Direct ultrasound-assisted extraction and characterization of phenolic compounds from fresh houseleek ( Sempervivum marmoreum L . ) leaves

The effects of ultrasound power and frequency on the yield of total extractive substances (TES), total phenolic content (TPC), total flavonoid content (TFC) and antioxidant activity (AOA) of fresh houseleek leaves extracts obtained by direct ultrasound-assisted extraction (DUAE) were studied. Preliminary extraction of plant material was performed using methanol, acetone and 2-propanol by Soxhlet extraction. It was found that maximum TES yield could be obtained by methanol extraction (2.91±0.02), followed by acetone and 2-propanol with a TES yield of 2.32±0.01 and 2.01±0.03 g per 100 g of fresh plant material, respectively. In the fresh houseleek leaves extracts obtained by DUAE and methanol as the chosen solvent, TPC, TFC and AOA were in the ranges of: 40.5–85.9 mg gallic acid/g dry extract, 12.7–19.3 mg rutin/g dry extract and 24.6–108.2μg/ml, respectively. The results showed that the increase in the ultrasound power and extraction time have positive and significant (p < 0.05) effects on the TPC, TFC and AOA, while the increase in the ultrasound frequency leads to a decrease in the TPC, TFC and AOA of the extracts. A chromatographic analysis of crude extract identified the following: kaempferol 3-O-(6’’-O-malonylglucoside)-7-O-glucosyde, kaempferol 3-O-glucoside-7-O-rhamnoside, luteolin 5-O-(6’’-O-malonylglucoside), kaempferol 3-O-(6’’-O-acetylglucoside)-7-O-rhamnoside, genkwanin 5-O-glucoside, luteolin 5-O-(6’’-O-malonylglucoside), kaempferol 3-O-(6’’-O-malonylglucoside), kaempferol 3-O-rhamnoside, quercetin, genkwanin 4’-O-glucoside and hyperoside.

three extraction techniques, including indirect ultrasound assisted extraction (UAE), classical solvent and Soxhlet extraction from fresh S. marmoreum leaves were investigated, and it was found that the obtained extracts possessed good antioxidant activity and exhibited antimicrobial activities against yeast such as Aspergillus niger and Candida albicans, while the range of TPC and TFC were 0.56-0.73mg gallic acid and 0.51--0.62mg rutin per g of fresh plant material, respectively.However, no reports of the identification and quantification of the main constituents in S. marmoreum leaves were found.
The extraction procedure is a crucial step in the isolation of valuable bioactive compounds from plant material.Development and improvement of extraction techniques with optimization of the operating conditions is a very important activity in every specific case.In order to overcome the drawbacks of classic extraction techniques (time and energy consuming, possible degradation of the target compound, high costs), non--conventional extraction methods (ultrasound, microwave and enzyme-assisted extraction, subcritical and supercritical fluid extraction) are increasingly being used.UAE is one of the "green" extraction techniques, which is recognized as an efficient technique that increases yield, while significantly reducing extraction time and solvent consumption.The main advantages of sonication, i.e., the reduction of particle size, cell disruption, greater penetration of the solvent into the plant material and easier release of extractable compounds, as well as the enhancement of mass transfer, are attributed to the acoustic cavitation phenomena [13,14].With respect to different equipment and scales, UAE could be performed with indirect (an ultrasound cleaning bath) or direct (a horn or probe system, ultrasound reactor) sonication [15], but both UAE techniques satisfy green extraction principles [16,17] and could be easily scaled up [18].So far, UAE has been widely applied in extraction of various bioactive compounds from plants [19][20][21][22][23], while recent comparative studies of indirect and direct UAE indicate higher efficiency of direct sonication for different target compounds [24][25][26][27].Great potential and advantage of direct UAE (DUAE) for extraction of pharmacologically useful compounds from different plant materials are due to a stronger cavitation effect caused by a much higher specific ultrasound energy than that of indirect UAE [26].
As a continuation of our previous work [12] where fresh houseleek (Sempervivum marmoreum L.) leaves were extracted by various extraction techniques (classic, indirect UAE and Soxhlet extraction), the aim of this study was to examine the effects of low/high ultrasound frequency and power during DUAE on the yield of total extractive substances (TES) and phytochemical constituents such as total phenolic content (TPC), total flavonoid content (TFC) and the antioxidant activity (AOA) of houseleek leaves.In addition, HPLC analyses of crude houseleek extracts were performed.

Soxhlet extraction
Fresh houseleek leaves (10 g) were extracted with solvents (100 mL) of different polarity (methanol, acetone and 2-propanol) using the Soxhlet apparatus until the complete exhaustion of the plant material (150 min).The extracts were evaporated by a rotary vacuum evaporator at 40 o C. The obtained yield represents the total extractive substance (TES) content in the plant material.

Direct ultrasound-assisted extraction (DUAE)
DUAE was carried out in the ultrasonic laboratory reactor URS 1000 (ELAC Nautik Communications GmbH, Kiel, Germany, internal diameter: 106 mm; height: 200 mm; total volume of the reactor: 1.7 L).An ultrasonic transducer (25 cm 2 , frequency range: 40-2500 kHz; power range: up to 250 W) was an integral part of the reactor bottom.The ultrasonic power was monitored on a digital display, while the temperature in the extraction system was measured with a type-K thermocouple digital thermometer (Symmetry, Leskovac, Serbia) and maintained at 25±0.1 o C by thermostated water that circulated through the reactor jacket.Fresh and chopped houseleek leaves (15 g) were submitted to extraction by methanol (150 mL) for 2.5, 5, 10, 20, 40 and 60 min.One series of experiments was performed at the constant ultrasound power of 20 W and three different frequencies (42, 211 and 1038 KHz), while the second series was performed at the constant ultrasound frequency of 211 kHz and three different powers (20, 30 and 75 W).The levels of ultrasound frequencies and powers, as well as their combinations, were chosen during preliminary experiments.Liquid extracts were separated from the solid phase by vacuum filtration and evaporated to dryness on a rotary vacuum evaporator (40 °C).

Determination of total phenolic content
The total phenolic content (TPC) was determined using the Folin-Ciocalteu (FC) reagent [28].The measurements were carried out in triplicate and the calculations were done using a gallic acid calibration curve (R 2 = 0.9922).TPC was expressed as mg of gallic acid equivalents (GAE) per gram of dry extract.

Determination of total flavonoid content
The total flavonoid content (TFC) in the extracts was determined using the aluminium chloride method [29].Measurements were carried out in triplicate and calculations were done using a rutin calibration curve (R 2 = = 0.9994).Total flavonoid content was expressed as mg of rutin equivalents (RE) per g of dry extract.

Determination of DPPH radical-scavenging activity
The scavenging activity of the houseleek leaves extract against the DPPH free radical was determined according to the method described elsewhere [30].Capability to scavenge the DPPH free radical (I / %) was calculated by the equation: A s -absorbance of the sample with DPPH, A b -absorbance of the sample with added methanol, A c -absorbance of DPPH solution with added methanol.Efficient concentration values (EC 50 ) were calculated according to the experimental data [31] by using a nonlinear regression model and SigmaPlot 2000 Software (trial version).
The flavonoid component in the sample extracts was identified by comparing houseleek extract retention time spectra with the retention times of standard spectra.Solutions of all the standards were prepared in methanol (1.0 mg/ml) and diluted to a series of concentrations ranging from 5 to 200 µg/ml.On the basis of the obtained peak areas and the concentration of the standard solution, calibration curves were constructed and concentrations of the components in the samples determined.

Statistical analyses
All measurements were performed in triplicate, and the results were expressed as the mean ± standard deviation.Tukey's test was used for multiple statistical comparisons (WINKS SDA software package, TexaSoft, 6 th ed., Cedar Hill, Texas, 2007 Trial version) and values were considered statistically significant at p < 0.05.

Solvent selection
Selection of the best solvent for extraction of TES from the plant material is the most important step in any method of extraction.Since the high water content in fresh houseleek leaves (88.1%) significantly affects the polarity of the extraction system, as well as due to the fact that the amount of water in water/organic solvent mixtures had a higher impact on the extraction efficacy than the solvent itself [33], it was important to test more solvents of different polarities and properties.In order to select the best solvent for further experiments, the plant material was extracted by the Soxhlet apparatus using three solvents of different characteristics, methanol, acetone and 2-propanol, and the obtained yields of TES were 2.91±0.02,2.32±0.01 and 2.01±0.03g per 100 g of fresh plant material, respectively.
The results published by other authors [33,34] also suggest that higher polarity solvents are a better choice for extraction of plant bioactive compounds than lower polarity solvents.Addition of water to organic solvents could increase the polarity of the solvent system, which according to the available reports [35] played an important role in UAE efficiency, while Barbero et al. [36] pointed out that addition of water to methanol does not increase the extraction yield of capsaicinoids from peppers obtained by USE.

Effect of ultrasound power and frequency on the total extractive substances yield
Changes in TES yield from fresh houseleek leaves during DUAE are shown in Table 1.
Considering that after approximately 20 min more than 90% of TES was extracted, and taking into account the technical and economic feasibility of the process, 20 min was chosen as the optimal time for DUAE of fresh houseleek leaves.By increasing the ultrasound power from 20 to 30 and 75 W at the constant frequency (211 kHz), the TES yield increased by 2.6 and 5.8%, respectively.Statistically significant differences (p < 0.05) between yields of TES obtained at different ultrasound powers were observed.
Increasing the ultrasound power increased the number of created and collapsed micro cavitation bubbles which are responsible for physical, mechanical, chemical and thermal effects of the ultrasound.Implosion of cavitation bubbles on the surface of plant material generates strong liquid microjets which could improve solvent penetration into the plant material and enhance the diffusion process, resulting in an increase in the TES yield [13,15].The enhancement of extraction efficiency with an increase in the ultrasound power was also reported for UAE of fresh rosemary [37] and boldo leaves [38].
By increasing the ultrasound frequencies from 42 to 1038 kHz at the constant ultrasound power (20 W), the TES yield increased slightly (2.5%).It was determined that there were no statistically significant differences in the yields of TES obtained at frequencies of 42 and 211 kHz.It has been reported previously by Wang et al. [39] that extraction yield did not increase significantly with the increase in ultrasound frequency.However, effects of ultrasound frequency largely depend on the nature of the plant matrix.This is the reason that some authors achieved the best results at moderate ultrasound frequencies, while others give priority to lower ultrasound frequencies [40].
In our previous experiment, after 40 min at 25 °C, the TES yields from fresh houseleek leaves were 2.3 and 2.4 g/100 g of fresh plant material for the classical (silence) extraction and indirect UAE technique, respectively [12].When compared to the DUAE, it can be concluded that under the same operating conditions (plant material, plant material to solvent ratio, temperature, solvent) and independently of the applied ultra-sonic power and frequency, the same yield of TES can be obtained in half the time.These findings could be attributed to a higher ultrasound power per gram of the extraction suspension introduced to the extraction system during the DUAE.Similar results were also observed for the DUAE of bioactive compounds from Hypericum perforatum [24], silymarin from Silybum marianum seeds [25], ginseng saponins from ginseng roots and cultured ginseng cells [26], and oil from crushed dill seeds [27].

Effects of ultrasound power and frequency on the total content of phenolic and flavonoid compounds
The TPC and TFC of houseleek extracts obtained under different process conditions by DUAE are shown in Table 2.
Generally, increasing the ultrasound power at the constant frequency of 211 kHz positively affects the content of bioactive compounds in houseleek extracts.Namely, by increasing the ultrasound power from 20W to 30 W, the contents of total phenolic (11.5%) and flavonoid (5.0%) compounds were increased, while further increase in the ultrasound power up to 75 W, increases the contents of total phenolic (31.7%) and flavonoid (7.8%) compounds compared to the contents obtained at 20 W. Statistically significant differences in the TPC obtained by extraction at different ultrasound powers were observed at the 5% probability level (p < < 0.05).
This result is probably due to the improved mechanical effects of the ultrasound, such as the better solvent penetration into the plant material cells and easier release of target compounds into the solution.However, if the power continues to increase, a degeneration of polyphenolic compounds caused by extensive local temperature and pressure generation in the moment of cavitation bubble collapse can occur [15].

Extraction technique
Total phenolic content mg (GAE/g dry extract) As can be seen from the Table 2, increasing ultrasound frequency from 20 to 1038 kHz at the constant ultrasound power (20 W) negatively affects TPC and TFC in the houseleek extracts.This reduction of bioactive compounds in the obtained extracts could be attributed to the fact that low frequency ultrasound (16--100 kHz) generates bigger cavitation bubbles, resulting in a more violent cavitation bubble collapse and higher localized temperatures and pressure which enhances the cavitational effect [44].Further, as the ultrasound frequency increases the production of free radicals (hydroxyl and atomic hydrogen), which could induce the degradation of phenolic compounds, also increases [45].The extraction of houseleek leaves was performed with methanol, a solvent which does not give rise to such a large proportion of radicals under cavitation [46], but the water present in the fresh houseleek leaves could.
Compared to our previous results [12], DUAE is more effective in TPC extraction than other extraction methods (indirect UAE, classical and Soxhlet extraction) which might be due to the strong cavitation effect causing an intensification of mass transfer [24].On the other hand, direct application of the ultrasound did not significantly affect the TFC in houseleek extracts.Different effects of direct sonication on flavonoid compounds can be explained by the fact that UAE efficiency is strongly related to the type of extracted compounds [47].Also, it was published earlier that direct ultrasound could lead to the degradation of some flavonoids, while flavonoid sensitivity depends on their chemical structure, number and type of substituents, as well the position of the hydroxyl group in the flavonoid molecules.Biesaga et al. [48] showed that a higher number of hydroxyl groups in flavonoid promoted the degradation of flavonoids, while sugar and methoxyl groups protected the flavonoid from degradation under sonication.

Antioxidant activity
The scavenging activity of the houseleek extracts obtained under different extraction conditions against stable DPPH radical are shown in Figure 1.Antioxidant activities of the extracts are compared using efficient concentration (EC 50 ) which could be defined as the concentration of the extract that causes reduction of DPPH concentration by 50%.For the interpretation of the results, higher EC 50 values indicate lower antioxidant activity.
Increasing the ultrasound power, as well as decreasing the ultrasound frequency, significantly affects the AOA of the extracts.It could be calculated that the AOA of houseleek leaves extract obtained under 75 W was four times higher than that of the AOA extract obtained under 20 W. The positive effects of ultrasound power on the antioxidant activity of the extracts were previously published for curry [42] and fresh Euphorbia tirucalli tree leaves [49].
Significantly lower AOA was found for extracts obtained at the ultrasound frequency of 1038 kHz than that obtained at 42 kHz, probably because of radical reactions and degradation of bioactive compounds as these processes were reported as the major effects of sonication at higher ultrasound frequencies (>500 kHz) [50].
Effects of different ultrasound power and frequency levels on AOA were consistent with the previously described effects on the TPC and TFC in houseleek extracts.The good correlation between EC 50 values and the TPC (R 2 = 0.999) and TFC (R 2 ˃ 0.784) also confirms that those bioactive compounds are mainly responsible for the antioxidant activity of the extracts.Finally, as can be seen from Table 2, the extract of fresh houseleek leaves with the highest TPC, TFC and AOA was obtained at the ultrasound power of 75 W and frequency of 211 kHz.So, taking into account the results in the present study and in order to maximize the extraction of antioxidant phenolic compounds from fresh houseleek leaves, the optimal DUAE conditions were suggested as follows: extraction time of 20 min, methanol as solvent, ultrasound power of 75 W and frequency of 211 kHz.The extract obtained under the proposed optimal conditions was further subjected for preliminary characterization.

HPLC/DAD analysis of extracts
The HPLC chromatogram of crude extract obtained under proposed optimal DUAE conditions is presented in Figure 2.
Most of the peaks in the chromatogram were not identified due to the lack of referent compounds, but seven flavonoids were identified including: genkwanin leaves juice.Kaempferol glycosides prevail, while quercetin glycosides are less characteristic.For unambiguous identification retention times, UV and mass spectra of kaempferol and quercetin were compared to those of a reference compound.However, flavonoid variation of the houseleek was studied only at the aglycone level and detailed data on glycosylation pattern of Sempervivum flavonols cannot be found in other sources [51].Similar phenolic profiles were detected for S. tectorum species with the presence of quercetin and kaem- pferol glycosides as the most abundant components [52] and scutellarein-7-rutinoside found in S. rhutenicum [53].The analysis of houseleek flower extracts shows the presence of carboxylic acids derivate as 3,4--dihydroxycinnamic acid and 1,4-dihydrocaffeic acid [54].The flavonoid aglycone composition analysis of some Sempervivum species after acidic hydrolysis showed that kaempferol was the principal flavonoid of all the species [55].

CONCLUSION
The results emphasize that higher ultrasound power and lower frequency of DUAE were better for the extraction of a bioactive compound from the plant material.In addition, the extract of fresh houseleek leaves with the highest content of TPC, TFC and AOA was obtained at the ultrasound power of 75 W and the frequency of 211 kHz.Hence, the extraction time of 20 min, methanol as the solvent, ultrasound power of 75 W and the frequency of 211 kHz were suggested as the optimal conditions for the DUAE of fresh houseleek leaves.
The results emphasize the possibility of successful application of DUAE in the extraction of bioactive compounds from fresh plant material.A great increase in the yield of TES was observed when DUAE was applied at a constant frequency and variable power.An increase in the frequency of ultrasonic waves at a constant power does not significantly affect the yield of total extractive substances.
Power and frequency of the ultrasound wave have a greater influence on antioxidant activity, and the total phenolic and flavonoid content in the extracts of houseleek.The good correlation between the total of phenolic and flavonoid content and antioxidant activities of houseleek extracts obtained at different frequencies and power of ultrasonic waves suggests that phenols are mainly responsible for the antioxidant activity of the extracts.
The most abundant components in the extract identified by HPLC-DAD are:

Figure 1 .
Figure 1.Antioxidant activity of the Sempervivum marmoreum L. extracts obtained at: a) the constant frequency of 211 kHz and different ultrasound powers; b) the constant ultrasound power of 20 W and different frequencies (methanol, 25 °C, 20 min.1:10 g/mL).Error bars are not shown when they are smaller than the symbols.

Table 1 .
Total extractive substances (TES) from fresh houseleek leaves during DUAE; data were expressed as the mean of three

Table 3 .
Compounds detected in Sempervivum marmoreum leaves methanolic extract obtained under optimal conditions