EXTRACTION OF PHENOLIC IN MANGO PEELS

S. M. V. Palmeira1; L. M. Gois1 and L. D. Souza2

Chemical Engng. Department, UFBA, Salvador, Bahia, Brazil. silvana.palmeira@gmail.com, lmario@ufba.br Tecnology Dto, UEFS, Feira de Santana, Bahia, Brazil diasliane@hotmail.com Abstract - Scientific studies show that mango peel is a rich source of phenolic compounds, substances with antioxidant activity. To contribute to the future use of this material, this paper aims to make an optimization of yield extraction of phenolic compounds from mango peels. The effect of the parameters mango variety, peel drying time, ethanol concentration and solvent temperature were selected to study the optimization of the yield extraction. A two-level, full four factorial design (24) was used in the design of the experiments and in the analysis of the results. With preliminar results it was performed a star expeimental design around the center point to obtain a response surface. The maximum yield of 33.7% of total phenolics was obtained for variety Espada. The optimum conditions were ethanol concentration of 70% as extraction solvent and mango peel drying time of 10 hours. Keywords?? Mango, phenolic compound, extraction l lI. INTRODUCTION Considering the importance of bioactive compounds on human health, studies have been undertaken for its characterization and quantification in different foods. Recent research performed by Berardini et al (2005) revealed that mango peels (Mangifera indica L.) is composed of various phenolic compounds. The authors characterized approximately fifteen phenolic compounds in the mango peels, showing the feasibility of obtaining flavonoids and also xanthones in the peel of the fruit. Clinical studies have found that the phenolic compounds, in addition to antioxidant activity, have antiviral, antibacterial and anti-inflammatory activity [Zgrka and Kawka (2001), cited in Santos and Hasman (2002)]. Other epidemiological studies have shown that phenolic compounds present in fruits, grains and vegetables have contributed significantly to the reduction of chronic and degenerative diseases in populations whose diet have high intake of these foods. There is evidence that the intake of polyphenols is inversely related to incidents of cancer. Mango grown in 85 countries and is considered one of the most important tropical fruit in the world, as 50% of tropical fruit produced are mangos. According to FAO (Food and Agriculture Organization of United Nations), the world's supply of mangos in 2004 was approximately 26.3 million tons. Brazil annual production is about 0.84 million tons, and participation of 3% of the total volume offered. [Manica and Oliveira (2005)]. The Northeast region of Brazil accounts for more than 50% of national production, with emphasis on the So Francisco Valley, with about 250,000 tons / year. Part of this production is consumed as fresh fruit in the domestic market, another part is exported and another part is absorbed by the juice processing industries. Byproducts of processing amount 3560% of the total fruit weight, are discarded in landfills. The exploitation of this natural waste product, becomes attractive because of its importance from both an environmental and socioeconomic point of view. So, to contribute to the future use of this material, this study aims to make an optimization of yield extraction of the phenolic compounds from mango peels. A full four factors (24) factorial design and response surface methodology was used in the design of the experiments and in the analysis of the results. The effect of the parameters mango variety, peel drying time, ethanol concentration and solvent temperature were selected to study the optimization of the yield extraction. To measure the content of total phenolic was performed the Folin-Ciocalteau method and UV-vis spectrophotometry. lII. METHODS lA. Reagents The reagents used in this study consisted of pure gallic acid, sodium carbonate anhydrous, reagent FolinCiocalteau, ethanol, and pure water. The reagent FolinCiocalteau is a polymer ion complex solution formed by molybdate, tungstate and phosphoric acid (H3PMO12O40 and H3PMo12O40). This reagent was used in the oxidation of the phenolate, reducing them to blue oxides of molybdenum and tungsten (W8O23 and Mo8O23). lB. Samples Peels were from two varieties of mango, "Espada" and Tommy Atkins, both acquired in open markets in the region of Feira de Santana/Bahia/Brazil. Fruits of the variety Espada were purchased in october of 2006 and of the variety Tomy Atkins in march of 2007. The

fruits were promptly peeled with stainless steel knife and the peels were stored in Polyethylene plastic bags at frozen temperature under -5C in freezer. So, before extraction processing, samples of variety Spada were stored for five months and Tommy variety for only a week. lC. Experimental Procedure

implies 16 runs, whose analysis matrix is shown in Table, below. Table 2 : Analysis matrix Test Mangovariety peel Drying time (h)Solvent Temperature (oC)Ethanol Concentration (% v/v)1Espada (-)15 (-)70 (-)50(-)2Tommy (+)15 (-)70 (-)50(-)3Espada (-)24 (+)70 (-)50 (-)4Tommy (+)24 (+)70 (-)50 (-)5Espada (-)15 (-)90 (+)50 (-)6Tommy (+)15 (-)90 (+)50 (-)7Espada (-)24 (+)90 (+)50(-)8Tommy (+)24 (+)90 (+)50(-)9Espada (-)15 (-)70 (-)0 (+)10Tommy (+)15 (-)70 (-)0 (+)11Espada (-)24 (+)70 (-)0 (+)12Tommy (+)24 (+)70 (-)0 (+)13Espada (-)15 (-)90 (+)0 (+)14Tommy (+)15 (-)90 (+)0 (+)15Espada (-)24 (+)90 (+)0 (+)16Tommy (+)24 (+)90 (+)0 (+)

The general procedlE. Results ures for sample preparation, extraction and determinaThe results presented in Table 3 were subjected to tion of the concentration of phenolic are represented in statistical analysis using Statistica software version 6.0 the flowchart shown in Figure 1. According to the flow diagram, mango peels were dried and grinded. Later, the to determine the effect of each extraction parameter on material obtained was mixed with the solvent specified total phenolic content of the extracts. for extraction. The suspension was kept in a process of dynamic steeping for 30 minutes and then filtered under Table 3 Extracts total phenolic concentration vacuum. Finally, extract phenolic content was determ- Experiment 1Experiment 2 16 83.0 55.0 57.5 15140.0103.0121.5 14 68.0133.0100.5 13158.0113.0135.5 12 60.0 68.0 66.5 11113.0113.0113.0 10 65.0 Average 9150.0285.0217.5 8165.0178.0171.5 7370.0340.0355.0 6178.0203.0190.5 5365.0340.0352.5 4198.0200.0199.0 3340.0363.0351.5 2315.0203.0259.0 1393.0370.0381.5 70.0 76.5 ined by the method of color reaction of Folin-Ciocalteau TestTotal Phenolics concentration (ppm) [Georg (2005)] and a Cary Varian fiber optical UV-vis spectrophotometer gave a direct measurement of the phenolic concentration at 760 nm. Figure 1 - Experimental procedure flow chart D. Experimental Design The effect of the parameters mango variety, peel drying time, ethanol concentration and solvent temperature were selected to study the optimization of the yield extraction of total phenolics in mango peels. It was performed a two-level, full four factorial design (24) showed in a tabular form through the matrix notation for experimental levels in Table 1, below. Table 1: Matrix notation for experiment levels ParameterLevels(-)(+)Mango Variety EspadaTommyPeel drying time (h)15.024.0Ethanol Concentration (% v/v) 50.0 0.0 Solvent Temperature (C) 70.090.0 As it can be seen, the factor level settings were the mango variety, the peel drying time, the ethanol concentration in a ethanol/water mixture and the solvent temperature used in the process. Considering the 24 factorial design described later, not counting replications, there are sixteen different ways of combining settings, which

E.1 The calculation and evaluation of the effects With the results in Table 3, the effects of each parameter on extract phenolic content were calculated, at the significance level of 5% (p ? 0.05). So, ten effects were calculated, classified as follow: four primary (variety, drying time, temperature and concentration of solvent)

and six, resultants from interactions between two factors. Following is a brief discussion about them. E.1.1 - The primary effects Table 4 illustrates the values of the main effects found. It can be observed first that the largest effect is the concentration of the solvent (-171.50), followed by the variety of mango (-113.37) and the effect of peel drying time (-32.25). On the other hand, the effect of temperature of the solvent was non-significant (p> 0.05). Table 4: Analysis of variance showing significance of the primary effects ParmeterEffectpVariety-113.3750.000000Drying time (h)-32.2500.022893Solvent Temperature (oC) -17.7500.186493Solvent Concentration171.5000.000000 Figures 3, 4 and 5 shown below, clearly identify the influences of these three parameters on the yield of extraction through the values of the mean concentrations of phenolic compounds found. Figure 3 compares the values of global average of total phenolic concentrations with the two solvents used (pure water and ethanol 50%). In this case the averages of total phenolic concentrations obtained with pure water (1 to 8 tests) was 111.1 ppm and with ethanol 50% (9 t0 16 tests) was 282.6 ppm.
300

Figure 4 compares compares the values of global average of total phenolic concentrations of the extracts obtained from two varieties of mango. It was taken the arithmetic mean of the concentrations of phenolic compounds obtained in the tests where we used only the variety Espada (1,3,5,7,9,11,13 and 15), and the average of the tests where we used only the variety Tommy (2,4,6,8,10,12,14 and 16). The overall average for the 1st test group was 253.5 ppm, while for the 2nd group of tests the average was 140.1 ppm .
300

Phenolics Concentration (ppm)

200

Espada
100

Tom my

0
1 2

Figure 4 - Effect of the variety of mango on phenolic concentration of the extracts Analyzing the data shown in the figure 4, it can be observed that the extracts obtained from the variety "Espada" presented a concentration of phenolics higher than in the extracts obtained with variety Tommy. According to Sharma et al (2001) this variation in total phenolic content was mainly due to genetic differences that exist between the varieties of mango. Figure 5 compares the concentrations of the extracts obtained from the two peel drying times used.The arithmetic mean of phenolic concentrations obtained in the tests where we used only the drying time of 15 hours (1,2,5,6,9,10,13 and 14), and the average of the tests where we used only the drying time of 24 hours (3,4,7,8,11,12, 15 and 16), were compared. Thus, the effect drying time on extractability of phenolics was also evaluated. On average 212.9 ppm phenolic were found in extracts from mango peels after 15 hrs of drying, while only 180.9 ppm phenolics after 24 hrs.

Phenolics Concentration (ppm)

200

Espada
100

Tom my

Figure 3 - Effect of ethanol concentration on phenolic concentration of the extracts. The figure is showing that ethanol-water is a better solvent for extraction of phenolics from mango peel than pure water. So we can conclude that the use of ethanol solution results in a higher yield of extraction. Rangkadilok et al (2005) also attributed this behavior to the variation of solubility of different compounds in water. According to these authors, the use of alcohol as a solvent can increase the solubility of poorly soluble compounds in water. One can also relate this reference to the findings of Prabha and Patwardhan (1986) [cited by Berardini et al, (2005)] who claimed that the ellagic acid compound, a slightly soluble in water, is present in greater proportion in mango peels.

300
Phenolics Concentration (ppm)

E.2. The optimization Considering the results discussed later, that the variety Espada and using the ethanol-water mixture resulted in higher yield of extraction. Moreover, that was also verified that increasing the peel drying time reduced the yield of the extraction, new experiments were performed. It was fitted a second design experiment using only variety Espada peels. It was taken a star planning around the center point, defined by drying time of 10 hours, and ethanol concentration of 70%. To obtain the star planning it was used the experimental design showed in Table 6. Table 6: Structure and results of star planning TestEthanol Concentration (%)Peel drying time (h)Yield (%)160829,7280829,2801229,9601229,7701034,47010 33,0701033,7561030,27012,829,8841030,9707,229,7 With the results of star planning showed in the table 6 and using the software Statistica was obtained the response surface and their respective contour plots showed in Figure 6.

200

100

15 hours 24 hours

Figure 5 - Effect of peel drying time on phenolic concentration As seen, we can conclude that the increase in drying time results in decreasing of the extracts phenolic concentration and therefore the yield of extraction. The explanation of this effect may be that the prolonged thermal treatment caused the decomposition of some compounds, however there were no references in the literature to justify this fact. E.1.2. The effects of interactions between the parameters The effects of interactions between the parameters were obtained from 6 combinations of parameters: Mango variety versus peel drying time Mango variety versus temperature of the solvent Mango variety versus ethanol concentration Peel drying time versus temperature of the solvent Peel drying time versus ethanol concentration Temperature of the solvent versus ethanol concentration According to Table 5, where is showed the effects of interaction between parameters, it is clear that only the interaction between mango variety/ethanol concentration had a significant effect on the extraction process, while the others had no significant effect (p ? 0.05). Table 5: Analysis of variance showing significance of interaction effects between the parameters DryingInteraction betweenSolvent 7.0000.611627 Drying timeSolvent ConcentrationConcentration Variety Solvent Temperature 41.7500.005754 Variety Temperature parametersEffectp Solvent Solvent Concentration -5.6250.682931 VarietySolvent Temperature18.6250.184713 time - Drying time 4.2500.757407 12.6250.363122

Figure 6 - Response surface and contour plots showing the effect of ethanol

concentration and drying time on the extraction yield of phenolic compounds. Looking at figure 6 we can see that the maximum yield was reached 33, 7% and this can be obtained from an ethanol concentration between 65% and 75%, and the drying time between 9 and 11 hours. Concerning to the concentration of the solvent, the results are in agreement with the literature, which claim that the best point was in a range of solvent concentration tbetween 60% to 80%. It can be cited a study published by George (2005) that performed the optimization of Folin Ciocalteau method in phenolic extraction in fruit juice. The author observed that the best solvent was 70%ethanol-water solution. .Mukhopadyan (2006, apud Luthria, 2006) found that the 60% methanol-water solution was more effective in extracting phenolic compounds than pure solvents acetone, methanol and ethanol when used in Cimicifuga racemosa. Along the same lines, Keinnen (1990), apud Lima, (2004) also noted that both 80% methanol-water and ethanol-water are also more efficient in the extraction of phenolic compounds in leaves of Betula pendula. III. CONCLUSIONS The results obtained in this study suggest that the parameters ethanol concentration and peel drying time directly influence the yield extraction of phenolic compounds in mango peel. Furthermore, mango peels of the variety 'Espada' had higher total phenolic content than the variety 'Tommy Atkins'. With respect to the solvent used in extraction, ethanol-water solution proved to be much more effective than pure water. Also was found that increasing peel drying time decreases the extraction yield. Finally the highest empirical yield of extraction was 33.7% based on dry matter, obtained from the variety "Espada" as follows: ethanol concentration between 65% and 75% and drying time between 9 and 11 hours. lREFERENCES Berardini, N,,Fezer R.,Conrad J., Beifuss, U., Carle, R. Schieber, Andras. Screening of Mango (Mangifera indica L.) Cultivars for their contents of flavonol Agric. Food Chem. 53 (5). 1563-1570. (2005). Georg, S., Brat, P., Alter, P, Amiot, M . Rapid determination of poliphenols and vitamin C in plantderived products Agric. Food Chem.. 53 (5). 1370-1373 (2005). Lima, V.L.A.G., Mlo, E.A., Maciel, M.I.S., Silva, G.S.B., Lima, D.E.S. Fenlicos totais e atividade antioxidante do extrato aquoso de broto de feijomungo (Vigna radiata L.) Rev. Nut. Luthria, D.L. Influence of sample preparation on the Test of phytochemicals Beltsville. (2006)

Mamede, M.E.O., Pastori, G. M. Compostos fenlicos do vinho: Estrutura e ao antioxidante, Curitiba: B. CEPPA:; 22, 2. jul/dez (2004). Manica, I., Oliveira, M.E.J. Principais pases e quantidades de fruta produzidas no mundo TODA FRUTA (2005). Pino, J.M., Munoz, Y., Pilar, M.M., Marbo, R. Volatiles from Mango Cultivars J. Agric. Food Chem,. 53, 6, (2005) Rangkadilok, N., Worasuttayangkurn, L., Bennett, R.N., Satayavivad , J., Identification and quantification of polyphenolic compounds in longan (Euphoria longana Lam.) fruit Agric. Food Chem.. 53, 5, 1387-1392. (2005) Santos, V.C., Hasman, F.R.I.C Extrao de compostos fenlicos de hidrolisado hemicelulsico de palha de arroz empregando termosseparao, Cong. Bras. Eng. Qum. em Inic. Cient. Maring, Paran, Brazil (2002). Sharma, R.R., Singh, C.N. Chhonkar, O.P., Goswami, A.M. Singh, S.K. Polyphenol oxidase activity as an index for screening mango (Mangifera indica L.) germplasm against malformation. Plant Genetic Resources Newsletter FAO. IPGRI. (2005)