Phytochemical Profile and In Vitro Antioxidant Properties of Essential Oils from Powder Fractions of Eucalyptus camaldulensis Leaves

Dried leaves of Eucalyptus camaldulensis were finely grinded and fractionated by sieving into four granulometric classes (<100 µm, 100 - 200 µm, 200 - 355 µm and >355 µm). The obtained powder fractions were used for essential oil (EO) extraction by hydrodistillation and their phytochemical profile and in vitro antioxidant activities were evaluated. The mother powder (unsieved powder) was used for comparison. Particle size exerted a significant influence (p < 0.05) on the phytochemical composition and in vitro antioxidant properties of the EOs. Comparatively, the mother powder had the highest contents of α-pinene (55.6%), camphene (3.4%) and limonene (3.7%), while 1,8-cineole (26.6% and 22.4%), exo-fenchol (5.6% and 3.5%),


Introduction
In recent years, the interest for plant-derived bioactive products has increased owing to their numerous health and nutritional benefits. Plants and their products are rich sources of phytochemicals including terpenes and polyphenols having a wide variety of biological activities, such as anti-inflammatory, antimicrobial, and antioxidant activities [1] [2] [3]. With regard to antioxidant activity, it is well known that the antioxidants could attenuate the oxidative damage of a tissue indirectly by enhancing natural defenses of cell and/or directly by scavenging the free radical species. Because of that, growing attention has been paid to phytochemicals characterized as antioxidant natural products.
Eucalyptus camaldulensis is a medicinal plant with a wide spectrum of application in traditional medicine in several countries around the world. Numerous studies conducted on its essential oil (EO) have demonstrated its high antioxidant potential, mainly due to the great diversity of its phytochemicals. This is how Salem et al. [4] attributed the strong antioxidant activity of this EO to a high content of spathulenol. However, α-pinene has shown a strong antioxidant effect on the DPPH radical, the hydroxyl radical and the superoxide anion [5] [6] [7] [8]. Moreover, the high antioxidant effects of several terpenes like thymol, carvacrol, borneol, 1,8-cineole and α-terpineol in Eucalyptus camaldulensis leaves have been reported [9] [10] [11] [12]. Since EO phytochemicals can accumulate in various parts of plants, extraction represents a main process for concentrating these active compounds. Hence, it ensures separation of these compounds from plant cellular matrix and extraction conditions determine the quality and/or quantity of extracted compounds [13] [14]. In this regard, plant EOs are commonly obtained using hydrodistillation. In this process, water is used for extraction of plant active compounds as it has been recognized as safe solvent [15]. Unfortunately, this method usually requires long time with increased risk of degradation of thermo-labile constituents and results in low yields of extraction depending on the rigidity of plant material [16].
Grinding followed by particle size fractionation consists of producing powders of the plant and spraying it on a column of sieves with decreasing mesh diameters. Theoretically, the molecules are distributed according to the size of the particles, which makes it possible to concentrate on a fraction of a group of biomolecules [17] [18] [19]. Using this technique, many authors demonstrated that powder fractionation resulted in concentrating polyphenols in finer particle size [20] [21] [22]. Otherwise, our previous study reported that finer fractions (<180 µm and 180 -212 µm) of Dichrostachys glomerata, Adansonia digitata, Boscia senegalensis and Hibiscus sabdariffa concentrated more polyphenols, micronu-trients and antioxidant activities as ethanolic extract [23] [24] [25]. However, there is no study on the effect of powder fractionation on the EO phytochemicals in relation to their antioxidant activities. Yet, the application of this new approach which combines the advantages of sieve fractionation and hydrodistillation would allow efficient access to the phytochemicals of E. camaldulensis leaves. The high-antioxidant powders or their EOs could be useful as functional ingredients in nutraceutical formulations.
Therefore, this study aims to assess the effect of particle size fractionation by sieving on the concentration of phytochemicals and the antioxidant activities of the EO in powder fractions from E. camaldulensis leaves. To achieve this, the study focused firstly on assessing the mother powder granulometry and the EO yields from sieving fractionation. Secondly on comparing the phytochemical contents and the antioxidant activities (DPPH, ABTS, reducing power) of the EOs of different powder fractions. Mother powder was also analyzed and served as a control.

Plant Material
The fresh leaves of E. camaldulensis were used as plant material in the present study. This plant material was collected from Dang locality (Ngaoundere, Cameroon: latitude 7'42'46; longitude 13'55'59) in October 2017. Identification was made in comparison with the sample N˚4039 SRFT/am of the National Herbarium of Cameroon. After washing and cleaning, the leaves were dried in the shade of a hangar with air circulation at room temperature which varied from 14˚C to 38˚C. The leaves are spread on a clean sheet in thin layers and turned over frequently for seven days.

Plant Grinding and Particle Size Analysis
The dried leaves of E. camaldulensis were ground in the Moulinex robot blender mill supplied with a sieve drilled with 1 mm trapezoid holes. The analytical method used by Nguimbou et al. [26] was applied for the analysis of particle size distribution of the obtained powder. The measurement was carried out using laser by Mastersizer 3000 (Malvern Instruments, Orsay, France) supplied with the Aero S wet dispersion unit. The chosen size estimator was the particle size in volume and classical granulometric parameters were determined: D10, D50 and D90 (where Dx means that x% of the volume of particles has a diameter inferior to Dx). The span, a common parameter related to the width of particle size distribution was calculated as follows:

Powder Sieving
The plant powder was separated into granulometric classes using a series of three sieves of various apertures (100 µm, 200 µm and 355 µm) selected on the basis of the particle size analysis. The powder was sieved according to procedure used in previous studies [23] [27]. For that, 100 g of powder passed through sieve columns using an Analysette 3 Spartan apparatus (Fritsch, Idar-Oberstein, Germany) to obtain fraction powders. Sieve shaker vibration amplitude was set at 0.5 mm for 10 min. Thus, the fraction of the powder retained on each sieve is recovered and weighed for the calculation of the mass fraction of each granulometric fraction. The following powder fractions were obtained: <100 µm, 100 -200 µm, 200 -355 µm and >355 µm with a moisture content of about 10% -12%. Non fractioned (mother powder) powder was taken as control. The powdered samples were stored at 10˚C in polyethylene bags and placed at room temperature (25˚C ± 2˚C) until they were used.

Essential Oils Extraction
The EO were extracted from each powder fraction by hydrodistillation using an adapted device of the Clevenger type for 5 hours. The EO collected by decantation was filtered through a column of anhydrous sodium sulfate. The essences obtained were introduced into dark bottles and stored at 4˚C protected from light. The essential oil yield was expressed as volume of the distillate (V) per kg of dry leaves (Ms) and calculated as follows:

Chemical Composition Analysis
The analysis of the EO was carried out using a Varian CP-3380 type chromato-

DPPH Radical Scavenging Capacity
The DPPH anti-free radical activity of the EO samples was evaluated in accordance with the method of Zhang and Hamauzu [28] with slights modifications. A volume of 0.5 mL of the methanolic solution of each EO extract at a concentration of 2 mg/mL was added to 1 mL of the DPPH solution (0.025 g/L). In parallel, a negative control was prepared by mixing 0.5 mL of methanol with 1 mL of the methanolic solution of DPPH. The optical density recorded at 517 nm using the methanolic solution for the blank tube and ascorbic acid as standard after 60 min of incubation in the dark and at laboratory temperature (25˚C ± 2˚C). The percentage of free radical scavenging effect was calculated as follows:

ABTS Radical Scavenging Capacity
The ABTS anti-free radical activity of the EO samples was determined according to the method described by Re et al. [29]. ABTS radical was produced with the mixture of 7 mM ABTS and 2.45 mM potassium persulfate and incubated at room temperature in the dark for 16 h before use. After incubation, the absorbance of the solution at 734 nm was adjusted to 0.70 ± 0.02 by dilution with 95% ethanol. Then in a test tube, 2 mL of this diluted ABTS solution and 150 μL of each ethanolic solution of the different EO extracts at the concentration of 2 mg/mL were introduced and well stirred, incubated at laboratory temperature (25˚C ± 2˚C) for 10 min and the absorbance recorded at 734 nm. Ascorbic acid solution (1 mg/mL) was used as a standard. The percentage of free radical scavenging effect was calculated as follows:

Total Reducing Power
The total reducing power of the methanolic/water extracts was investigated using the method developed by Oyaizu [30]. One milliliter of the extract of each sample was mixed with 2.5 mL of phosphate buffer (200 mM, pH 6.6) and 2.5 mL of 1% potassium hexacyanoferrate [K3Fe (CN) 6] solution. The mixture was incubated in a water bath at 50˚C for 30 min, cooled, mixed with 2.5 mL of 10% trichloroacetic acid solution and centrifuged at 3000 rpm for 10 min. The supernatant (2.5 mL) was removed and mixed with 2.5 mL of distilled water and 0.5 mL FeCl 3 (1.0%), allowed to react for 10 min at room temperature and the absorbance was measured at 700 nm. Ascorbic acid solution was used as standard. The reducing power is determined by referring to the calibration curve of ascorbic acid and expressed in terms of mg ascorbic acid equivalent per g of dry matter (mg AAE/g DM).

Statistical Analysis
Analyses were performed in triplicate and the results presented as mean ± SD.
Statistical significance of differences between sample means was determined using analysis of variance (ANOVA) followed by a Duncan's test at 95% confidence level using statgraphics 15.1. Principal component analysis (PCA) has been used to classify and correlate the factors that define powder fractions, phytochemical compounds and antioxidant activities and was realized by XL-STAT 2019. Table 1 presents the particle size characteristics of the mother powder of E. camaldulensis leaves studied. Particle size distribution of the mother powder of E. camaldulensis leaves was polymodal with a volume of fine particles around 34 µm (D10), a median volume of particles around 195 µm (D50) and a volume of large particles around 735 µm (D90). The major volume of E. camaldulensis mother powder corresponded to large particles, whereas fine particles were less numerous (10% of a total volume of the powder). The low span value of 2.3 reflects a low dispersion of the mother powder and confirmed the correct running of the grinding process. Indeed, a dispersion is said to be wide if its span is greater than 3 [22].

Particle Size Characteristics of the Mother Powder
The fact that the major volume of E. camaldulensis powder corresponded to large particles can be explained by the expected non-sphericity of large particles, as ground fibers generally lead to large rod-shaped particles. So, fine particles can stick to larger particles and thus be retained by sieves of mesh size exceeding their diameter [27]. As the median distribution of the mother powder being located around 200 μm, the fractionation sieves were chosen as follows: a 200 μm sieve for studying the particles around the median; a 100 µm sieve for studying particles below the median and a 355 µm sieve for studying particles above the median.

Sieved Mass, Mass Fraction and EO Yield
The sieving fractionation was carried out on E. camaldulensis mother powder in order to check if the employed sieving procedure was efficient in producing powder fractions having well-different particle size (Table 2). It can be seen in Table 2  Based on these results, one would expect differences on EO extraction yields in these powder fractions.
V. D. Noumi et al. As shown in Table 2, the mother powder of E. camaldulensis has the best EO yield (1.6 mL/kg) compared to the powder fractions. These values are much lower than those obtained by El Hajji et al. [31] whose EO yields were comprised between 15 and 16 mL/kg for dried E. camaldulensis leaves. Observed differences in yield could be explained by the fact that after drying, the leaves were ground into a fine powder and then sieved, while our predecessors did so on dried and whole leaves. From this point of view, it could be admitted that, during the grinding and sieving processes of the leaves, there is loss of volatile compounds of low molecular weight. Indeed, EO are highly volatile liquid compounds synthesized in specialized histological structures often located on or near the plant structure [32]. The work of Cheftel et al. [33] demonstrated that the yields of EO are markedly reduced depending on the types of drying: in the shade, activated by ventilation, assisted by lamps or using a resistance. Even more, Akhihiéro et al. [34] demonstrated that there is progressive decrease of the EO yield from 8.4 to 5.1 mL/kg as the particle size of Cymbopogon citratus leaves decreases from 20 mm to 4 mm. Cryogenic grinding makes it possible to preserve sensitive active ingredients such as EO and vitamins [35], which has not been the case in this work. The leaves were crushed using an electric mill without a temperature stabilization device (thermogenic) which would further explain the great loss of volatile compounds. However, sieving fractionation allowed to concentrate more EO in the 100 -200 µm fraction (0.80 mL/kg), which has the highest yield compared to the rest of the fractions: 200 -355 µm (0.68 mL/kg), <100 µm (0.40 mL/kg) and >355 µm (0.32 mL/kg). This result could be explained by the fact that in Myrtaceae, EO are trapped in the secretory pockets of specialized cells [36]. From this point of view, we can admit that 100 -200 µm fraction concentrates the maximum of secretory cells which remained unex-ploded after the grinding operation; these EO were released later under the constraints of temperature and pressure during hydrodistillation.  [41]. Even more, the results show a higher content of L-borneol (12.6%), L-trans-pinocarveol (5.5%), α-campholenol (4.2%) and exo-fenchol (5.6%) in the <100 µm fraction. These values are higher than those of the other fractions and that of the mother powder. These results could be explained by the fact that terpenes react with oxygen (in the air) to form new compounds through the auto-oxidation mechanism [42]. In fact, in aerobiosis, oxygen promotes the oxygenation and hydroxylation reactions of the double carbon bond favoring the  transformation of hydrocarbon terpenes into oxygenated terpenes [43]. In this regard, we can admit that the reduction in the size of the particles by the grinding/sieving process, increases the contact surface with atmospheric oxygen. As a result, the EO self-oxidation process is more important in small particles compared to large particles. This justifies the reduction in the level of hydrocarbon terpenes, in particular α-pinene and limonene in the <100 µm and 100 -200 µm fractions.

Chemical Composition of the EO Samples
According to Neuenschwander et al. [44], the structure of α-pinene has four possible oxidation sites and its auto-oxidation depends on three factors: the ambient temperature, the oxygen flow and the exposure time. So, the auto-oxidation of α-pinene can lead to the formation of L-trans-pinocarveol [45].
Furthermore, the work of Rio [46] demonstrated that, atmospheric oxidants (hydroxyl radical OH, Nitrate NO 3 , Ozone O 3 ) react with terpenes promoting the self-oxidation process. This is how limonene could react with the OH radical to form α-terpineol, as well as the oxidation of α-pinene in air which leads to the formation of derivative compounds [47]. However, the degradability of aerobic terpenes varies depending on the type of molecule and depends on the possible presence of a group. In this respect, Wilson and Hrutfiord [48] reported different rates of degradation mainly associated with molecular type. Alcohols are degraded to 99%, hydrocarbons to 75% and ketones to 12%. This aerobic metabolism can consist of several chemical reactions such as the hydroxylation of allylic carbons, the oxygenation of carbon-carbon double bonds, or the oxidation of alcohols to ketones as well as the breakdown of certain cyclic structures.
Otherwise, the transformation of limonene into α-terpineol by Penicillium digitarum is initiated by the epoxidation of the double bond in position 8, 9 and followed by the reductive rupture which forms α-terpineol [49].

Antioxidant Activity of EO Samples
The antioxidant activity of EO samples from E. camaldulensis leaves powders was measured in terms of radical scavenging ability and total reducing power. In Table 4 is reported the antioxidant activity of the EO assessed by different tests.
The DPPH assay is known to provide reliable information concerning the antioxidant capacity of specific compounds or extracts across a short time scale.
From the DPPH assay, the maximal antioxidant activity of EO was recorded for the <100 µm fraction (90.62% ± 0.13%) followed by the 100 -200 µm fraction (73.31% ± 3.13%) whereas, the EO of mother powder had the lowest DPPH free radical scavenging activity (70.37% ± 21%). Compared to ascorbic acid (96.29% ± 0.28%) used as antioxidant reference, the fractions of E. camaldulensis powder were less active against the DPPH and ABTS radicals. Similarly, Barra et al. [50] demonstrated significant anti-radical activity with DPPH in EO of E. camaldulensis leaves collected in different localities of Sardinia in Italy. This author attributes the variation in anti-radical potential to the difference in phytochemical compounds in the EO. This suggests that, the particle size fractionation af-  (Table 4). Thus, the highest DPPH radical scavenging activity of the EO in the <100 µm fraction would be due to its high concentration of oxygenated terpenes [51]. Earlier studies have indicated that oxygenated terpenes such as thymol, carvacol, borneol, 1,8-cineole, α-terpineol, have a significant antioxidant effect, mainly due to their redox properties [9] [11] [12]. Thus, the high contents in phytochemical compounds of EO of finer powder fractions are likely responsible for their DPPH scavenging capacity. The antioxidant activity of EO samples was also measured in terms of radical scavenging ability using the ABTS assay. Table 4 displays the ABTS radical scavenging activity of the EO in the different granulometric fractions and mother powder of E. camaldulensis leaves. Significant differences (p < 0.05) were denoted between the antioxidant activity of the different fractions and unsieved powder. As for the DPPH test, the maximal antioxidant activity of EO was recorded for the <100 µm fraction (89.59% ± 0.09%) followed by the 100 -200 µm fraction (70.46% ± 2.83%). Compared to the mother powder, the 100 -200 µm and <100 µm fractions were 1.5 times and 2 times more active against the ABTS radicals, respectively. For most of the samples tested, the effect obtained with the DPPH method is not correlated with that obtained by the ABTS method; the results differ by these two methods. These variations can be explained by the mechanisms involved in the antioxidant reactions of radicals. The EO samples were dissolved in ethanol for the ABTS test, while methanol was used for DPPH test. This suggests that, solubility of the EO in different solvents can influence their effect. Indeed, the antioxidant activity of the tested compounds depends on the agent used and insisting on the mechanism of action of the antioxidant [52]. Other factors, such as the stereo-selectivity of the radicals or the solubility of the compounds in the different test systems can also influence the capacity of each EO to react and to reduce the different radicals [53]. Antioxidant activity deduced by the ABTS experiment allowed sorting powders according to their antioxidant activity in the following decreasing order: <100 µm, 100 -200 µm, 200 -355 µm, >355 µm and the mother powder. The reducing power of antioxidants is an important indicator of potential antioxidant activity. The antioxidant effect increases as a function of the reducing power, indicating that the antioxidant properties are concomitant with the development of reducing power [30]. In Table 4  tion-based assays [54]. In fact, in the reducing power assay, the presence of reductants (antioxidants) terpenoids causes the reduction of the Fe 3+ or ferricyanide complex to its ferrous form [55]. As stated by Malecky [36], oxygen that has a number of oxidation (II) easily gives electrons during oxido-reduction reactions, while molecular oxygen is the final electron acceptor, important in reduction reactions. Thus, terpenes which have a low number of functional groups (hydroxyl and carbonyl) have a reduced activity in the reactions [55]. However, the antioxidant activity of the majority compounds tested separately gives lower results compared to the activity of the total EO [56]. This suggests that, finer powder fractions richer in phytochemicals concentrate the various compounds of the EO, which act synergistically for much greater antioxidant activity [57].

Principal Component Analysis and Correlations
Principal component analysis (PCA) is one of the most widely used methods for multivariate analysis. It was performed to present an overview of the similarities and differences between the EO from powder fractions and to highlight the correlations between the terpene compounds and the antioxidant activities ( Figure   2). This analysis makes it possible to identify more clearly the fractions having the EO with different contents of phytochemical compounds and therefore dif-  In this regard, it was observed a strong correlation between the exo-fenchol variable and the anti-free radical test with DPPH (r = 0.82, p < 0.05) as well as the total reducing power (r = 0.85, p < 0.05). Concerning the anti-free radical test with ABTS, the strongest correlation (r = 0.87, p < 0.05) was observed with 1,8-cineole. Thus, exo-fenchol and 1,8-cineole appear to be the terpenes with higher antioxidant potential. Pearson's correlation between terpenes and the different antioxidant tests makes it possible to note that the different variables, namely 1,8-cineole, exo-fenchol, L-trans-pinocarveol, L-borneol and α-terpineol are positively correlated, in particular with regard to the inhibition of DPPH, ABTS radicals and reducing power. These strong correlations (r > 0.5) would be due to the fact that the presence of the oxygen molecule as well as the carbon-carbon double bond in the structure of terpenes makes them more willing to redox reactions which can therefore reduce free radicals in solutions [54].

Conclusion
This study permitted to investigate the effects of size reduction technology on the phytochemical compounds, and antioxidant properties of EO from E. camaldulensis leaves. EO of E. camaldulensis leaves is mainly composed of α-pinene, 1,8-cineole and α-terpineol. The particle size fractionation by sieving made it possible to significantly concentrate the α-terpineol, L-borneol, L-trans-pinocarveol and exo-fenchol in the EO of the <100 µm and 100 -200 µm fractions of the leaves of E. camaldulensis. However, it is important to control grinding speed, particle size and hydrodistillation conditions to avoid phyto-chemicals degradation by heat and oxidation. The distribution of molecules as a function of particle size revealed the EO of the <100 µm and 100 -200 µm fractions as the most effective in reducing the concentration of DPPH and ABTS in solution. Ultimately, the results of this work show that particle size fractionation by sieving concentrates the phytochemical compounds in finer fractions of the powders of E. camaldulensis leaves, while enhancing the antioxidant potential of their EOs. Overall, these finer powder fractions offer high potential in the management of oxidative reactions associated with human and plant pathologists.