Orientador: Prof. Dr. Jesuí Vergílio Visentainer

 Data da Defesa: 17/03/2016



INTRODUCTION. The yellow passion fruit corresponds to 97 % of Brazilian cultivars, and its production has increased more than 90 % between 2005 and 2010, becoming the seeds of the fruit into a considerable residue which is characterized as source of lipids, fibres and phenolic compounds, and it has been commonly exploited for animal feed or for cosmetic industry. Studies report that extracts of the seeds of passion fruit have anti-inflammatory properties, antioxidant, vaso-relaxants, anti-adipogenic, antidiabetic and antitumor, and still present benefits in the synthesis of collagen and reduction of melanogenesis in the skin, probably due to the high content of piceatannol, an analogue stilbene of resveratrol. The exploitation of the plan residues as natural antioxidants source requires the investigation of the best extraction conditions of the compounds that may be performed using the surface response methodology and mathematical simulation techniques such as the simplex algorithm. Furthermore, the vegetable extracts can be microencapsulated for prolonging their active functions, and the spray drying technique é widely used for this purpose. AIMS. The aim of this study was to optimize both the extraction of antioxidants from residual seed of passion fruit and the microencapsulation of the extracted compounds, and apply analytical technics for evaluating the phenolic composition and antioxidant capacity of the hydroalcoholic extracts and microencapsulated, being theses last one also evaluated for its hygroscopic behavior and for the antioxidant stability in function of different conditions of water activity. MATERIAL AND METHODS. Article 1. The residual seeds from the production of frozen pulp of passion fruit were washed and dried at room temperature, and then defatted with hexane to the study of non-lipid portion wich is still poorly investigated. Extracts of the defatted seeds of passion fruit were produced according to the conditions determined in 15 experiments by Doehlert experimental design, to evaluate the effects of the proportion of solids in the extraction (X1, 2.5 % – 25.0 %, w.v-1 ), and the concentrations of ethanol (X2, 20.0 % – 95.0 %, w.v-1 ) and hydrochloric acid (X3, 0.0 – 3.0 %, w.v-1 ) in the extraction solution on the concentration of total phenolic content (TPC) and its phenolic composition, and antioxidant capacity based on the methods of DPPH, ABTS and FRAP. Article 2. The extract obtained with optimized conditions was microencapsulated by spray drying, using a central composite design with 17 experiments to evaluate the effects of the amount of maltodextrin used as wall material (X1, 4,89 – 30,11 %, m v-1 ), inlet air temperature (X2,119,77 ºC – 170,23 ºC) and feed flow (X3, 0,43 L h-1 – 0,77 L h-1 ), on the same responses considered in the optimization of the extraction of antioxidants. Article 3. The microencapsulated extract produced with de optimized conditions for spray drying was evaluated for its hygroscopic behavior using moisture sorption isotherms, and its antioxidant capacity and piceatannol concentration was monitored in all the range of water activity considered in the study. The mathematical models obtained via multiple linear regression for each dependent variable was evaluated by ANOVA and their regression coefficients evaluated by t-test, at 5% of significance. The correlation among the responses was 11 measured using Pearson coefficient, which was used to choose the variables for sequential optimization via simplex algorithm, in order to determine the extraction and microencapsulation conditions that resulted in the best responses related to antioxidants. Extractions and microencapsulations with the optimized conditions were performed, and the values of the experimental responses were compared with the predicted values to validate the mathematical models and the optimization procedure. RESULTS AND DISCUSSION. Article 1. For the stage of optimization of extraction of antioxidants from residual passion fruit seed, the mathematical models were adequate (p < 10-3 , R2 > 0.9748) when predicting responses for total phenolic compounds, piceatannol concentration and antioxidant capacity based on methods of DDPH, ABTS and FRAP. Ethanol most affected all responses and piceatannol was strongly sensitive to acid. The highest concentration of TPC was 3134.00 mg GAE 100 g-1 obtained with 2.5% of defatted seed, 62.45% ethanol and 1.61% HCl. For piceatannol, results were 750.46 mg 100 g-1 , 17.39% DS, 66.62% ethanol and 0.00% HCl. Article 2. For the stage of microencapsulation by spray drying of the extract produced with optimized condition on the first stage, the predictive models were significant (1.00 10-6 < p < 1.07 10-3 ; 0.8819 < R 2 adjusted < 0.9978), showing low standard errors and did not presented lack of fit. The responses were most affected by the amount of maltodextrin than inlet air temperature and feed flow. Moisture and piceatannol were the two less correlated responses (r = -0.6861) and they were selected for simplex optimization. When moisture was the most important response, the lowest value of 5.06% (w w-1 ) was obtained with 4.89% (w v-1 ) of maltodextrin, inlet air temperature of 147.22ºC and feed flow of 0.43 L h-1 , leading to piceatannol concentration of 1404.53 mg 100 g-1 . If piceatannol was considered the most important, 1932.53 mg 100 g-1 were reached (3.31 fold higher than in the seeds), but moisture increased to 5.96% (w w-1 ) using 4.89% (w v-1 ) of maltodextrin, inlet air temperature of 119.79ºC and feed flow of 0.70 L h -1 . The simplex algorithm allowed selecting different extraction or microencapsulation conditions depending on the expected response. Article 3. On the stage of the study of the hygroscopic behaviour of the microencapsulated extract from passion fruit seed obtained with the optimized conditions, it was found that the isotherms were classified as GAB isotherms, exhibiting a Type II behavior. Isosteric heat of sorption decreased when the relative humidity increased and critical equilibrium moisture reached 61%. The microencapsulated extracts retained their antioxidant capacities within the evaluated conditions, what suggests that it may be applied in food at any water activity value. CONCLUSION. The experimental designs applied for evaluating the process variables related to the extraction of the antioxidants from residual passion fruit seed provided satisfactory predictive models for total phenolic compounds, piceatannol and antioxidant capacity. The simplex algorithm was flexible and reliable on multiple and simultaneous optimization of the responses of hydroalcoholic extraction and microencapsulation by spray drying. KEY WORDS: Passiflora edulis seed, phenolic compounds, piceatannol, antioxidant capacity, response surface methodology, simplex algorithm.


Artigos Publicados Vinculados a Tese: