Abstract: This research has been carried out to determine the filtration effect on the phenolic content and on the visual characteristics of olive oil. The influence of the filtration system on the phenolic and water content of virgin olive oil that differed in the year of production, production system and the olive variety was measured. Samples were filtered in the laboratory using two different systems (cotton or filter paper plus sodium sulphate anhydrous). Qualitative and quantitative variation of the phenolic fraction of virgin olive oils was evaluated by HPLC-DAD-MS and correlated with their water content (by Karl Fischer titration) and oxidative stability under forced conditions (by OSI). Colorimetric assays were also carried out in order to calculate the effect of filtration on the visual characteristics of virgin olive oil. After filtration the oxidative stability index decreased and in particular, filtration with cotton showed a significant loss of hydroxytyrosol, a phenol endowed with high antioxidant activity. One interesting behaviour was highlighted: Filtration with either cotton or paper plus anhydrous sodium sulphate led to an apparent increase in the phenolic content. These apparently contradictory data can be explained by considering that the reduction of the water content permits a higher availability of phenolic compounds that remain in oil and are extracted with the methanol-water mixture. Lastly, the filtration of virgin olive oil produced a loss in the intensity of green color and an increase in its lightness.
INTRODUCTION
Polyphenols is a term widely used to designate substances that possess a benzene ring bearing one or more hydroxyl groups, including functional derivatives (Harborne et al., 1989). Polyphenols are polar compounds that can be found in the olive fruit; however many of these compounds are modified or lost during the production process of virgin olive oil (Brenes et al., 1995). The final quantity of polyphenols is also influenced by the cultivar, climatic conditions during growth and degree of ripening (Di Giovacchino et al., 2002; Cerretani et al., 2005). Virgin olive oil is dominated by secoiridoid derivatives, followed by flavonoids and phenolic alcohols. The presence of secoiridoid derivatives provides an indication of the degradation pathways for the phenolic oleosides present in olive paste and wet pomace (Artajo et al., 2007). These derived compounds appear in virgin olive oil and possess antioxidant activity and a lower polarity compared to those in olive fruits (as glycosidic compounds). The partition coefficients between olive oil and water depend on the structure of these compounds and the number of hydroxyl groups.
The most abundant secoiridoids in virgin olive oil are the dialdehydic forms of elenolic acid linked to hydroxytyrosol or tyrosol (3,4-DHPEA-EDA or p-DHPEA-EDA) and an isomer of the oleuropein aglycone (3,4-DHPEA-EA). This can be explained by the fact that DHPEA-EA and 3,4-DHPEA-EDA are the compounds with the highest partition coefficients, as reported by Servili et al. (2005).
It is known that olive oil has a low quantity of water (Fregapane et al., 2006; Mendez et al., 2007) and for this reason olive oil can be considered as a water-in-oil emulsion. The presence of phenolic compounds in virgin olive oil and their high antioxidant activity can be explained by the so-called polar paradox (Porter et al., 1989) dictating that: Polar antioxidants are more effective in non polar lipids, whereas non-polar antioxidants are more active in polar lipid emulsions.
Frankel et al. (1994) demonstrated that interfacial phenomena are key to a better understanding of antioxidant action in heterogeneous foods and biological systems. The orientation of phenolic compounds in the oil-water interface and the active surface of water droplets influence protection against the oxidation of oil. This study concluded that lipophilic antioxidants are more effective in an oil-in-water emulsion system than in bulk oil, while an opposite trend has been found for hydrophilic antioxidants.
An important factor to consider is the visual appearance of virgin olive oil, as it will strongly influence consumer preference. Color is an intrinsic characteristic of each food product and helps to identify it, to the extent that the consumer is disconcerted if the color changes. From a hedonistic point of view, the color of olive oil can be considered an important organoleptic attribute that is a basic criterion in assessing quality, according to consumer preferences (McEwan, 1994; Pagliarini et al., 1994).
The compounds responsible for the color of virgin olive oil are chlorophylls, carotenoids and flavones (as apigenin and luteolin). Chlorophylls give olive oil its yellow-green color, carotenoids contribute in the yellow-red range (Mínguez-Mosquera et al., 1991) and flavones, having an absorbance maximum at around 330-350 nm, provide a yellow color.
Usually, the color of food is measured in L*a*b*. The L*a*b*, or CIELab, color space is an international standard for color measurements, adopted by the Commission Internationale d’Eclairage (CIE) in 1976. L* is the luminance or lightness component, which ranges from 0 to 100 and parameters a* (from green to red) and b* (from blue to yellow) are the two chromatic components, which range from -120 to 120 (Papadakis et al., 2000; Segnini et al., 1999; Yam et al., 2004). Nevertheless, the measurement of color is not currently required by regulations established by the European Economic Community (European Union Commission, 1991) to assess the quality of olive oil.
The aim of this report was to evaluate how different filtration processes (normally carried out during virgin olive oil production) affect the characteristics of virgin olive oil. In particular, oxidative stability, water content, the presence of each phenolic compound and color changes of virgin olive oil have been investigated. Eight types of virgin olive oil have been examined that were filtered using two different filtration systems (cotton or filter paper plus anhydrous sodium sulphate).
In our knowledge this is the first study in which olive oils with different origins have been analyzed in order to determine the effects of filtration focusing on the phenolic profile by using a separative technique as HPLC. Furthermore, the filtration systems used have been those that are traditionally applied in small mills.
MATERIALS AND METHODS
Apparatus
All HPLC analyses were performed using a HP 1100 Series instrument (Agilent
Technologies, Palo Alto, CA, USA) equipped with a binary pump delivery system,
degasser, autosampler, diode-array UV-VIS Detector (DAD) and Mass-Spectrometer
Detector (MSD). The analytical HPLC column used was a C18 Luna column,
5 μm, 25 cmx3.0 mm (Phenomenex, Torrance, CA, USA), with a C18
pre-column (Phenomenex) filter. The mobile phase flow rate was 0.5 mL min-1.
All analyses were carried out at room temperature.
The CIELab color space analyses were carried out using a ColorFlex instrument (HunterLab, Reston, VA, USA). To evaluate oxidative stability an eight-channels Oxidative Stability Instrument (OSI) (Omnion, Decatur, IL, USA) was used. The water content of virgin olive oils was obtained using a TitroMatic 1S instrument (Crison Instruments, S.A.; Alella, Barcelona, Spain).
Reagents and Standards
The standard used for HPLC quantification (3,4-dihydroxyphenylacetic acid)
was obtained from Sigma-Aldrich Inc. (St. Louis, MO, USA). Methanol, acetic
acid, acetonitrile and n-hexane were from Merck and Co. Inc. (Darmstadt,
Germany). All solvents were HPLC-grade and filtered through a 0.45 μm nylon
filter disk (Lida Manufacturing Corp., Kenosha, WI, USA) prior to
use. Double-deionized water with a conductivity less than 18.2 MΩ was obtained
with a Milli-Q system (Millipore, Bedford, MA, USA). Hydranal-Titran 2 and Hydranal-solvent
oil (solvents used to measure the water content with the volumetric titration
of Karl Fischer) were from Riedel-deHaën (Seelze, Germany).
Samples
Eight samples of virgin olive oil were obtained from different geographic
zones in Italy. They differed in the production year (oxidative state), production
plant (traditional and continuous) and storage conditions.
The analysis were carried out on July-September 2006 in the laboratories of the Department of Food Science of the University of Bologna, Cesena (Italy).
Filtration Processes
Two different filtration systems were utilized to reduce the water content
of virgin olive oils: (a) cotton and (b) paper and sodium sulphate anhydrous.
| • | Cotton: Fifty gram of virgin olive oil was passed through 0.5 g of cotton. |
| • | Paper and sodium sulphate anhydrous: Fifty gram of virgin olive oil were passed through filter paper. Next, 100 g of anhydrous sodium sulphate per L of oil was added to the sample and the oil was shaken in order to eliminate the water. |
Extraction of Polar Phenolic Fraction
Phenolic compounds were extracted from virgin olive oil by a liquid-liquid
extraction method according to Pirisi et al. (2000). The dry extracts
were dissolved in 0.5 mL of a methanol/water (50/50, v/v) solution and filtered
through a 0.2 μm syringe filter (Whatman Inc., Clinton, NJ, USA). The extracts
were frozen and stored at -43°C.
Determination of Phenolic Compounds by HPLC-DAD-MSD
Determination of the phenolic fraction was performed using an HPLC-DAD/ESI-MSD
equipped with a reverse phase C18 LunaTM column according
to Rotondi et al. (2004). Phenolic compounds detected at 280 nm were
quantified using a 3,4-dihydroxyphenylacetic acid standard calibration curve
(r2 = 0.9987). Phenolic compounds were tentatively identified based
on their UV-vis and mass spectra (Table 1) obtained by HPLC-DAD/ESI-MSD.
| Table 1: | Absorption maxima and fragmentation patterns using the ESI
interface of the compounds under study |
Evaluation of Oxidative Stability under Forced Conditions
These analyses were carried out in an eight-channel Oxidative Stability
Instrument (OSI) following Carrasco-Pancorbo et al. (2005) analytical
protocol.
Determination of Water Content in Virgin Olive Oil
The water content was analyzed with a TitroMatic 1S instrument. This measurement
uses a Karl-Fischer titration based on a bivoltametric indication (2-electrode
potentiometry). A solution of chloroform: Hydranal-solvent oil (a methanolic
solvent) 2:1 (v/v) was used to dissolve the sample and Hydranal-Titran 2 was
used as a titrating reagent. Each sample was introduced three times and the
quantity of sample was measured with the back weighing technique. The sample
was dissolved in the solution of chloroform: Hydranal-solvent (2/1, v/v) oil
and the titrating reagent was added until the equivalence point. The quantity
of water was expressed as mg of water per kg of oil (n = 3).
Statistical Analysis
Data were analyzed using Statistica 6.0 (Statsoft, Tulsa OK, USA) statistical
software. The significance of differences at a 5% level between averages was
determined by a one-way ANOVA using Tukey’s test.
RESULTS AND DISCUSSION
Effect of Filtration on the Water Content
In most olive oil samples the water content decreased significantly after
filtration (Table 2). With the exception of sample 7, the
control sample always had a higher water content, whereas samples that had been
filtered with cotton showed a significant decrease in the quantity of water.
However, samples that had been filtered using paper and anhydrous sodium sulphate
presented a significant decrease in only four of the eight samples (samples
2, 3, 4 and 8).
Filtration with cotton, termed filtration Bari style, is especially widespread in the olive oil industry plants located in the South of Italy and as is shown from experimental data in Table 2, can be considered effective in reducing the water content.
Evaluation of Oxidative Stability under Forced Conditions
The Oxidative Stability Index (OSI) decreased after the filtration with
cotton and paper plus anhydrous sodium sulphate (Table 2).
This effect was more pronounced in samples that showed a higher OSI value. For
example, the OSI of sample 2 decreased about 4.4% with cotton filtration and
by 18.0% with anhydrous sodium sulphate plus paper; sample 6, another control
sample with a high oxidative stability, showed a reduction of 15.0% with cotton
filtration and 19.0% with anhydrous sodium sulphate plus paper. However samples
4 and 5 that had a low OSI value compared to the control sample, demonstrated
a decrease of 2.9 and 8.1%, respectively, with cotton filtration and a reduction
of 2.3 and 14.8%, respectively, with anhydrous sodium sulphate plus paper filtration.
| Table 2: | Analytic results of virgin olive oils: Oxidative stability
(OSI, RSD% = 2.3), Water content (mg of water per kg of olive oil) and Phenolic
compounds (mg of analyte as 3,4-dihydroxyphenylacetic acid per kg of olive
oil). Quantification of the Individual Components (n = 7) (±SD) |
Paper+SAA, Filtration by paper and sodium sulphate anhydrous;
HYTY, Hydroxytyrosol; TY, tyrosol; VA, Vanillic acid; DMOA, Decarboximethyl
oleuropein aglycon; Pin, pinoresinol; DLA+AcPin, Decarboxymethyl ligstroside
aglycon+acetoxypinoresinol; Ol Agl, Oleuropein aglycon; LA, Ligstroside
aglycon, Letter(s) a-c in brackets indicate statistically significant differences
(HSD Tukey p<0.05) |
|
Considering these results, it can be surmised that the oxidative stability of virgin olive oils is lower when the water content is decreased (after filtration), which is related either to a loss of phenolic compounds or a reduction in their antioxidant activity. The decrease of antioxidant activity, as mentioned before, can be explained by the polar paradox. In fact, phenolic compounds, being polar molecules, have a higher activity in a water-in-oil emulsion. However after filtration the water content is reduced. As a consequence, the antioxidant capacity of these compounds diminishes, probably due to their particular orientation around small droplets of water.
Evaluation of Behavior of Individual Phenolic Compounds after Filtration
Hydroxytyrosol, decarboxymethyl oleuropein aglycon and oleuropein aglycon
are, in that order, the phenolic molecules of virgin olive oil having the highest
antioxidant activity (Carrasco-Pancorbo et al., 2005). In general, hydroxytyrosol
showed a significant decrease after filtration with cotton with respect to the
control sample. This behavior can be explained considering the partition coefficient
between olive oil and water of this compound (Kp = 0.01 as reported Servili,
2005), which makes it more soluble in water than other phenols.
As reported in Table 2, the concentration of several phenolic compounds seemed to increase after filtration, but is related to the fact that filtration reduces the water content even though the loss of phenolic compounds is not proportional. In fact, it is assumed that the majority of phenolic compounds located around water droplets remain in olive oil.
It is also possible to hypothesize that extraction of phenolic compounds in control samples does not allow for complete recovery of these analytes; indeed, when the analytes are in a more polar matrix the affinity of phenolic extraction to the solvent (methanol/water, 60/40, v/v) is lower and their separation is more difficult. On the other hand, if the extraction with a hydroalcoholic solution is done after the partial elimination of water, phenols are more available to the solvent mixture.
This study represents a novelty in the fact that this apparently increase of phenolic content has been explained by the performance of a study about the variation in water content of virgin olive oil. This type of study has never been considered before by other investigations carried out about filtration effect (Fregapane et al., 2006).
Filtration Effect on the Colorimetric Parameters
A large amount of the particles in suspension was retained by the filtration
system. As shown in Table 3, the luminosity of olive oil (L*
value) increased after filtration. Furthermore, when the control sample had
a deep green color, the a* value increased after filtration and the intensity
of green color was minimized; whereas if the control sample was light green,
the a* value decreased and the contribution of green color was more apparent.
The b* value had a tendency to increase because the yellow color was more evident
when the olive oil had been filtered.
| Table 3: | Values of L*a*b* coordinates of the eight virgin olive oils
studied: unfiltered samples (control sample) and filtered samples (cotton
and paper plus anhydrous sodium sulphate) (±SD) |
| Paper+SAA, filtration by paper and sodium sulphate anhydrous | |
CONCLUSIONS
By study of two filtration systems it can be concluded that the oxidation stability decreases after filtration due to elimination of water. This could be due both to the decrease of the concentration of phenols with a higher antioxidant activity, particularly hydroxytyrosol and to the decrease of antioxidant activity of phenolic compounds when the water content is lowered. Furthermore, filtrated olive oils had a higher component of yellow color, luminosity and in some cases, the intensity of green color diminished. Presently, consumers have more knowledge about olive oil and would choose a veiled and deep green oil over one that is transparent and light green oil (filtered). However there are also consumers that prefer transparent oils. Thus filtration may reduce the quality of virgin olive oils (oxidative stability decrease) and many consumers may not prefer these products due to their unfavorable visual characteristics.