All chemicals and reagents were of analytical reagent grade. Potassium hydroxide, Hydrochloric acid and Nitric acid were purchased from Panreac (Spain). Acetone, methanol, ether and hexane were supplied from Carlo Erba (France).

Olive leaves were harvested from the Olive Tree Institute farm of Sfax, Tunisia (34°43N, 10°41E) in April 2011. Nineteen-year-old olive trees (Olea europaea L.) were used. Trees were spaced 4 m×6 m and subjected to the same olive cultivation practices. The sandy soil had an organic matter content of 1.3%; 12.3% CaCO₃; 1.2% N and a pH of 7.8. Four varieties were selected for this study: Chemlali (CL), Chemchali (CH), Chetoui (CT) and Zarrazi (ZR). Samples of olive leaves (120 g), characterized as being from the present year (aged 3 months), were collected from each variety. All foreign matter was removed from the material, especially adhering soil or sand. Since the moisture contents of the leaves were about 50%, the samples were oven dried at 37 °C for 48 h. Olive leaves from each variety were separately milled in a blender to pass a 1 mm screen and stored in light protected glass bottles for further use.

Samples of each olive leaf powder (1 g) were ashed at 550 °C in a muffle furnace for 4 hours. The ashes were then cooled in a desiccator and used for the determination of calcium, potassium, magnesium, sodium, iron, zinc and copper contents (AOAC, 1984). Therefore, a known amount of ash was dissolved in 4 mL of concentrated HNO₃ in a porcelain crucible. The solution was stirred and evaporated to dryness on a hot plate set at 120 °C. After that, the crucible was returned to the furnace at 550 °C for 1 hour. The ash was dissolved in 10 mL of concentrated HCL and then filtered into a 50 mL volumetric flask. The ash residue was rinsed three times and made up to volume with deionized water. The mineral constituents present in the olive leaves were analyzed separately using an atomic absorption spectrophotometer (Hitachi Z6100, Japan). A calibration curve for each of the mineral to be determined was prepared using the standard solutions before the readings were obtained.

Total chlorophyll, chlorophyll a and b contents were determined by the spectrophotometric method according to AOAC (1984). Briefly, fresh olive leaf samples (5g) were ground into a Waring Blendor and then 25 mL of 85% acetone solution were added. The samples were homogenized for 3 min. The homogenate was filtered through filter paper Whatman no. 1, with a Büchner funnel under vacuum. The filter cake residues were then washed with an 85% acetone solution. The extraction procedure was repeated until the tissue was devoid of any green and washings were colorless, and then the filtrate was brought to a final volume of 100 mL. An aliquot of 25 mL of acetone extract was transferred into a separator containing 50 mL of ether. After that, distilled water was added until it was apparent that all fat solution pigments had entered the ether layer. Washings of ether solutions were continued with distilled water until all the acetone was removed. Then the ether solution was transferred to a 100 mL volumetric flask, diluted to volume and mixed. The absorbance readings of this solution were recorded at wavelength 660 and 642.5 nm using a UV/Vis spectrophotometer (Shimadzu, Japan). The concentrations of total chlorophyll, chlorophylls a and b (ppm) were calculated as: Total chlorophyll = 7.12 A 660 + 16.8 A 642.5 Chlorophyll a = 9.93 A 660 - 0.777 A 642.5 Chlorophyll b = 17.6 A 642.5 - 2.81 A 660

The olive leaf powder (50 g) was placed in a dark flask and homogenized with 250 mL of hexane (1:5, w/v). After mixing for 4 h in a shaker (Selecta, Spain), the mixture was centrifuged for 15 min at 1000 g. The supernatant was then filtered through a filter paper (Whatman no. 2). The extraction procedure was repeated twice. The solvent was removed by a vacuum rotary evaporation at 40 °C. The extract was pooled, drained under a stream of nitrogen and then stored in a freezer until analysis.

The hexane extracts of olive leaves were filtered with a millipore membrane filter (0.45 μm) prior to gas liquid chromatography analysis. Samples (1 mg) were converted into their corresponding fatty acid methyl esters using a methanolic solution of potassium hydroxide (2 M). The mixture was maintained at 100 °C for 1 h. The reaction was stopped with 0.5 mL of distilled water. Then, the extracted fatty acid methyl esters were dissolved in heptane. GC analyses were performed on a GC-17-A SHIMADZU (Japan), equipped with a hydrogen flame ionization detector and a capillary column: Carbowax (15 m, 0.25 mm).The column temperature was fixed at 180 °C and the injector and detector temperatures were set at 230 °C and 250 °C, respectively. Nitrogen was the carrier gas. Fatty acid methyl esters were identified by comparison of their retention times with respect to pure standards purchased from Sigma and analyzed under the same conditions. Fatty acid methyl esters were quantified according to their percentage area, obtained by the integration of the peaks. The results were expressed as a percentage of individual fatty acids in the extract.

The thermal properties of olive leaf extracts in hexane were determined using a differential scanning calorimeter (NETZSCH-Gerätebau GmbH Thermal analysis DSC 204, Germany). The hexane extract (8±0.1 mg) was weighed in a DSC-aluminium pan. An empty DSC-pan was used as an inert reference. The sample and the reference pans were then placed inside the calorimeter and were quickly cooled to −60 °C at a speed of 5 °C·min⁻¹, held at this temperature for 15 min, and heated to 100 °C with a heating speed of 5 °C·min⁻¹ and the DSC thermographs were recorded during the melting transition. The peak temperature is the temperature maximum of a thermal transition. The onset temperature is the temperature where the extrapolated leading edge of the endotherm intersects with the baseline.

Analyses were carried out in triplicate. The values of different parameters were expressed as the mean ± standard deviation. The results were statistically analyzed by one-way analysis of variance (ANOVA) and Tukey tests using SPSS (Version 11, SPSS Inc., Chicago, USA). Statistical significance was accepted at a level of p<0.05.

The mineral compositions of the olive leaf varieties namely Chemlali (CL), Chemchali (CH), Chetoui (CT) and Zarrazi (ZR) are shown in Table 1. It can be seen that the ash contents of olive leaves ranged from 6.60 to 9.82%. The values were in the same range of those reported by Delgado-Pertinez et al., (2000) for olive leaves (6.5 to 9.6%). The ash amounts of olive leaves are considered high in comparison with other vegetal products such as pumpkin and watermelon seed kernels (3.21% and 3.6%, respectively) (El-Adawy and Taha 2001).

The mineral composition of olive leaves showed that calcium [9.25 mg·g⁻¹ d.m. (CT)−10.39 mg·g⁻¹ d.m. (CL)] was the predominant mineral. Similar values (9.296 mg·g⁻¹ d.m.) were reported by Lee et al., (2005). Calcium develops and maintains strong bones and teeth and enhances the use of other nutrients. Therefore, calcium supplements may be used to prevent and to treat calcium deficiencies.

Potassium amounts in the olive leaves ranged from 4.47 to 9.14 mg·g⁻¹ d.m. Potassium is a micronutrient that plays an important role in the regulation of the heartbeat. Besides, it maintains fluid balances and helps muscles contract. An adequate intake of calcium and potassium contributes to preventing cardiovascular diseases (McCarron and Reusser, 2001). Thus, the presence of calcium and potassium in olive leaf powder extracts could enhance the beneficial effects on health of other phytochemicals such as phenolic compounds and tocopherols.

The potassium concentrations of CL and CT olive leaf varieties were significantly (p<0.05) different from those of CH and ZR. The mineral contents of olive leaves could be influenced by the differences in the nutrient status of the soils in which the olive tree grows and by the environmental effects around the tree. Higueras et al., (2012) studied the uptake of metallic trace elements from soils for olive-tree growing in the Almadén mercury mining district. The authors found good correlations between soil and olive leaf contents for major elements from the soils (Fe, Al, Mg and Ca), indicating a direct and important uptake of these elements from the soil. However, some element contents in leaves do not show any relationship with the element contents in the soil (Ce, Cr, Hg, Nd, Pb, Zn). These variations have been explained by the different availability of the elements in the soil. Several methods for estimating nutrient availability and their limitations were reviewed by Marschner and Rengel (2012).

The effects of soil chemistry on plants were also studied by Lynch et al., (2012). In fact, the soil nutrient availability, pH, aeration and low molecular weight organic solutes were among the soil chemical factors affecting the plant root growth and development.

The magnesium contents varied from 1.5 to 3 mg·g⁻¹ in olive leaves. The magnesium average value measured for Spanish olive leaves was 1.5 mg·g⁻¹ (Higueras et al., 2012). Sodium was present in relatively low concentrations in the olive leaves.

Chlorophylls are very common pigments, which give color to plants and several algae. The color of olive leaves is mainly related to their chlorophyll content, as this compound is the main pigment of green vegetables and masks the bright color of carotenoids.

The total chlorophyll, chlorophyll a and chlorophyll b contents of olive leaves are presented in Figure 1. The total chlorophy1l contents of olive leaves ranged from 1132.33 to 1795.93 ppm. These values were in the same range obtained for watercress leaves (1400 ppm) (Gonçalves et al., 2009). The chlorophyll content of olive leaves seems to vary with variety. Indeed, the amounts of chlorophyll a were significantly (p<0.05) different among varieties [281.69–854.59 ppm] being higher for the CL variety which also had the highest contents of total chlorophyll. The chlorophyll b content of olive leaves ranged from 851.83 to 1114.23 ppm. The influence of genetic factors on chlorophyll content was confirmed by Bojovic and Markovic (2009) who investigated the variability of the chlorophyll content of wheat leaf with cultivar. Furthermore, chlorophyll content is usually affected by various environmental factors such as temperature, photoperiod, nitrogen, water and degree of maturity (Chen et al., 2011; Lee et al., 2011).

Chlorophylls are used as additives to food products due to their color and physico-chemical properties. In fact, chlorophyll molecules are extracted and used as natural pigments in processed foods (Gutiérrez-Rosales et al., 1992). They can be chemically modified before being incorporated into food products (e.g. replacement of Mg²⁺ by Cu²⁺ in chlorophyll). Further, chlorophyll precursors and chlorophyll derivatives are used in medicine for photodynamic treatments (Schoefs, 2002).

The hexane extract yields of olive leaves are shown in Table 2. The percentages varied from 3.11 to 3.88 %. The extraction yields of the olive leaf varieties were statistically similar (p>0.05). This can be explained by the fact that the olive trees were grown in the same region with similar rainfall. Additionally, the olive leaves were harvested during the same season. Martin-Garcia and Molina-Alcaide (2008) found higher yields (7.81–9.76%) for Spanish olive leaves using the soxhlet technique for the extraction. This difference can be attributed to the effect of heat on fat extraction and to the progressive depletion by the successive and multiple washing cycles by the soxhlet extraction. Based on the fact that the temperature of extraction can affect the initial quality of the fat, cold extraction was selected in this study.

The fatty acid composition of olive leaf extracts in hexane is shown in Table 2. The major fatty acids found in the olive leaf samples were linolenic acid (C_18:3) [30.02–42.16%], oleic acid (C_18:1) [18.28–26.36%] and palmitic acid (C_16:0) [18.22–22.42%]. The fatty acid composition of olive leaves seems to vary slightly with variety. For instance, the fatty acid composition of the CL variety is statistically similar to that of ZR, except for the amounts of linoleic acid.

Linolenic acid is a polyunsaturated omega-3 fatty acid which is metabolized to eicosapentaenoic acid, a precursor of eicosanoids with anti-inflammatory and antithrombotic activity (Ruiz et al., 2002). Many researchers have found evidence that linolenic acid is related to a lower risk of cardiovascular diseases (William, 2000; Simopoulos, 2002; Schacky and Harris 2007).

The percentage of polyunsaturated fatty acids [44.50–58.29%] in olive leaves was higher than that of saturated fatty acids [23.30–28.99%]. The olive leaves exhibited higher levels of saturated fatty acids than olive oil (21.22%) (chemlali variety) (Dabbou et al., 2010). Besides, the olive leaves showed higher levels of polyunsaturated fatty acids in comparison with olive oil (18.11%), which contains about 17.48% of linoleic acid and 0.63% of linolenic acid. The polyinsaturated fatty acids could make olive leaves specifically prone to oxidation. However, this could be counterbalanced by the levels of antioxidants that protect the olive leaf extract.

The polyunsaturated fatty acids including the omega-3 and omega-6 families detected in the plants constitute an important class of phytochemicals due to their generalized beneficial health effects (Guimarães et al., 2009).

Linoleic acid is observed at appreciable percentages in olive leaf extracts [14.48–16.52%]. This essential fatty acid has become increasingly popular because of its beneficial properties for the skin, including anti-inflammatory, acne reduction and moisture retention properties (Darmstadt et al., 2002). Therefore, it is often used in making soaps, creams and emulsifiers. Thus, the olive leaf extracts containing polyunsaturated fatty acids constitute a potential source of fatty acids which can be used to manufacture dietary supplements.

In addition, the use of olive leaves in animal feeding appears to increase the content of unsaturated fatty acids and lower the content of saturated fatty acids in animal milk (Fegeros et al., 1995; Moline-Alcaide and Yanez Ruiz, 2008) and therefore provides an appreciable dietary nutraceutical value. Notably, the fatty acid composition in the milk fat produced by the animals fed olive leaves had more oleic and linolenic acid and less myristic and palmitic acids. This trend is due to the fatty acid composition of olive leaves, which are rich in unsaturated fatty acids.

DSC is a fast and direct way to assess the quality of fats and to study their physical properties. DSC heating profiles help to explore the nature of the phase transition taking place during the melting of fats and oils (Gloria and Aguilera, 1998).

Olive leaf extracts in hexane showed the same melting profile (Figure 2). Olive leaf extracts exhibited a simple thermogram with a single peak after melting in the DSC. The melting of a fat results in volume expansion and a negative heat effect (Tan and Che Man, 2000). This showed the endothermic transition upon heating from a crystalline solid to the liquid state. The thermal parameters from the DSC melting curves of olive leaf extracts are shown in Table 3. The onset temperatures of the olive leaf extracts were significantly different (p<0.05). The values ranged from 26.50 to 30.10 °C.

The enthalpy of melting is the heat energy required for melting. This is calculated by integrating the area of the DSC peak on a time basis. Significant differences (p<0.05) were also encountered for the melting enthalpy of the samples and values ranged from 31.57 to 54.97 J·g⁻¹. The thermal parameters and the phase transitions in fats may be affected by the compositional changes between the samples such as fatty acid chain length, degree of unsaturation and nature of distribution of fatty acids in triacylglycerol species. The DSC application upon heating allowed for discriminating among oil samples from olives of different cultivars and/or harvesting periods. The thermal properties of monovarietal extra virgin olive oil samples were found to correlate well with the chemical composition i.e. triacylglycerols, diacylglycerols, total and free fatty acids and oxidation status (Chiavaro et al., 2008).

A mathematical model based on a simple regression procedure was also developed to correlate melting parameters to mass fractions of mono and polyunsaturated fatty acids (Fasina et al., 2008).

Regarding the peak temperatures of olive leaf extracts, the values ranged from 54.59 to 54.80 °C. The CH variety exhibited the lowest peak temperature value (54.59 °C) and it required a smaller amount of energy for melting (31.57 J·g⁻¹). This can be due to the higher level of unsaturated fatty acids and lower level of saturated fatty acids in the CH olive leaf extract. Indeed, the fatty acid composition influences the melting point values of the samples. Fats containing higher amounts of saturated triacylglycerol species commonly demonstrate a higher melting point than those which are highly unsaturated. Further, the peak temperatures of CL (54.79 °C) and ZR (54.80 °C) were statistically similar (p>0.05). This could be attributed to the similarities observed in the fatty acid composition between the two varieties.

de Man et al. (1991) consider the melting point as the peak temperature calculated from a DSC melting curve. The melting point is used to characterize the fats and is related to their physical properties, such as hardness and thermal behaviour.