1. Introduction
The autumn brings the finest displays, with various yellow, gold, orange, coppery red, crimson, cream and amber exfoliates on thin leaves plates to reveal the autumnal and winter leaf colors. For several weeks in late autumn, every
Parrotia persica (DC) C. A. Mey., Hamamelidaceae leaf turns to a different color, including red, crimson, orange and yellow. In winter, the
P. persica leaves turn to cream and amber.
P. persica grows naturally on the Caucasus and in Caspian forests [
1]. Aerial parts of
P. persica (leaves and stems) were extracted with methanol, the chemical constituents present in the obtained extract were purified by column chromatography and the flavonoids were determined by different spectroscopic techniques. [
2].
P. persica leaves were extracted with various organic solvents and the extracts obtained were evaluated for antibacterial activity [
3].
In autumn, with the beginning of leaf senescence, chlorophyll molecules rupture. The hydrolysis of chlorophyll reveals the phytol and chlorophyllide molecule [
4]. One pathway for tocopherol biosynthesis uses phytol derived from chlorophyll [
4,
5]. Tocopherols are compounds known for their antioxidant activity [
6]. The tocopherols are present in soybean and wheat germ oils [
7]. The methods for obtaining tocopherols include extraction from natural sources or chemical synthesis [
8]. On the market, the mixture of tocopherols is available in vegetable oils [
8]. Tocopherols are present in different vegetable oils [
9].
Supercritical carbon dioxide (SC CO
2) extraction has higher extraction rates, efficiencies and selectivity compared to extractions using solvents [
10]. SC CO
2 extraction has been used for the recovery of tocopherols from soybean oil deodorizer and tocopherols were concentrated from 11% to 13% [
11]. The operational conditions were from 40 °C to 80 °C and from 90 to 170 bar [
11]. It was concluded that the yield of the process reduces with the increase in pressure at constant temperature [
11]. Various SC CO
2 extractions were carried out or simulated for canola, lampante olive oil, sunflower oil deodorizer distillate and esterified olive oil deodorizer distillates [
12,
13]. The recovery of tocopherols from the deodorized distillate of the vegetable oil using SC CO
2 has been reported [
14,
15]. The effective separation of linoleic acid, stigmasterol and squalene from tocopherols was investigated [
14]. At 40 °C and 90 bars, the separation between tocopherols and linoleic acid had an efficiency of 93.1% and the concentration factor showed that the tocopherols were completely separated from the fatty acids [
14]. The optimal conditions for separation between tocopherols and squalene was at 40 °C and 350 bar [
14]. The separation between tocopherols and stigmasterol was at 40 °C and 250 bars, with the efficiency of 99.0% and concentration factor of 103.6 [
14]. At the same operating conditions, tocopherols remained with squalene because the separation was not efficient at low pressures [
14]. The recovery of tocopherols, contained in the esterified soybean sludge, can be up to a maximum of 40 wt% [
15]. SC CO
2 has been used for the extraction of tocopherol concentrates from olive tree leaves [
16]. The tocopherol extraction rates were determined in the function of pressure (25–45 MPa), particle size (0.25–1.5 mm), solvent flow (0.5–1.5 SL·min
−1) and temperature (40–60 °C) [
16]. At 40 °C, pressure of 25 MPa, solvent flow rate of 1 SL·min
−1 and particle size of 1.5 mm, the tocopherol concentration in the extract was 74.5% [
16]. The Soxlet extraction using hexane showed no tocopherol content, which was most probably due to thermal and oxidative degradation [
16]. Tocopherol extraction from the yellow and amber
P. persica leaves has not been performed, up to now. The main goals of the present study were as follows: to find the lipophilic phytochemical profile of yellow and amber
P. persica leaf extracts obtained by SC CO
2, to investigate the influence of process parameters on SC CO
2 extraction and to find the optimal extraction conditions related to the extraction yield and α-tocopherol amount.
3. Discussion
The phytochemical compounds present in the yellow and amber leaves were as follows: neophytadiene, hexahydrofarnesyl acetone, octadecanal, 1-octadecanol, phytol, squalene and α-tocopherol (
Table 2 and
Table 3). The yellow leaves contained α-cadinol, while in the amber leaves, it was not detected. The β-sitosterol was present in amber leaves.
The effect of pressure, temperature and CO
2 flow rate was investigated using the BBD with three levels for each factor. The BBD consisted of fifteen experiments (
Table 1). The average particle size, the mass of the plant material in the extractor and the extraction time were kept constant during all experiments. The extraction time was kept constant, due to the results obtained where extractions longer than 90 min did not significantly increase the extraction yield. The extraction yield varied from 0.94% (30 MPa, 50 °C and CO
2 flow rate 1 kg·h
−1) to 1.55% (30 MPa, 50 °C and CO
2 flow rate 3 kg·h
−1) for the yellow
P. persica leaves (
Table 1). For the amber
P. persica leaves, the extraction yield was 0.82% (30 MPa, 60 °C and CO
2 flow rate 2 kg·h
−1), while the highest extraction yield was 1.39% (20 MPa, 40 °C and CO
2 flow rate 3 kg·h
−1) (
Table 1). The interaction coefficient pressure and CO
2 flow rate for yellow leaves and pressure and temperature for amber leaves had a significant influence on the extraction yield. The quadratic interaction coefficients for CO
2 flow rate showed the statistically significant influence on the extraction yield from the yellow leaves. The second order polynomial models were used to express the total extraction yield in the function of independent variables and are expressed by Equation (1) for the yellow leaves and Equation (2) for amber leaves. From Equation (1), the quadratic terms of pressure (
X12) and temperature (
X22) were statistically insignificant and this reflection can be concluded from the values of the regression coefficients. Interactions between pressure and temperature (
X1·
X2) and interactions between temperature and CO
2 flow rate (
X2·
X3) can be removed from the model because they have no significant effect on the extraction yield. In case of Equation (2), the two-way and quadratic interactions have no significant effect on the extraction yield. The RSM analysis confirmed that the temperature and CO
2 flow rate are highly statistically significant model parameters that influence the process.
The calculated optimal conditions for obtaining the maximum extraction yield were, for both plant materials, 30 MPa, 40 °C and flow rate of 3 kg·h
−1. The higher selectivity of the total extraction yield was obtained at lower temperatures and higher flow rates. On the selected surface plots, it can be observed that the total extraction yield increases with the increase in the CO
2 flow rate (
Figure 1). With the increase in pressure and CO
2 flow rate, the total extraction yield increases in both plant materials analyzed. The total extraction yield is the highest at lower temperatures.
Three extractions under optimal conditions were carried out to validate the developed mathematical model for the total extraction yield. The obtained experimental data are in the range of the predicted values and are within the limits set by the relevant confidence intervals with the deviation of ±5%. This confirms the acceptable accuracy of the assumed mathematical models and reliability of the BBD.
The highest relative amount of α-tocopherol in yellow leaves was 81.67 mg per 100 g (20 MPa, 40 °C and CO
2 flow rate 3 kg·h
−1), while the lowest relative amount was 22.65 mg per 100 g (30 MPa, 60 °C and CO
2 flow rate 2 kg·h
−1) (
Table 2). In amber leaves, the highest α-tocopherol relative amount was 321.68 mg per 100 g (20 MPa, 40 °C and CO
2 flow rate 3 kg·h
−1), while the lowest relative amount found was 90.27 mg per 100 g (30 MPa, 60 °C and CO
2 flow rate 2 kg·h
−1) (
Table 3).
In the case of the α-tocopherol relative amount, the response variables were positively correlated with pressure (
X1) and temperature (
X2) and the quadratic terms of pressure (
X12) and quadratic terms of temperature (
X22) were regarded as statistically significant. The linear and quadratic terms of CO
2 flow rate (
X3) and (
X12) turned out to be statistically insignificant (
p > 0.05). All the two-way interactions were statistically insignificant. It can be concluded from the statistical analysis of regression coefficients that the α-tocopherol relative amount depends on the pressure (
X1) and temperature (
X3) for both plant material analyzed (
Table 5).
The effects of process parameters on the studied response variables, the α-tocopherol relative amount, were drawn in the form of response surface figures (
Figure 2). The total α-tocopherol relative amount increases when increasing the pressure to 16.54 MPa for yellow leaves and to 15.85 MPa for amber leaves and then it decreases. The highest α-tocopherol relative amount was obtained at 40 °C. At the CO
2 flow rate of 2.65 kg·h
−1 for yellow leaves and 2.77 kg·h
−1 for amber leaves, the highest α-tocopherol relative amount was obtained. A further increase in the CO
2 flow rate led to the decrease in α-tocopherol relative amount.
Three extractions under optimal conditions were carried out to validate the developed mathematical model for the α-tocopherol relative amount. The obtained experimental data were as follows: 80.03 mg of α-tocopherol per 100 g of the dry plant material for the yellow leaves and 315.3 mg of α-tocopherol per 100 g of the dry plant material for the amber leaves. The results obtained were in the range of the predicted values and relevant confidence intervals with the deviation of ±5%. This permitted the acceptable accuracy of the assumed mathematical models.
The parameters that influenced the α-tocopherol extraction yield from the various leaves investigated were not uniform. The highest α-tocopherol extraction yield, in the case of olive leaves, was 10.10 mg per 100 g of leaves at 25 MPa, 40 °C, CO
2 flow rate of 1 SL·min
−1, particle diameter 1.5 mm and extraction time of 120 min [
16]. Under the same parameters, only the extraction time was 60 min, the α-tocopherol extraction yield was 6.94 mg per 100 g of leaves [
16]. In case of
Eugenia involucrata, the highest α-tocopherol extraction yield was 68.27 mg per 100 g of leaves when the pressure was 20 MPa, temperature was 60 °C and the CO
2 flow rate was 4 mL·min
−1 [
20]. The α-tocopherol extraction yield in the case of
Pandanus odorus leaves increased with the increase in pressure and decreased when increasing the temperature [
21]. At a pressure of 200 kg·cm
−2 and at 40 °C, the yield was ~300 ppm after 3 h of extraction, while at the same temperature, but at pressure of 80 kg·cm
−2, the α-tocopherol extraction yield was 134 ppm [
21]. In
Vitis vinifera leaves, α-tocopherol was only detected when the extraction parameters were as follows: pressure of 30 MPa, temperature of 40 °C, particle diameter of 10 mm and CO
2 flow rate of 80 kg·m
−3 [
22]. The
P. persica yellow and amber leaves had the highest α-tocopherol extraction yield at the pressure of 20 MPa, temperature of 40 °C and CO
2 flow rate of 3 kg·h
−1. In the case of
E. involucrate, the temperature was the highest and obtained the highest α-tocopherol extraction yield compared to olive,
P. odorus,
V. vinifera and
P. persica leaves. Yellow
P.
persica leaves have a higher α-tocopherol extraction yield compared to
E. involucrate. The highest α-tocopherol extraction yield was demonstrated by amber
P. persica leaves.