Optimization of Processing Parameters for the Extraction of Essential Oil from Orange Rind A.F. Evangelista, N. Narain and R.R. Souzaa Department of Chemical Engineering Federal University of Sergipe University Campus “José Aloísio de Campos” Av. Marechal Rondon, S/N, Rosa Elze 49100-000, São Cristóvão, Sergipe Brazil
J.C.C. Santana School of Chemical Engineering State University of Campinas University Campus “Zeferino Vaz” Av. Albert Einstein, 500 6066 Barão Geraldo 13083-970 Campinas, São Paulo Brazil
Keywords: oil, orange, response surface methodology, yield, standardization Abstract The residues generated in the orange juice processing industries in Brazil are generally donated or sold at an extremely low price, and these are consequently used in animal feed or as organic fertilizer. For this reason, the present work was planned to give a better economic viability to these residues through the extraction of oil which could be used later in food and cosmetic industrial products. A 23 factorial with star design was utilized and the factors such as: solvent concentration, time and temperature of extraction were varied while the process yield was evaluated as response. The factors varied according to the planned experimental matrix and the models were evaluated by ANOVA while for optimization of parameters, response surface methodology (RSM) was employed. A unitary type extractor Foss Tecator of Soxtex was employed while various mixtures of alcohol and water were used as solvent. The extracts were incolored and possessed characteristic orange aroma. Through ANOVA analysis, a square model was adjusted to fit the results. The RSM analysis revealed the yield to be parabolic in nature between solvent concentration and time, and saddle point for temperature. The optimum regions for essential oil extraction pertained to the mean values of solvent concentration and of time but of minimum temperature for extraction. This work demonstrates that it is possible to aggregate value to these residues through the extraction of essential oils which could be used as subsidiary products and commercialized by industries. INTRODUCTION The production of essential oils constitutes an important economic activity as these are widely used in food, cosmetic and pharmaceutical industries. Although the major utilization of these oils happens to be in the food sector (spices, condiments, flavor agents and refrigerants) and cosmetics (perfumes and clean hygienic products), it is also used in pharmaceutical products as vegetable drugs rich in volatile oils are employed in nature for the preparation of water extraction and/or galenic preparation simples. Furthermore, many volatile oils are utilized in function of its therapeutic properties and for aromatization of pharmaceutical oral drugs (Boss, 2000). The conventional method for obtaining essential oils is extraction with organic solvents such as hexane. The problems encountered with this method are that it is timeconsuming and requires treatment at high temperature to remove the remaining solvent in the product. These characteristics could affect the organoleptic food quality (Lamera et al., 1997; Mendes et al., 1997). There are several extraction methods and these vary according to the localization of essential oils in the plant. The common methods are: extraction with solvent, serving as carrier water vapors, utilization of supercritical CO2, pressing procedures or by efloration (Boland et al., 1991; Lamera et al., 1997; Mendes et al., 1997; Penedo and Coelho, 1997). a
[email protected]
Proc. IIIrd IS on Trop. and Subtrop. Fruits Eds.: M. Souza and R. Drew Acta Hort. 864, ISHS 2010
479
The extraction with organic solvent is one of the most efficient procedures within the various methods cited. However, the product cost could vary according to the solvent used, energy expenses and on price of the raw material used as essential oil source (Penedo and Coelho, 1997). Taking into consideration that Brazil is one of the major world producers of citrus fruit juice, it is important to look for a detailed study on commercial exploitation of this sector. Since the industrial residues are currently being utilized as organic fertilizer or as an animal feed, the extraction of oils would be one of the promising sources for its development which will consequently improve the quality of life of the agricultural producers, generating employment within the community besides aggregating value to these materials in citrus producing regions. MATERIALS AND METHODS Materials The orange rinds utilized in this study were obtained from citrus processing industry. A pretreatment was given to residue consisting after a previous treatment. Among the solvents utilized were solutions of water/alcohol in varying concentrations according to the conditions presented in planning matrix for experiments (Table 1). Methods 1. Liquid-Solid Extraction. A unitary type of extractor Foss Tecator – Soxtex, which had six extraction units and provision for temperature control, was used. Extraction time was varied in experiments. Six samples of approximately 2 g of orange rind taken in extraction thumbles and these were extracted with 6 fractions of 100 ml of solvent, according to the planned proportions of these experiments design. Once the extraction was concluded, the essential oils were collected, concentrated in rotary evaporator under reduced pressure and stored in amber colored vials at 5°C (Araújo, 2001). 2. Planning and Optimization of Experiments. A 23 factorial planning was realized with star design; consisting of alcohol concentration (CAlcohol), extraction time (t) and extraction temperature (T) as factors which could influence on the process yields as shown in Table 1. The variables codification was as follows: x1 = (CAlcohol – 50)/ 25 (1) x2 = (t-105)/15 (2) x3 = (T-175)/25 (3) Evaluation for adjusting the data to find an appropriate model was done by utilizing the ANOVA methodology while optimization for processing parameters was carried out by Response Surface Methodology (RSM) according to Barros Neto et al. (2001). RESULTS AND DISCUSSION The extracts obtained possessed characteristic orange aroma and these were incolor in appearance, which demonstrated that the properties of these essential oils were retained in the extracts. Table 1 presents the data on various experiments undertaken by selecting the values of x1, x2 and x3. These results reveal the experimental values of extraction yields, obtained from each experimental design. These served as a base for finding regression by minimum square method and thus obtained the models with the objective of posterior evaluation of these by ANOVA methodology, which is demonstrated as follows: The Eq. 4, which is presented below, describes the optimum model for the evaluation of dependence on extraction yield (Y) under effects of factors of time (t) and temperature (T) of extraction. It is perceived that the dependence is of square type with both the factors. Y=5.3300-0.0513CAlcohol+0.0756t-0.1076T-0.0654CAlcohol2-0.2318t2....+ 480
(4)
...+0.1621T2 +0.0176 CAlcohol. t+0.0174 CAlcohol.T -0.0114 t.T The results presented in Table 2 were obtained by application of ANOVA for the optimum model. In these are presented the values of analysis of variance explainable, maximum explainable and multiple correlations (R2). How close these values be of 100 (1º and 2º) and 1.0 (3º), small will be the error accumulated in the model. It is perceived that the values obtained for these cited parameters fall within the expected. Thus from these concepts, we can conclude that the model presents low error values due to the (variances and R2) and low errors due to the methods employed in analysis. The two last columns in the Table 2 present F test values, being that the first (Fcalc/Ftab) indicates that the model is significative, or the data calculated approximates to the experimental values while the second (Ftab/Fcalc) indicates that the data are adjusted and describes well the surface response. For the first test to be valid, it is necessary that the value of Fcalc should be four times the values of Ftab (Barros Neto et al., 2001). This condition approximates the values thus demonstrating that the model is significant within the scope of this work. In the case of the second test, the condition is inversely proportional or Fcalc has to be four times lower than Ftab (Barros Neto et al., 2001). However, for the model, we observe that Fcalc value is two times lower than the Ftab, which signifies that the data could not describe the surface, although the model being better adjusted to the process (Barros Neto et al., 2001). Base on Eq. 4, we could obtain the Figures 1, 2 and 3 which present the response surfaces to facilitate the process optimization. Figure 1 reveals that the extraction yield is higher in the manner that the solvent concentration and the extraction time meet at a region near to their mean values. In Figure 2, there is a concurrence of the results obtained in Figure 1, as the factor solvent concentration should be in the medium range in order that the yield will be higher, but the temperature presents two optimum regions, both in the extreme of the curve. By undertaking the analysis as shown in Figure 3, we arrive at the complete optimization of the process, as it is perceived that by maintaining the extraction time in the range close to mean value and by maintaining the extraction temperature at extreme regions, we arrive at the maximum yield in extraction of essential oils under the studied conditions. However, in higher temperatures, there is more energy consumption and consequently, the cost becoming higher in obtaining the product. Thus the range of values obtained at lower temperature due to economic reasons would be the better option utilized in this process. It could hence be stated for the conditions studied in these experiments that the optimum region for the extraction of essential oils from the residues of orange rinds by solid-liquid extraction is that which tends to be close to the mean values of CAlcohol (50% of alcohol) and t (105 min of process), and lower values of T (175°C), once that the increase in the last parameter will elevate the product cost. CONCLUSION The product obtained had pleasing aroma, color and appearance characteristic to the essential oils from orange rinds, showing that the extraction method was efficient in the retention of principal characteristics of this product. The solid-liquid extraction process of essential oils from orange rinds must be done under conditions of utilizing a mixture of 50% alcohol in water having a heat treatment at 175°C for 105 min so that the yield could be maximum and a quality product could be obtained. This work shows that it is possible to aggregate value by utilizing residues from citrus juice processing industries, generating a new source of economic development which could improve the quality of life in agricultural producers, creation of new jobs and income for these communities. ACKNOWLEDGEMENTS Authors thank to National Consul for Science and Technologic Development 481
(CNPq) and to the FundoVerde-Amarelo (convention FINEP-EMDAGRO), from Brazil, for the financial supports received. Literature Cited Araújo, J.M.A. 2001. Química de alimentos. Teoria e prática. Vol.1, 2nd edition, EDUFV, Viçosa, MG, Brazil. Barros Neto, B., Scarminio, I. S. and Bruns, R.E. 2001. Como Fazer Experimentos: Pesquisa e Desenvolvimento na Ciência e na Indústria. Coleção Livros-Textos, Vol. 1, 1st edition, EDUNICAMP, Campinas, SP, Brazil. Boland, D.J., Brophy, J.J. and House, A.P.N. 1991. Eucalyptus leaf oils: use, chemistry, distillation and marketing. INKATA/ACIAR/CSIRO, Melbourne. Boss, E.A. 2000. Análise do desempenho de plantas de extração de óleos convencionais e de processos supercríticos. School of Chemical Engineering, State University of Campinas, Campinas, Sp, Brazil. (Diss.) Lamera, A.C.P., Coelho, G.L.V. and Mothe C.G. 1997. Extração de lipídeos da amêndoa de castanha de caju com CO2 supercrítico. Ciência e Tecnologia de Alimentos 17(4):405–407. Mendes, M., Oliveira, J.V., Uller, A. 1997. Fracionamento de óleos de citros utilizando fluidos supercríticos. Ciência e Tecnologia de Alimentos 17(4):441–445. Penedo, P.L.M. and Coelho, G.L.V. 1997. Purificação de óleos vegetais por extração com CO2 supercrítico. Ciência e Tecnologia de Alimentos 17(4):380-383.
Tables Table 1. Planned matrix for experimental design. Assay 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Yex x1 x2 x3 Ycalc 5.3018 -1 -1 -1 5.3440 1 -1 -1 5.1100 5.1292 -1 1 -1 5.4420 5.4406 1 1 -1 5.3940 5.3384 -1 -1 1 5.0560 5.0746 1 -1 1 5.0070 4.9716 -1 1 1 5.2240 5.1678 1 1 1 5.1300 5.1352 0 0 0 5.3260 5.33 0 0 0 5.3760 5.33 0 0 0 5.2970 5.33 -1.682 0 0 5.2010 5.2313 0 -1,682 0 4.5410 4.547 0 0 -1,682 5.9400 5.9696 1.682 0 0 5.0370 5.0587 0 1,682 0 4.7550 4.8014 0 0 1,682 5.5850 5.6076
x1, x2 and x3 are coded variables for alcohol concentration, extraction time and extraction temperature, respectively.
482
Table 2. Variance analysis for obtaining the optimum empirical model. Variation source
Square sum
Regression 1.5782 Residual 0.0182 Lack of fitting 0.0150 Error 0.0032 Total 1.5916 % variation explained % maximum variation explained Determining coefficient (R2)
Freedom degree 9 7 5 2 16
Square means 0.1754 0.0026 0.0030 0.0016
Fcalc
Ftab
67.397
3.68
1.881
19.3
99.159 99.799 0.9916
Fcal and Ftab are, respectively, the calculated and tabled Test F.
Figures
3,956 4,094 4,233 4,371 4,51 4,648 4,787 4,925 5,064 5,202 above
Fig. 1. Effect of extraction time and solvent concentration on the extraction yield.
483
5,073 5,188 5,303 5,418 5,532 5,647 5,762 5,877 5,992 6,107 above
Fig. 2. Effect of temperature and solvent concentration on the extraction yield.
4,419 4,597 4,776 4,954 5,133 5,311 5,489 5,668 5,846 6,024 above
Fig. 3. Effect of extraction time and of solvent concentration on extraction yield.
484
