Antioxidant Activity of Phenolic Compounds in Bulk Camellia Oil and Corresponding Oil in Water (O/W) Emulsions

The antioxidant activities of gallic acid, propyl gallate, (+)-catechin and quercetin in bulk camellia oil and corresponding emulsions were investigated using hydroperoxides and nonanal as indicators of oxidation. In bulk oil, these phenolic compounds showed antioxidant activities with the order of gallic acid>quercetin>propyl gallate> (+)catechin when measured using hydroperoxides and gallic acid>quercetin>(+)-catechin>propyl gallate using nonanal. In emulsions, propyl gallate and (+)-catechin showed antioxidant activities for the entire duration of the experiment, while quercetin and gallic acid displayed antioxidant activities for 10 days, and then pro-oxidant activities thereafter. Results suggested that in bulk oil, the antioxidant activity was regulated by phenol polarity and hydrogen atom donating ability too. In emulsions, the antioxidant activity of phenolic compounds seems to be related to their affinity toward the emulsifing agent rather than their polarity.


INTRODUCTION
Camellia oil (Camellia oleifera) is one of the four major tree-bearing oils (palm, olive and coconut oils) with health benefits, since it has abundant antioxidants, such as tocopherols, polyphenols etc. Data related to camellia oil polyphenol content have only become available in the last decade. Liu and Zhao (2002) used TLC and phenolic-specific spray reagents to detect phenols in Camellia sinensis seed oil. They suggested that polar antioxidant compounds played a major role in the stability of the oil. Zhong et al. (2006) measured total phenols and reported the phenolic profile of a water/methanol extract of cold-pressed camellia oil. The addition of exogenous phenols, caffeic acid and tyrosol, resulted in a decrease in peroxides and secondary oxidation products in bulk camellia oil during heating (Zhong et al., 2007).
On the other hand, the performance of phenolic antioxidants in oil-in-water emulsions is of interest from numerous viewpoints. In human nutrition, it has been recently proposed that tea polyphenols can modify the emulsification of dietary lipids and this may change their digestion and absorption (Shishikura et al., 2006) In food and health studies, there has been much interest in antioxidant activity in emulsions due to the fact that lipids are more susceptible to oxidation in an emulsion due to the high surface/volume nature of the emulsion facilitating the interaction of pro-oxidants in the aqueous phase with high surface area dispersed lipids (Shahidi and Zhong, 2011). To date there have been no studies on the oxidation of camellia oil emulsions.
Individual polyphenols play different roles in bulk oil and oil-in-water emulsion due to different molecular structures. For several decades, the "Polar Paradox" theory has been the dominant theory to explain the different behaviors of antioxidants in heterogeneous phase (bulk oil and emulsions) (Porter et al., 1989;Porter, 1993). It has been suggested that the activity of an antioxidant depends on its partition affinity between the oil-water and oil-air interfaces, where oxidation occurs. In bulk oil, hydrophilic antioxidants distributed at the oil-air interface can better protect lipid oxidation than lipophilic antioxidants dissolved in the lipid phase. In emulsions, lipophilic antioxidants, distributed at the oil-water interface, better protect lipid oxidation than hydrophilic antioxidants dissolved in the aqueous phase. Additionally, hydrogen bonding between hydrophilic antioxidants and water may reduce the ability of the phenolic compound to donate hydrogen atoms to inhibit oxidation (Frankel et al., 1994Schwarz et al., 1996;Pekkarinen et al., 1999;Zhao and Hall.III, 2007;Mattia et al., 2009).
Despite the popularity of the polar paradox theory, there is recognition that it may not account for all situations of antioxidant, lipid, emulsifier, etc. In fact the interfacial particle size of droplet (surface/volume ratio) and type of emulsifier (cationic, anionic, and neutral) may affect the oxidative stability of emulsions. These additional factors are now being considered in response to studies showing nonlinear relationships between polarity and antioxidant activities in emulsions (Laguerre et al., 2010;Shahidi and Zhong, 2011) and some phenolic compounds acting as pro-oxidants in emulsions Mattia et al., 2009). More comprehensive factors (i.e., emulsifier, concentration of antioxidants, lipid character, surface/volume ratio) may be responsible for antioxidant activity, apart from polarity of the antioxidant.
In the present study, the antioxidant activity of phenolic compounds in bulk camellia oil and corresponding emulsions was investigated and some mechanistic interpretations will be discussed. Antioxidant activities of individual phenolic compounds were evaluated by determining hydroperoxides  and nonanal (Hall et al., 2005) which represented the primary and secondary oxidative products, respectively.
Preparation of bulk oil and oxidation: 0.8 mL of 2 mM gallic acid, propyl gallate, (+)-catechin and quercetin methanol solution was added into 50 mL vial, respectively, and purged under nitrogen prior to addition of 20 mL of oil. Sample without phenolic compound addition was used as control. All samples were placed randomly in an incubator shaker (ZHWY-2102C, Shanghai Zhicheng Analytical Instrument Manufacturing Co., Ltd, China) in the dark at 60°C (Zhao and Hall.III, 2007). Oxidated oils were collected after 0, 2, 4, 6, 8 days for further determination.

Preparation of emulsions and oxidation:
The emulsions (40%, w/w) were prepared using the modified method described by Frankel et al. (1996). Forty percent of camellia oil, 59% of deionized water and 1% of 20 were transferred into an Erlenmeyer flask.
Determination of oxidised products: The content of hydroperoxide was determined by the method from Crowe and White (2001). The nonanal was measured by headspace solid phase microextraction-gas chromatography (SPME-GC). Oil (0.19 g), emulsions (0.19 g) was added to vial (15 mL, supelco) which was contained dodecane (IS), respectively, and sealed with Teflon lined septum, the sealed vials were placed in incubator (IKA, RCT basic, Germany) at 95°C, The SPME needle (DVB/CAR/PDMS, 50/30 µm, 1 cm, Supelco) was inserted into the headspace of vial and left exposed for 10 min, removed and immediately desorded at the GC (Shimadzu-2014) injector at 250°C with splitless mode. Separation was achieved on a SP 2340 column (60 m×0.25 mm×0.2 µm film thickness, Supelco) using nitrogen as carrier gas with flame ionization detection (FID, 250°C). The oven program started at 40°C for 2 min, then at the rate of 3.00 °C/min to 160°C for 2 min, increasing at 12°C/min to 220°C with final isothermal period of 5 min.
Statistical analysis: Data analysis was done using IBM SPSS statistics 19.0. All analyses were performed in triplicate.

RESULTS AND DISCUSSION
Antioxidant effect of phenolic compounds on bulk camellia oil: On the basis of hydroperoxides formation, gallic acid, propyl gallate, (+)-catechin and quercetin showed antioxidant activities compared with control (p<0.05), respectively. While, individual phenolic compound represented different antioxidant efficiency, the antioxidant activities were ranked as the following order: gallic acid>quercetin>propyl gallate>(+)-catechin>control (Table 1). Among them, gallic acid was the best antioxidant compared with others (p<0.05). On the basis of nonanal formation, the rank of antioxidant activities was similar to hydroperoxides formation except that (+)-catechin played a better role than propyl gallate (p>0.05) after the 4 th day. Therefore, the sequence of antioxidant activities using nonanal indicator was: gallic acid>quercetin> (+)-catechin>propyl gallate>control (Table 2).
Gallic acid, propyl gallate, (+)-catechin and quercetin (Fig. 1) involve in different affinity between lipids and aqueous phase. Gallic acid has a great affinity with aqueous phase due to the abundant of        The values are represented by means±SD, n = 3. Values within a column by no letter or the same letter are not significantly different (p>0.05)   The values are represented by means±SD, n = 3. Values within a column by no letter or the same letter are not significantly different (p>0.05) hydroxyl group which endow the molecular structure a high polarity (Mattia et al., 2009). Propyl gallate consists of propoxy and gallic acyl groups which has affinity with lipids and aqueous phase, respectively, hence, it can play some interfacial properties. (+)-Catechin has been considered an amphiphilic compound due to its two benzenic rings and a central eterocyclic oxygenated ring structure (Mattia et al., 2009). Quercetin belongs to a lipophilicity compound and is the less polar among these phenolic compounds (Burda and Oleszek, 2001) which characterized by its pyrone ring structure. As described above, the polarity of these phenolic compounds were as the following order: gallic acid>catechin, propyl gallate>quercetin (Mattia et al., 2009). According to the "polar paradox", hydrophilic antioxidants were more effective than lipophilic antioxidants in bulk oil. The sequence of antioxidant activity of this research for individual phenolic compounds (see above) was not following the "polar paradox" absolutely. It was supposed that polarity is not the only factor to determine efficiency of antioxidants. Mattia et al. (2009) reported that the ability (amount) of hydrogen atoms donating of these antioxidants following the order of quercetin>catechin>gallic acid, propyl gallate. Additionally, gallic acid was faster than quercetin for donating hydrogen atoms. Regarding of the inhibition of nonanal formation, (+)-catechin was a better antioxidant than propyl gallate, it was suggested that catechol structure was more effective in prohibition of hydroperoxides decomposition than that of pyrogallol structure.
According to "polar paradox", interface is the site where oxidation occurred in emulsion. Hence, high surface/volume ratio of emulsion makes it more susceptible to oxidation. In this study, propyl gallate, (+)-catechin showed antioxidant activity in emulsion with a smaller particle sizes (high surface/volume ratio) than others (Fig. 2). It seems to have a contradictory with the theory of "polar paradox". This phenomenon should be explained that the concentration of propyl gallate or (+)-catechin may be enough for colonising the interface. Under this circumstance, the surface/volume ratio is not the dominant factor in emulsion oxidation.

CONCLUSION
In bulk camellia oil gallic acid, propyl gallate, (+)catechin and quercetin showed different positive efficiency in inhibiting lipid oxidation. While in camellia oil emulsion gallic acid, quercetin showed antioxidant efficiency during the initial 10 oxidating days, then pro-oxidant activity after the 10 th day of oxidation; Propyl gallate, (+)-catechin played antioxidant roles during the whole period of the oxidating. These factors confirm that interface is the site where oxidation occurred in bulk oil and emulsions. In bulk oil, the antioxidant activity was regulated by polarity and ability, rapidity of donating hydrogen atoms of phenolic compounds. In emulsions, the antioxidant activity of phenolic compounds prefer to the affinity toward emulsifier (tween 20) rather than their polarity.