Effect of methanol extracts of rosemary and olive vegetable water on the stability of olive oil and sunflower oil

Extractos metanólicos de fenoles de hojas secas de romero y filtrados de agua de vegetación de la aceituna, en combinación con BHA, se añadieron al aceite de oliva (mezcla de aceite de oliva refinado y virgen, 3 a 1) y al aceite de girasol, evaluándose sus efectos antioxidantes usando condiciones aceleradas. Estas condiciones incluyeron el test del horno de oxidación (a 63 °C) y el método de conductividad (Rancimat a 120 °C). También se aplicó al proceso de fritura a 180 °C. Los extractos metanólicos de fenoles y el BHA se añadieron a cada aceite en las siguientes concentraciones: 200 ppm de extracto de romero, 200 ppm de extracto de agua de vegetación de la aceituna, 100 ppm de extracto de romero + 100 ppm de BHA, 100 ppm de extracto de agua de vegetación + 100 ppm de BHA y 200 ppm de BHA. En general, el efecto antioxidante de los aditivos fenólicos de romero y de BHA tuvo el siguiente orden: 200 ppm de extracto de romero > 100 ppm de extracto de romero + 100 ppm de BHA > y 200 ppm de BHA. La adición de 200 ppm de extracto de agua de vegetación y 100 ppm de extracto de agua de vegetación + 100 ppm de BHA mostró un efecto antioxidante similar al de 200 ppm de BHA.


INTRODUCTION
Natural and synthetic antioxidants are used to delay or to retard the oxidative deterioration of fats and oils.The major food antioxidants are phenolic compounds and are generally referred to as phenolic antioxidants (Sherwin, 1990).Recently, consumers show a growing tendency to reject all synthetic additives to foods, including synthetic antioxidants.Thus, there has been an increased interest on the use of natural antioxidants, especially those extracted from herbs, spices and other plant materials (Manganari y Oreopoulou, 1991;Tsuda et al., 1993;Frankel et al., 1997;Chen étal., 1998;Hall et al., 1998).Several substances including alcoholic extracts of different spices, paprika and rosemary, show good antioxidant activity.Rosemary extracts exhibited higher antioxidant effect compared to other natural extracts (Houlihan et a!., 1984;Houlihan et al., 1985).According to Satue et al., (1995) antioxidant compounds extracted from virgin olive oil increased the oxidative stability of refined olive oil.Antioxidants present in virgin olive oil, contribute to its oxidative stability (Kiritsakis, 1998;Blekas y Boskou, 1998).Vegetable water, a main by-product of olive fruit processing, contains sufficient amount of polyphenols (18-125 mg/g), which may act as antioxidants and prevent lipid oxidation (Unal, 1994).
The present study was undertaken to gain knowledge on the effect of rosemary leaves and of vegetable water extracts on olive oil (blend of refined and virgin olive oil, 3 to 1) and sunflower oil, under accelerated oxidative conditions.

Rosemary Leaves and Vegetable
Refined olive oil and sunflower oil, obtained from a local market, was used in this study.Virgin olive oil was provided by a local producer.Parts three to one (w/w) of refined and virgin olive oil were mixed and blend olive oil was formed.Rosemary leaves were collected from the garden of the Mediterranean Agronomic Institute of Chania (MAICh).The leaves were left to dry at room temperature.Vegetable water from a double phase decanter, from the olive oil mill of the Subtropical and Olive Trees Institute of Chania, was used.

Methanol Extracts of Rosemary Leaves
The methanol phenolic extracts of rosemary were prepared according to the method proposed by Wu et al. (1982).500 g of dried rosemary leaves were ground to pass through 0.5 mm sieve, and phenols were extracted with 3 L of methanol (95%) at 60 °C for 2 h.The mixture was filtered and the residue was extracted again with 2 L of methanol.The combined filtrate was bleached with 100 g of active carbon for 15 min, and then filtered to yield a light-brown filtrate.The methanol solution was concentrated to 300 ml by a rotary evaporator and then filtered to remove any precipitates.

Vegetable Water Extracts
The vegetable water extracts were prepared according to the method proposed by Balice and Cera (1984).50 ml of vegetable water, were filtered through a Watman 41 filter paper and then acidified with HCI (pH= 2), and saturated with NaCI.Then^ 50 ml of ethyl acetate were added and the mixture was stirred for 10 min.Ethyl acetate solution was dried over anhydrous sodium sulphate.

Addition of Antioxidants to Oil Samples
100 g of each oil (blend olive oil and sunflower oil) was transferred to 250 ml flask and the exact amount of the phenol extract or the synthetic antioxidant was added to the oil.Methanol phenolic extracts and the synthetic antioxidant were added alone and in combinations as follows: 200 ppm rosemary extract, 200 ppm vegetable water extract, 100 ppm rosemary extract + 100 ppm BHA, 100 ppm vegetable water extract + 100 ppm BHA and 200 ppm BHA.After complete mixing, the solvent was evaporated at 55 °C under vacuum, in a rotary evaporator.The volume was completed with oil and the mixture was left to equilibrate.

Oven Test
Oil samples (50 g each) containing the additives were transferred into a series of capped transparent glasses and put in an oven at 63 °C.The peroxide value (PV) of the oil samples was periodically determined by the AOCS (1989) method.

Conductivity Method
The antioxidant activity of the tested additives was also determined using a Rancimat apparatus (Metrohm AG, CH-9100 Herisau, Switzerland) (Laubli and Bruttel, 1986).3 g of oil samples were used.Temperature was set at 120 °C and the flow rate at 20 L/h.The induction period was recorded automatically.

Frying Process
The frying process started 10 days after the additives were added to the oils.During the 10 days period all treated samples and the control were stirred occasionally for 20 min.350 g of potatoes were fried (French fries) in 900 g of oil at 180 °C for 15 min.The frying procedure was repeated eight times.The total frying lasted 120 min.30 g of oil sample was weighted after each frying and stored in the refrigerator until the time of analysis.Peroxide value and specific absorbance (K232, K270 and AK) measurements as described by Kiritsakis (1998) were conducted to measure the degree of oxidation.The phenol content of the oil samples was also determined during the study, by the Gutfinger (1981) method.

Statistical Analysis
The Statistical analysis was carried out using the LSD test and DUNCAN'S Multiple Range test to evaluate the significant differences between the samples.

Evaluation of Additives by the Oven Test
Table I shows the effect of methanol phenolic extracts on peroxide formation in oil samples when (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://grasasyaceites.revistas.csic.es the oven test was used.In the control samples of blend olive oil, the PV was 77 meq/kg oil at 120 days, while the control samples of sunflower oil reached 85 meq/kg oil in 12 days.This was expected due to the faster oxidation (Chan et al., 1982) of the polyunsaturated sunflower oil and to the presence of natural antioxidants in blend olive oil, which act as potent antioxidants (Papadopoulos and Boskou, 1991).The relative inhibitory effect of the different additives in blend olive oil was in the following order: 200 ppm rosemary extract > 100 ppm rosemary extract + 100 ppm BHA > 200 ppm BHA.Neither 200 ppm vegetable water extract nor 100 ppm vegetable water extract + 100 ppm BHA improved the oxidative stability of the samples.Samples containing 200 ppm rosemary extract exhibited the highest oxidative stability for both oils.Manganari and Oreopoulou (1991) also found high antioxidant effect of rosemary extracts when tested in lard and in corn oil.

Evaluation of Additives by tlie Conductivity IVIethod
Table II shows that the results recorded by the Rancimat apparatus were consistent with those obtained by the oven test (Tables I).The induction period for the control samples of blend olive oil and those of sunflower oil was 5.55 and 2.03 h, respectively.Under these conditions, phenolic extracts increased the resistance of oil samples to oxidation.Vegetable water extract seemed to show similar antioxidant effect to BHA.Both oil samples containing 100 ppm rosemary extract + 100 ppm BHA, or 200 ppm rosemary extract showed longer induction period than oil samples containing the other additives (Table II).Blend olive oil containing 200 ppm rosemary extract exhibited longer induction time (11.2 h) than sunflower oil samples (3.25 h) as compared to their respective controls.

Evaluation of Additives during the Frying Process
Table III shows that K232 values for both oils increased with the frying time indicating the formation of conjugated hydroperoxides.At the end of the frying process the K232 values for blend olive oil containing no additives were lower than those of the control sunflower oil.The blend olive oil had initial K232 value 2.64 which increased to 4.30 by the end of frying.In contrast, sunflower oil had an initial value of 3.72 which increased to 10.11 (Table III).Blend olive oil samples containing 200 ppm rosemary extract showed the lowest K232 value (3.92) by the end of frying.Our results agree with others (Manganari y Oreopoulou, 1991;Houlihan etal., 1984).The K270 values increased significantly in both oils with frying time (Table IV), indicating a continuous formation of secondary oxidation products (Hamilton, 1989;Kiritsakis,1998).The initial K270 value for control samples of blend olive oil was 0.81 and increased to 1.19 by the end of frying.However, the control samples of the sunflower oil had initial value 2.43 which increased to 3.55, indicating the formation of more decomposition oxidation products in sunflower oil than in blend olive oil.The latter contains more oleic and less linoleic (Kiritsakis, 1998) and according to Nawar (1985) thermal oxidation of linoleic gives higher amount of decomposition products than oleic.However, samples containing 100 ppm rosemary extract + 100 ppm BHA, or 200 ppm rosemary extract showed the highest absorbance at 270 nm at the end of frying.These results were inconsistent with K232 values and indicate that the measurement of UV absorbance is not the best parameter to evaluate oil alterations during frying.Frying temperature and food humidity may affect the values.AK values did not provide additional information since values increased but not significantly in both oils with frying time (Table V).I) In both (blend olive oil and sunflower) oils a significant decrease in polyphenol content with the time was observed (Figs. 1, 2), due to the effect of frying temperature (Loliger, 1989).Natural polyphenols (60 ppm) present in the control samples of blend olive oil (control samples of sunflower oil contained no polyphenols), were more resistant to high temperature than polyphenols from the antioxidants extracts.Among the added extracts, those from vegetable water showed higher resistance.Samples to which 200 ppm polyphenols from the vegetable water were added, contained at the end of the experiment, the higher amount of polyphenols, which was 30.5 ppm and 18.6 ppm for blend olive oil and sunflower oil respectively.The presence of lower amount of polyphenols, at the end of the experiment, in samples to which 200 ppm of rosemary extract extract was added compared to those containing 200 ppm vegetable water extract (Figs. 1, 2) may explains the higher antioxidant effect of rosemary extracts.Polyphenols were probably used and the oxidation was retarded.
In conclusion, under accelerated oxidation conditions (63 °C in oven and 120 °C in Rancimat), methanol phenolic extracts of rosemary leaves and vegetable water exhibited antioxidant activity.Better results were obtained from rosemary than vegetable water.

Table I The effect of phenolic additives on peroxide formation in blend olive oil and in sunflower oil (Oven test at 63 °C)
Values with the same letter are not significantly different with Duncan's Multiple Range Test.BO = Blend olive oil (mixture of refined olive oil with virgin olive oil, 3/1).SO = Sunflower oil.Vw = Vegetable water extract.Ros = Rosemary extract.(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://grasasyaceites.revistas.csic.es