prepared on Rodem 6 with germicidal light and M10PPF, M10CTC and two ionex MB filters with additional Rodem UP4-MO with four ultrapure ionex MB filters.

Chromatographic separations were performed using thin-layer chromatography (TLC silica gel 60 F254 plate, Merck, Germany) and on glass columns using reverse phase silica gel (C18 90 reverse phase silica gel, fully endcapped—particle size 40–63 μm, Diaion HP-20—particle size 250–850 μm, MCI GEL CHP20P—particle size 75–150 μm, Sigma-Aldrich, Switzerland).

1H and 13C NMR spectra were measured on Varian MR400 spectrometer at frequency 400 and 100 MHz. 13C NMR spectra were decoupled against protons. The chemical shifts were referenced using an internal tetramethylsilane standard (δ1H = 0, δ13C = 0). The NMR measurements were performed in deuterated methanol CD3OD. Zeta potentials were measured on Brookhaven ZetaPlus analyzer using unfiltered samples dissolved in deionized water. Particle size was determined by dynamic light scattering on Brookhaven BI9000-200SM coupled with Lexel 95 laser using filtered samples. The compound’s surface tension was measured on a KRÜSS Processor tensiometer K100 using Wilhelmy plate method in superpure deionized water.

Extraction and purification

Hederasaponin B and hederacoside C were extracted from the leaves of Hedera helix and purified by reverse phase chromatography. Leaves of Hedera helix were collected at the Devínska Kobyla, Slovakia on March 2, 2020. 111.25 g of dried leaves were finely ground and extracted using ultrasonic extraction with 200 ml of 90% methanol four times for 15 min each. The resulting extract was freed from solid particles by filtration through a frit with a porosity of P4 (10–16 μm). The crude methanolic solution extract was then diluted with 60 ml of water and substantial amount of chlorophyll was extracted by petrolether. The methanol phase was removed on a rotary vacuum evaporator until only water remained. The aqueous phase after evaporation of the methanol was subsequently extracted with 200 ml and twice with 50 ml of butanol. The combined butanol fractions were evaporated and dried by vacuum to give a solid crude extract containing vast variety of secondary metabolites.

The crude extract was subjected to reverse column chromatography on 350 g of DIAION HP-20 (4.5 cm column width) as the stationary phase. A solution of methanol and 0.5% formic acid in water was used as a mobile phase. The sample was applied to the column in 30% methanol solution and eluted with a 500 ml of mobile phases with each 10% gradient (30–90% gradient). 200 ml fractions from the 30–60% gradient and 50 ml fractions from 70% gradient were collected and the presence of saponins was detected using thin-layer chromatography on silica gel with a mobile phase of chloroform:methanol:water—60:40:9. A mixture of methanol, sulfuric acid and anisaldehyde was used as a detection agent for the presence of saponins. Fractions containing the mixture of saponins were collected, the solvents were removed on a rotary vacuum evaporator and solid mixture was dried using high vacuum. The mixture of saponins was then subjected to reverse phase chromatography on 55 g of MCI GEL CHP20P (3 cm column width). The loading of the sample was done in 30% methanol and 0.5% formic acid in an aqueous solution. The sample was eluted with 200 ml of mobile phases with each 10% gradient and after reaching 50% methanol concentration elution was done with 5% gradient (30–50% and 55–85% gradient). The presence of saponins in fractions was detected, as previously mentioned, by silica gel thin-layer chromatography. The separated compounds were collected, dried and precipitated from methanol (5 ml) by addition of diethyl ether. The final pure saponins were filtered through a frit with porosity P4 (10–16 μm) and dried to give hederasaponin B (yield 0.125 g) and hederacoside C (yield 1.181 g). The structure of saponins was verified by 1H and 13C NMR and compared with the literature (Lavaud et al. 2001; Babadjamian et al. 1988).

Hederasaponin B

1H NMR (CD3OD, TMS) δ: 0.79 (3H, s), 0.84 (3H, s), 0.91 (3H, s), 0.94 (3H, s), 0.96 (3H, s), 0.98 (3H, s), 1.01 (3H, s), 1.15–2.06 (27H, m), 2.85 (1H dd, J = 14, 4.4 Hz), 3.08–4.09 (28H, m), 4.39 (1H, d, J = 8 Hz), 4.54 (1H, d, J = 4.8 Hz), 5.10 (1H, m), 5.25 (1H, m), 5.33 (1H, d, J = 8 Hz).

13C NMR (CD3OD, TMS) δ: 14.0, 14.7, 15.7, 16.4, 16.5 18.0, 22.7, 23.1, 24.8, 25.6, 27.2, 27.5, 30.1, 31.8, 32.0, 32.5, 33.5, 36.5, 38.5, 38.8, 39.3, 41.1, 41.5, 45.8, 46.6, 46.9, 47.1, 47.3, 47.6, 47.8, 48.0, 48.2, 55.6, 60.4, 62.3, 65.5, 67.0, 68.0 68.8, 69.2, 69.5, 70.7, 71.0, 71.7, 72.3, 72.5, 73.9, 75.4, 76.8, 78.1, 89.2, 94.3, 100.6, 101.5, 102.8, 103.4, 122.4, 143.4, 176.7

Hederacoside C

1H NMR (CD3OD, TMS) δ: 0.69 (3H, s), 0.80 (3H, s), 0.91 (3H, s), 0.94 (3H, s), 0.98 (3H, s), 1.06–2.05 (31H, m), 2.86 (1H dd, J = 14.4, 4 Hz), 3.20–4.09 (29H, m), 4.41 (1H, d, J = 8 Hz), 4.55 (1H, d, J = 5.2 Hz), 5.15 (1H, m), 5.25 (1H, m), 5.33 (1H, d, J = 8 Hz).

13C NMR (CD3OD, TMS) δ: 12.3, 14.0, 15.1, 16.5, 16.6 17.4, 22.7, 23.2, 24.9, 25.1, 27.5, 30.1, 32.0, 33.5, 36.2, 38.3, 39.3, 41.1, 41.6, 42.5, 45.8, 46.6 46.7, 46.9, 47.1, 47.4, 47.6, 47.8, 48.0, 48.2, 48.4, 60.4, 63.2, 63.4, 65.5, 67.7, 68.0, 68.7, 69.2, 69.5, 70.6, 70.8, 71.0, 72.3, 72.5, 73.9, 75.2, 75.4, 76.8, 78.1, 80.8, 94.4, 100.4, 101.5, 102.8, 102.9 122.3, 143.5, 176.7

Preparation of alpha-hederin

Alpha-hederin was prepared by alkaline hydrolysis (Luo et al. 2006). Hederacoside C (0.302 g–0.247 mmol) was dissolved in 10 ml of 1 M potassium hydroxide solution and stirred for 2 h at 80 °C. The reaction was then quenched by the addition of 1 M hydrochloric acid solution to give a slightly acidic pH (approx. pH 6) to neutralize all hydroxides. The reaction mixture was subsequently extracted with 100, 50 and 25 ml of butanol. The butanol fractions were collected, evaporated and dried to give a crude solid. The substance was then subjected to 35 g of C18 reverse phase silica gel (2.5 cm column width) column chromatography. Due to poor solubility, the substance was sorbed (1:1 ratio) on reverse phase silica gel and applied to a column in 30% acetonitrile and 0.5% acetic acid in an aqueous solution. Elution of the sample was 100 ml with 10% gradient and 15 ml fractions were collected. The presence of hydrolyzed saponin was detected using thin-layer chromatography on silica gel with a mobile phase of chloroform:methanol:water—60:16:2. After collecting and evaporating the fractions, the structure was verified by 1H and 13C NMR spectroscopy and compared with the literature to be alpha-hederin (yield 0.054 g) (Lavaud et al. 2001).

Alpha-hederin

1H NMR (CD3OD, TMS) δ: 0.69 (3H, s), 0.81 (3H, s), 0.90 (3H, s), 0.94 (3H, s), 0.97 (3H, s), 0.99 (3H, s), 1.10 (1H, m), 1.17 (3H, s), 1.23 (3H, d, J = 6 Hz), 1.28–1.97 (18H, m), 2.85 (1H dd, J = 14, 4.6 Hz), 3.29–3.90 (12H, m), 4.55 (1H, d, J = 5.2 Hz), 5.15 (1H, d, J = 1.6 Hz), 5.24 (1H, t, J = 3.6 Hz).

13C NMR (CD3OD, TMS) δ: 12.3, 14.9, 16.3, 16.5, 17.4, 22.5, 22.6, 23.1, 25.0, 25.1, 27.4, 30.2, 32.0, 32.4, 33.5, 36.2, 38.2, 39.1, 41.3, 41.6, 42.5, 45.8, 46.2, 46.7, 46.9 63.2, 63.3, 67.7, 68.7, 70.6, 70.7, 72.2, 72.4, 72.5, 75.2, 80.8, 100.5, 102.9, 122.2, 143.8, 180.4

Alpha-hederin is relatively poorly soluble in aqueous solutions, and since the measurements of surface activity are measured in water, it was necessary to increase the solubility by forming a salt. Alpha-hederin (0.054 g–0.072 mmol) was dissolved in a minimal amount of methanol and an equimolar amount of sodium hydroxide (0.0029 g–0.072 mmol) also dissolved in methanol was added. The final alpha-hederin sodium salt was precipitated from methanol by addition of diethyl ether. The precipitated salt was filtered through a frit with porosity P4 (10–16 μm) and dried by vacuum (Fig. 2) (Lavaud et al. 2001; Babadjamian et al. 1988).

Fig. 2
figure 2

Chemical structures of measured triterpene saponins (Lavaud et al. 2001; Babadjamian et al. 1988)

Determination of physicochemical properties

Critical micelle concentration

The critical micelle concentration was determined based on surface tension measurements. The measurement was performed on a KRÜSS Processor tensiometer K100 using the Wilhelmy plate method. The measurement of the surface properties of saponins was carried out using the modified previously described methodology (Lukáč et al. 2014). Deionized water was used in the preparation of samples for the experiment. During measurement of hederacoside C and hederasaponin B, the measurement temperature was maintained at 25 ± 0.1 °C. During the measurement of alpha-hederin sodium salt, the temperature was kept at 37 ± 0.1 °C, which should be the normal temperature of the human body. The increased temperature in the measurement of alpha-hederin salt was used mainly to improve the solubility of the saponin in water. Surface tension measurements were taken every 6 min (for 1–2 h) to ensure that equilibrium was established and that the difference in surface tension was less than 0.05 mNm−1. The critical micellar concentration (cmc) and the surface tension at the point of the critical micellar concentration (γcmc) were determined from the point where the curve of surface tension and the logarithm of the concentration (c) intersect. By measuring the dependence of the surface tension on the surfactant concentration in aqueous solutions, the area corresponding to one saponin molecule at the water/air phase interface (Acmc) was determined. Equation for calculation Acmc is

$${\text{A}}_{{{\text{cmc}}}} = {1}0^{{{18}}} /{\text{N}}_{{\text{A}}} \Gamma_{{{\text{cmc}}}}$$
(1)

where NA is Avogadro’s constant. Γcmc is absorbed amount of surfactant which was calculated using the Gibbs absorption isotherm:

$$\Gamma_{{{\text{cmc}}}} = - \left[ {{\text{d}}\gamma /{\text{d log c}}} \right]_{\tau } /\left( {{2}.{3}0{3}i{\text{R}}T} \right)$$
(2)

where γ is the surface tension (mNm−1), c is the concentration of the surfactant and i is the prefactor; number of ions in one monomer in micelle (i = 1 for hederacoside C and hederasaponin B and i = 2 for alpha-hederin), T is the absolute temperature and R the gas constant. The base solution was prepared in 100 ml volumetric flask, while the individual concentrations in the drop gradient concentration were diluted by different additions of the stock solution in a 25 or 10 ml volumetric flask.

Dynamic light scattering

The size of the micelles formed by saponins after cmc was measured by their hydrodynamic diameter (dh) which was determined by dynamic light scattering according to the modified method described in Pisárčik et al. article (Pisárčik et al. 2019). Dynamic light scattering was measured on Brookhaven BI-9000 with goniometer 200SM system coupled with Lexel 95 argon laser with 514.4 nm wavelength. The measurement was carried out at a fixed scattering angle of 90° and at the temperature of 25 °C and 37 °C for alpha-hederin controlled by thermostat. Saponin samples were prepared in deionized water, dissolved using ultrasound for 5 min and filtered through a 450 μm syringe filter to avoid undissolved particles. Each sample was subjected to five independent measurements of the autocorrelation function at the surfactant concentration of 4 times the cmc value.

Micelle diameter values were calculated from the particle size distributions resulting from the application of the constrained regularized algorithm CONTIN on the autocorrelation function (Provencher 1982).

The translation diffusion coefficient was calculated from the time correlation function using the method of cumulants. The method of cumulants was used for the calculation of the mean particle diameter from the expansion of logarithm of time correlation function into a series up to the second quadratic term. The diffusion coefficient was determined from the correlation function decay rate and the hydrodynamic diameter (d) was calculated from the diffusion coefficient (D) using the Stokes–Einstein formula:

$${\text{d}} = {\text{k}}T/\left( {{3}\pi \eta {\text{D}}} \right)$$
(3)

(η) is solvent viscosity, (k) is the Boltzmann constant and (T) is absolute temperature.

Zeta potential

Zeta potential measurements were performed on a Brookhaven ZetaPlus instrument according to a modified known method (Pisárčik et al. 2019). Determination of zeta potential on this device is based on particle electrophoresis (microelectrophoresis) which is the movement of charged particles influenced by an applied electric field in a liquid. Information about the electrophoretic mobility is contained in Doppler shifted frequency when the scattered light is shifted by the motion of particles when it is perpendicular to the laser beam. From the measured mobility using the Smoluchowski limit for the relationship mobility vs. zeta potential, zeta potential values of saponins were determined. The statistical mean values were calculated from a series of 30 zeta potential measurements. The samples (4 × cmc) were dissolved in deionized water using ultrasound (5 min) and subsequently measured as unfiltered at a temperature of 37 ± 0.1 °C. Filtered samples did not show any zeta potential.

Results and discussion

Extraction, isolation and purification

Hederasaponin B and hederacoside C were isolated from Hedera helix leaves. Extraction of the saponins was carried out by sonication of the dried and powdered leaves in 90% methanol for 15 min with four repetitions. The choice of solvents and conditions was based on the preference for the extraction of more polar compounds and also to improve the cavitation of cell membranes during sonication. Ultrasonic extraction of crude triterpene saponins from Hedera helix is well-known method as Gavrila et al. describes and the extraction conditions were chosen accordingly (Gavrila et al. 2022).

One of the reasons for choosing atmospheric pressure reverse phase column chromatography in this study was to establish a more affordable and available method of saponin isolation, while still achieving optimal separation and purification of the target compounds, compared to high-performance liquid chromatography (HPLC). HPLC is a widely used technique for compound separation and purification but can be expensive due to the equipment and consumables involved. According to our knowledge, HPLC methods are most frequently used for separation and purification of saponins from Hedera helix (Gavrila et al. 2022; Bezruk et al. 2020; Havlíková et al. 2015). The separation and purification of hederasaponin B and hederacoside C from the crude extract is also described in the literature using chromatography on a glass column under vacuum or at atmospheric pressure, but the chromatographies are often multiple, involving silica gel with a normal phase and the final purification is usually still performed on HPLC (Bedir et al. 2000; Hussien and Awad 2017). As part of the optimalization, we tried the use of a different mobile phase mixture with acetonitrile as an organic component, or the use of another reverse phase such as RP-C18. We found that reverse phase column chromatography on DIAION HP-20 and on MCI GEL CHP20P was sufficiently efficient and showed a significant simplification and reduction in the number of purification steps. Our new developed method can be valuable, especially in research where budget constraints are considered and in industries where chromatography is becoming a popular purification choice, without compromising the quality of the results obtained.

The final triterpene saponins were purified by precipitation from minimal amount of methanol to remove impurities that are well soluble in diethyl ether to give yield of 0.1% for hederasaponin B and 1.1% for hederacoside C from dry matter. Based on the information presented by Tatia et al., that the content of triterpene saponins in Hedera helix is 0.1–0.2% for hederasaponin B and 1.7–4.8% for hederacoside C, we can conclude that the method developed by us is highly effective in obtaining pure compounds. Alpha-hederin is found in Hedera helix only in very small quantities (0.1–0.3%), and therefore, its preparation by alkaline hydrolysis of hederacoside C with 29% yield is much more effective than isolation (Tatia et al. 2019). Final purification requires only one-step chromatography on C18 reverse phase silica gel, which effectively purifies the compound from hydrolysis byproducts leaving only a pure monodesmosidic triterpene saponin. Alpha-hederin has a relatively low solubility in water, and therefore, it is necessary to increase this property for measurement purposes of aqueous solutions. We achieved the improvement of solubility by simple ionization of the free carboxyl group with the methanolic solution sodium hydroxide.

Physico-chemical properties

Critical micelle concentration (cmc) depends significantly on the structure of the molecule. Hederacoside C is structurally more hydrophilic, therefore the measured value of cmc is 3.10 × 10–4 ± 6 × 10–6 mol·dm−3, which is higher compared to hederasaponin B. The absence of a hydroxyl group in the aglycone portion of hederasaponin B highly affects its lipophilicity. The cmc value of this saponin is 4.98 × 10–5 ± 4 × 10–6 mol·dm−3, and therefore, almost one order of magnitude lower compared to hederacoside C (Table 1).

Table 1 Values of measured and calculated surface properties of individual saponins

Our findings were compared with a study by Böttger et al., wherein hederacoside C was also analyzed using the Wilhelmy plate method. Differences were noted in cmc values, with Böttger et al. reporting a value of cmc = 8.2 × 10–5 mol·dm−3, differing significantly from our measurement. This may be due to variations in experimental conditions, specifically a 5 °C temperature difference, and the use of a buffer solution compared to our measurement in pure deionized water (Böttger et al. 2012). This significant difference in the measured and published values by Böttger et al. is mainly based on the use of ionic compounds as solubilizers when measuring the cmc, which depending on the concentration can effectively reduce the cmc as described by Akhlaghi and Riahi and also by Bajpai (Akhlaghi and Riahi 2019; Bajpai 2018). Taking hederasaponin B into consideration, we did not find any cmc measurements in the literature to compare with, and therefore, measurements we performed on this saponin gave us unstudied new information stated in Table 1.

The surface tension of alpha-hederin in its sodium salt form was measured at the physiological temperature of 37 °C due to insufficient solubility at 25 °C. Alpha-hederin salt showed an unusual property during the measurement, which can be observed by the break in the curve in the pre-micellar region (Fig. 3), we conclude that it is pre-micellar aggregation, which is characterized by the formation of small aggregates of surfactant, similar to what Lukáč et al. observed with other amphiphilic compounds (Lukáč et al. 2014). To our knowledge cmc of alpha-hederin was only measured in 0.1 M phosphate-buffered saline solution at 20 °C by Wilhelmy plate method. Despite the use of different conditions and solubility enhancement method, the value measured by Böttger et al. (cmc = 1.3 × 10–5 mol·dm−3) is only slightly different from the value that we stated simulating physiological conditions (Table 1) (Böttger et al. 2012).

Fig. 3
figure 3

Correlation of surface tension γ and the logarithm of concentration; from left hederacoside C (●), hederasaponin B (▲) and alpha-hederin (♦)

The surface tension at cmc is also dependent on the structure of the saponins. The more lipophilic hederasaponin B shows a lower value of this quantity γcmc = 55.3 mN·m−1 compared to hederacoside C. Monodesmosidic alpha-hederin with its missing three saccharides is the most lipophilic compound compared to other measured saponins and therefore shows the lowest γcmc value. The results well describe decreasing γcmc with decreasing polarity. When we compare the values measured by us with the values from the literature by Stanimirova et al. (30–40 mN·m−1), Chen et al. (50 mN·m−1) and Rai et al. (37.4–49.9 mN·m−1), we see that the surface tension values for Quillaja triterpene saponins or saponin rich extracts from Camellia oleifera and Jatropha curcas are similar (Stanimirova et al. 2011; Chen et al. 2010; Rai et al. 2023).

By measuring the correlation of the surface tension and the concentration of the surfactant in aqueous solutions, the limiting value of the area of the liquid/gas surface per one surfactant molecule or surface area per molecule (Acmc) was calculated. The presence of a hydroxyl group in the aglycone of saponin occupies only a small space compared to the total volume of molecule, and therefore, this structural change was manifested only to a negligible extent; Acmc values of hederasaponin B and hederacoside C were practically identical (Table 1). When comparing Acmc values of Quillaja saponins that were measured by Stanimirova et al. (Acmc ≈ 1.05 nm2), we can conclude that the values calculated by us are very similar. A value of Acmc ≈ 1 nm2 indicates a lay-on configuration on the solution surface which is typical for bidesmosidic triterpene saponin structures. As for monodesmosidic saponins, the typical value is Acmc ≈ 0.3 nm2, and therefore, alpha-hederin salt Acmc value is quite unusual (Acmc = 1.28 ± 0.11 nm2). With the absence of a trisaccharide chain, monodesmosidic alpha-hederin salt should occupy a smaller area compared to the other measured saponins and adopt an end-on configuration (Stanimirova et al. 2011). This unusual behavior is most likely caused by the increased temperature as described by Hamieh and also by ionization and electrostatic repulsion effects which increases the surface area per molecule as mentioned by Patil et al. (Hamieh 2020; Patil et al. 1979).

The size of the micelles formed by saponins after cmc was measured by their hydrodynamic diameter (dh) with dynamic light scattering. The concentration used for measuring all samples was 4 × cmc. The mean values of dh are hederasaponin B dh = 168.33 ± 9.83 nm, hederacoside C dh = 7.5 ± 1.31 nm and alpha-hederin dh = 128.55 ± 3.61 nm, as calculated using the method of cumulants. When we compare the data obtained by the method of cumulants with the data from the CONTIN algorithm, we can see that these two methods gave relatively similar results (Fig. 4).

Fig. 4
figure 4

Particle size spectra obtained from the time correlation functions by CONTIN algorithm (diameter [nm], intensity [%]); from left hederacoside C, alpha-hederin and hederasaponin B

By comparing hederasaponin B and alpha-hederin we can observe a rather unusual behavior in terms of particle size, since alpha-hederin shows higher values of lipophilicity and in theory its micelles should be larger. The smaller value of the size of the micelles can be explained by the different shape aggregate formation like Verza et al. describes. When we look at the sizes of triterpene or steroidal saponin aggregates, we see that the sizes range from 144.8 ± 2.99 nm to 156.1 ± 1.67 nm as Verza et al. mentioned or 167 nm and 177 nm as in Chapagain and Wiesman article. Saponins showing the mentioned hydrodynamic diameter values are mostly described as spherical-shaped nanosized vesicles (Verza et al. 2012; Chapagain and Wiesman 2006). According to this information we assume that the measured saponins, which show similar size values, could also be behaving as spherical vesicles.

Hederacoside C is the most hydrophilic molecule out of the three and shows the lowest values of particle size. By comparing the values mentioned in the Mitra and Dungan articles with hederacoside C we can see that micelles formed by Quillaja bark triterpene saponins alone can reach hydrodynamic diameter values as low as 3.2–5.8 nm to 7–10 nm depending on concentration and temperature (Mitra and Dungan 2000, 1997). Looking at the difference between hederacoside C and the rest of the measured saponins, this relatively large difference can be explained by the ability of hederacoside C to form only micelles and not vesicles (Verza et al. 2012).

The zeta potential (ζ) of micelles formed by a saponin was measured only for the alpha-hederin salt. Hederasaponin B and hederacoside C do not have the required group in their chemical structure that could be ionized and show conductivity in an electric field (Fig. 2). For alpha-hederin, sufficient conductivity was ensured by its free carboxyl group, which was ionized by a sodium cation on the aglycone portion of the structure. The optimal conductivity for zeta potential measurement of alpha-hederin was at 4 × cmc concentration. Zeta potential was calculated from 30 measurements from which we excluded extremities as measurement errors (Fig. 5). The calculated mean zeta potential value of the alpha-hederin with deviation is ζ = − 15.64 ± 2.89 mV. A negative zeta potential indicates that the measured particles are negatively charged, which correlates with the alpha-hederin structure. The zeta potential for alpha-hederin was not mentioned before according to the available literature sources. When compared to the other saponin zeta potential measurements with single unsubstituted carboxyl group like Quillaja saponins (ζ from − 10.1 to − 12.5 mV) from Cibulski et al. study, we can see that alpha-hederin zeta potenatial shows higher negative values, therefore it could exhibit greater stability of the aggregates (Cibulski et al. 2022).

Fig. 5
figure 5

Zeta potential (mV) of alpha-hederin in deionized water (37 ± 0.1 °C)

Conclusions

In this study, we present an effective method of the isolation, purification and analysis of physicochemical properties of hederasaponin B and hederacoside C derived from Hedera helix leaves. A novel, cost-effective extraction method utilizing sonication in 90% methanol was used to obtain hederasaponin B (0.1%) and hederacoside C (1.1%) from dry matter. We optimized the isolation process using reverse phase column chromatography on DIAION HP-20 and on MCI GEL CHP20P at atmospheric pressure, providing an efficient alternative to high-performance liquid chromatography (HPLC). Alpha-hederin was efficiently obtained through alkaline hydrolysis of hederacoside C. The physicochemical properties of these saponins were studied. Structural differences influenced the critical micelle concentration (cmc), with hederacoside C showing a higher cmc due to its hydrophilicity. Surface tension measurements at cmc revealed distinct behaviors, with hederasaponin B displaying lower values than hederacoside C. Remarkably, alpha-hederin exhibited pre-micellar aggregation and a unique zeta potential of − 15.64 ± 2.89 mV. In addition, micelle size analysis demonstrated that hederacoside C forms micelles, differentiating it from vesicle-forming saponins. Our findings not only contribute essential data to the understanding of saponin behavior in aqueous solutions but also offer an optimized and economically viable approach for researchers. The characteristic properties observed in these saponins, especially alpha-hederin, represent promising avenues for further research, highlighting their potential pharmaceutical applications as they are essential surface-active natural products.