Statistical modelling of a facile process for the extraction of crude constituents of Curcuma longa.
Ethno medicinal values of Curcuma longa have been recognised since pre-historic times. It is used in ethno medicinal formulations for the treatment of spectrum of diseases notably heart disorder, liver problems, asthma, arthritis, gall bladder infections, digestive disorders and dysmenorrhoea (Adeniji, 2003). The herb is endowed with a colourant of superior tinctorial strength in comparison to synthetic dye of similar shade e.g. tartrazine (Henry, 1996) therefore, prefered to use in colouring delicasies such as soup and puddies. The rhizome is a herb of domestic commerce in western Nigeria and could contribute to import substitution therefore, such goal should be prioritised. Enhancing agronomical value essentially involves exploiting upstream and downstream opportunities that abound in agricultural produce. At the entry point of the opportunity chain lies feed stock, a product of primary process usually characterised with low technology process but less effective. It also reduces packaging and transportation cost in comparison to the large sized rhizomes. Using high technology, such feed stock could be processed to yield speciality natural products in nature of dietary colourant or herbal medicine principally inherentto C. longa. Catalogue of bio active principle and prospective pharmacological consequence of C. longa have been given by Henry (1996). Added to this, the active components of C. longa have been modified to enhance its antifungal, antibacterial (Misra et al., 2007) and anticancerous properties (Rojsitthisak et al., 2011). Crude extracts which serve as feed stock for colourant and bioactive components inherent in C. longa are of different chemical groups. Therefore, the components could be preliminarily separated using appropriate solvent system enhanced by physical variables such as time and temperature.
The aim of this study was to model a facile process using temperature, contact time and solvent system as independent variables for the extraction of crude constituents of C. longa with emphasis on dietary colourant and therapeutic components which could serve as feedstock for the production of specialized natural product with therapeutic (antibiotics) or colourant specificities consequently lending enhancement of agronomical value of C. longa.
Materials and Methods
Materials. Wholesome Curcuma longa rhizome (red ginger) fingers were obtained from central commercial market in Ado-Ekiti, Nigeria. All chemicals used were of analytical grade: ethanol and distilled water.
Extraction protocol. Curcuma longa rhizome without blemish was washed, peeled and dried. The dried product was milled and used in subsequent crude constituent extraction process. The schematic diagram for the selected bioactive constituent recovery process using two solvents mixture system, contact time, medium temperature as extraction independent variables is shown in Fig. 1. Details of extraction variables in the experimental design are presented in Table 1. The lowest and highest levels of independent variables were chosen from the results of preliminary investigations.
Experimental design. A central composite rotatable design for k=3 was used (Cochran and Cox, 1957). The 3-factor, 5-level design generated 20 sample combina tions comprising eight points peculiar to [2.sup.3] factorial, six star points and six central points for replication. The effects of independent variables namely: solvent-mixture ratio, temperature and contact time were noted for extraction of the crude constituents. The crude constituents markers namely: total phenolic content, tinctorial (colour intensity) index and relative total soluble solids of solution were evaluated. Step-wise regression analyses were performed on the data to yield equations for predicting extraction of crude constituents of C. longa with reference to designated functionality.
Analytical Methods. Relative total soluble solids determination. Refractive index of sample was measured using Abbey Refractometer (ABBE 325, ZUZI) and corresponding soluble solids was determined using a procedure adapted from sugar analysis method and reference to appropriate designated table. Result is reported as relative total soluble solids (Table 2).
Determination of colour density/polymeric colour. Colour density was determined according to the method described by Wrolstrad et al. (1982). Colour density was calculated as the sum of the absorbances at 420 nm and 510rrm.
Evaluation of total phenolic content. Total phenolic content was evaluated according to the method described by Taga et al. (1984). Briefly, a 100jj.L of Folin - Ciocalteau reagent (2N wrt acid Fluka Chemic AG - Ch-9470 BUCHS) was added to each sample (20 [micro]L) and well mixed after addition of 1.58 mL of water. After 30s, 300 [micro]L of 2% sodium carbonate solution was added and the sample tubes were left at room temperature for 2 h. The absorbance(A) of the developed blue colour was measured at 750 nm using Unicam Helios & uv/vis/spectrophotometer. A plot of A750 nm against corresponding concentration was used to calculate phenolic content using ascorbic acid as standard and result expressed in mg ascorbic acid equivalent/g sample.
Statistical analysis. The central composite orthogonal designed was analysed as reported by Cochran and Cox (1957). Each of the X-matrix was multiplied by the Y-column (response) to obtain corresponding sums of products that is 0y to 13y for [X.sub.0] to [X.sub.1] [X.sub.3].Consequently, the coefficients [b.sub.0] to [b.sub.13] were calculate as:
[b.sub.0] = 0.166338(0y)-0.056791 [SIGMA] (iiy) (1)
bi = 0.073224(iy) (2)
bii = 0.062500(iiy) + 0.006889 [SIGMA] (iiy) -0.056791(Oy) (3)
bij = 0.125000(ijy) (4)
The quadratic model was fitted using the regression coefficients and the predicted response calculated for each of the observed values. The model was observed for adequacy by subject to analysis of variance and residual analysis.
Results and Discussion
Literature suggests that active principle of C. longa is insoluble in water. However, it is worth noting that bioactive components of natural colours do not occur in isolation in nature, but usually glycosylated (Oszmianski et al., 2004).Hence, glycosylation is suggested to be the principal factor accounting for aqueous solubility of active constituents of C. longa as instantaneous colouration of hot water appeares on addition of C. longa powder. Thus, water is used as one of the solvents in this study due to ready availability and relatively low cost. Basically three technological parameters namely; phenolic content, tinctorial index and relative total soluble solids were evaluated for the extraction of crude constituents of C. longa. The constraints (Table 1) explored in this study, exerted varied extraction effects as shown by the three responses (Table 2) evaluated therefore, subsequently modeled for prediction of constituent extraction outcome that is dictated by constituent's functionality.
The central composite orthogonal design to fit the polynomial model for the extraction of the three classes of crude constituents in C. longa was accomplished as elicited by Cochran and Cox (1957). The computed sums of products and regression coefficients to fit the model are shown in Table 3.
Total phenolic content. One of the most important groups of natural product with respect to therapeutic or biological value is phenolics. Therefore, total phenolic content of the extracted crude constituents in the C. longa solution can give insight to therapeutic potential of the designated rhizome. Thus the quadratic model takes the form:
[P.sub.t] = 0.03549 + 0.00928[X.su.1] + 0.00388[X.sub.2] + 0.00715[X.sub.3]-0.00237[X1.sup.2]-0.00766[X2.sup.2]- 0.00731[X3.sup.2]-0.00063X1X2-0.00013X1X3-0.00063X2X3 (5)
The predicted total phenolic content pt for each of the experimental run and their respective residual are shown in Table 4. Examination of the residuals suggests that the fitted model was reasonably adequate. The assertion was confirmed on model testing. Added to this, the analysis of variance to test the fitness of the model is presented in Table 5. The first and second order terms were significant as shown by the higher calculated F-ratio in comparison with the tabulated values. However, since the calculated F-ratio for the lack of fit was lower than the tabulated value, adequacy of the fitted model is affirmed. One of the flexibilities affordable by this model is that it can be adapted for extraction of crude constituent of interest in terms of dietary colourant or bioactive (therapeutic) dominate.
Tinctorial index (Tp). Poor tinctorial index is the foremost drawback of colourant of natural origin. However, it is gratifying to note that, such negativity is not the case with shade intensity of C. longa. Colourimetric power (index) of C. longa is higher than available synthetic colourants of similar shade (Henry, 1996). Therefore, predictive model for tinctorial index of crude extract of colourant from C. longa deserves determination. Thus the quadratic model takes the form:
[T.sub.I] = 0.36412498 + 0.1173268[X.sub.1] + 0.03195495[X.sub.2] + 0.0449698[X.sub.3] - 0.019149[X.sup.2.sub.1] - 0.059799[X.sup.2.sub.2] - 0.054499[X.sup.2.sub.3] + 0.0025[X.sub.1][X.sub.2] + 0.025[X.sub.1][X.sub.3] - 0.025[X.sub.2][X.sub.3] (6)
The predicted tinctorial index ([T.sub.I]) for each of the experimental run and their residual are presented in Table 4. Their residuals suggest that the fitted model is adequate. The claim on adequacy of the model fitness was verified by conducting analysis of variance test (Table 5). Since the first and second order terms were significant as revealed by higher calculated F-ratio in comparison with tabulated values and the calculated F-ratio for the lack of fit were lower than the tabulated value, attesting adequacy of the model.
Relative total soluble solids (RTSS). Relative total soluble solids reflect a composite of colourant of the crude constituents of C. longa. Its value is principally dictated by solvent composition (mixture) and bio chemical nature of the crude constituents (i.e. colourant and phenolic compounds). Nevertheless, the other variables influence RTSS of C. longa constituents in solution. The quadratic model takes the form:
RTSS = 18.1668 + 4.02439[X.sub.1] + 0.642[X.sub.2] - 0.389[X.sub.3] - 1.63[X.sup.1.sub.2] - 0.215[X.sup.2.sub.2] - 0.215[X.sup.2.sub.3] + 0.125[X.sub.1][X.sub.2] + 0.875[X.sub.1][X.sub.3] - 0.125[X.sub.2][X.sub.3] (7)
Using same assessment protocol reported for previous model parameters, the fitness of RTSS model was evaluated using analysis of variance (Table 4) and residual analysis (Table 5) and found adequate.
Extraction of crude constituents of Curcuma longa using three process variables yielded constituent that could essentially be dietary colourant and antimicrobial bases. The models developed were found adequate to extract crude constituent of interest with respect to colourant or antibiotics from the C. longa powder. In addition, the preliminarily separated crude constituents could serve as feedstock for further processing to yield product of high functionality. Bearing in mind that C. longa is partly plant food, its constituents can be used as additive in food preparations that are subjected to heat but not necessarily exposed to light. Such foods are: bread, cake, potatoes, mayonnaise, rice, macaroni and frankfurters. The findings of this study also offer possibility for reducing transportation and packaging cost as a result of miniature volume of the crude constituents in comparison to the large size (bulk) of the milled rhizome.
Adeniji, M.O. 2003. Herbal Treatment of Human Diseases. pp. 23-42, Oynx International (Nig) Ltd, Ibadan, Nigeria.
Cochran, W.G., Cox, G.M. 1957. Experimental Design, 2nd edition, pp. 335-357, John Wiley and Sons, New York., USA.
Henry, B.S. 1996. G.A.F. Henry, J.O. Houghton, (eds.). Natural Food Colourants, 2nd edition, pp. 40-79, Blackie Academic and Professional-an Imprint of Chapman and Hall, London, UK.
Misra, K., Narain, U., Kapoor, N. 2007. Bio-active conjugates of curcumin having ester, peptide, thiol and disulfide links. The Journals Scientific and Industrial Research, 66: 647-650.
Oszmianski, J., Bakowska, A., Piacente, S. 2004. Thermodynamic characteristics of copigmentation, reaction of acylated anthocyanin isolated from blue flowers of Scutellaria baicalensis Georgi with copigments. The journal of the Science of Food and Agriculture, 84: 1500-1506.
Rojsitthisak, P., Wichitnithad, W., Nimmannit, U., Wacharasindhu, S.L. 2011. Synthesis, characterization and biological evaluation succinate prodrugs of curcuminoids for colon cancer treatment. Molecules, 16: 1888-1900.
Taga, M.S., Miller, E.E., Pratt, D.E.1984. Chia seeds as a source of natural lipid antioxidants. Journal of The American Oil and Chemical Society, 61: 928-932.
Wrolstrad, R.E., Culbertson, J.D., Cornwell, C.J., Mattick, L.R. 1982. Detection of adulleration in blackberry juice concentrate and wines. Journal of Association of Official Analytical Chemists, 65: 1417-1423.
Department of Food Technology Federal Polytechnic, PMB 5351, Ado-Ekiti, Ekiti State, Nigeria
(received February 25, 2012; revised November 12, 2012; accepted February 14, 2013)
* Author for correspondence; E-mail: email@example.com
Table 1. Process variables used in the central composite rotatable design (k = 3) levels Independent variable Code -1.682 -1 0 1 1.682 Time (min) [X.sub.1] 5 10 15 20 30 Temp ([degrees]C) [X.sub.2] 30 40 50 60 70 Solvent(water/ [X.sub.3] 1:0 2:1 1:1 1:2 0:1 ethanol) * * 0.5g of C. longa in100 mL solvent mixture Table 2. Range of values for total phenolic content, tinctorial index and relative total soluble solids Model parameter Range value Total phenolic content mg/g 0.001-0.051 Tinctorial index 0.03-0.45 Relative total soluble solids (%) 4-23 - Table 3. Regression coefficients for the quadratic model equation for extraction of crude constituent of C. longa Sum of Phenolic content Relative total Tinctometric products soluble solids property 0y 0.473 335 5.46 y 0.1267 54.96 1.6023 2y 0.053 8.77 0.4364 3y 0.0976 -5.31 0.61414 11y 0.308 205.35 3.5988 32y 0.223 227.98 2.9484 33y 0.229 227.98 3.0332 12y -0.005 1.00 0.02 13y -0.001 7.00 0.2 23y -0.005 -1.00 -0.2 [SIGMA] (iiy) 0.760 661.31 9.5804 Sum of Regression Phenolic content Relative total products coefficients soluble solid 0y b0 0.03549 18.1668 y b1 0.00928 4.02439 2y b2 0.00388 0.642 3y b3 0.007150 -0.389 11y b11 -0.00237 -1.63 32y b22 -0.00766 -0.215 33y b33 -0.00731 -0.125 12y b12 -0.00063 0.125 13y b13 0.00013 0.875 23y b23 -0.00063 -0.125 [SIGMA] (iiy) Sum of Tinctometric products property 0y 0.36412 y 0.11732 2y 0.03195 3y 0.04496 11y -0.01914 32y -0.05979 33y -0.05449 12y 0.00250 13y 0.02500 23y -0.02500 [SIGMA] (iiy) Table 4. Analysis of Variance (ANOVA) for the Predictive Model Equations Independent Statistical DF SS MS variable term Phenolic content First order 3 0.00207953 0.000693176 Second order 6 0.00149 0.0002483 Lack of fit 5 0.0002885 0.0000577 Error 5 0.0001635 0.0000327 Total 19 Tinctorial index First order 3 0.2295556 0.0765185' Second order 6 0.097061 0.01617684 Lack of fit 5 0.0401 0.00802 Error 5 0.0127 0..00254 Total 19 Relative total First order 3 228.872 76.291 soluble solids Second order 6 47.9911 7.998 Lack of fit 5 28.06 5.612 Error 5 6.831 1.366 Total 19 Independent F-Ratio Tabulated variable calculated 5% Phenolic content 21.1198 5.41 7.5933 4.95 1.76 5.05 Tinctorial index 30.125 5.41 6.3688 4.95 3.15748 5.05 Relative total 55.8497 5.41 soluble solids 5.857 4.95 4.1083 5.05 DF = Degree of Freedom; SS = Sum of Square; MS = Mega Square Table 5. Residual analysis of assessed parameters Total Phenolic content Expt Observed Predicted Residual run 1 0.003 -0.00355 0.00655 2 0.015 0.01653 -0.00153 3 0.006 0.00673 -0.00073 4 0.02 0.02429 0.00429 5 0.017 0.01227 0.00473 6 0.033 0.03183 0.00117 7 0.022 0.02003 0.00197 8 0.031 0.03707 -0.00607 9 0.006 0.013179 -0.007179 10 0.051 0.0444 0.0066 11 0.001 0.0073 -0.0063 12 0.026 0.02035 0.00565 13 0.003 0.00279 0.00021 14 0.026 0.02684 -0.00084 15 0.030 0.03549 -0.00549 16 0.039 0.03549 0.00351 17 0.032 0.03549 -0.00349 18 0.03 0.03549 -0.00549 19 0.038 0.03549 0.00251 20 0.044 0.03549 0.00851 Relative total soluble solids Expt Observed Predicted Residual run 1 14 12.70441 1.29859 2 18.2 18.75319 -0. 55319 3 16 13.9884 2.0116 4 20 20.53719 -0.53719 5 11 10.42641 0.57359 6 18 19.97519 -1.97519 7 12 13.21041 -1.21041 8 20 21.25919 -1.25919 9 4 6.788136 -2.788136 10 23 20.326184 2.673816 11 17 16.478936 0.521064 12 18 18.6386 -0.6386 13 17 18.213078 -1.213078 14 18 16.90448 1.09552 15 18 18.1668 -0.1668 16 19 18.1668 0.8332 17 20 18.1668 1.8332 18 18 18.1668 -0.1668 19 17 18.1668 -1.1668 20 17 18.1668 -1.1668 Tinctorial index Expt Observed Predicted Residual run 1 0.11 0.038926 - 2 0.30 0.21858 0.08142 3 0.18 0.147836 0.032164 4 0.38 0.337489 0.042511 5 0.15 0.128866 0.021134 6 0.44 0.4085196 0.03148 7 0.12 0.137745 -0.017745 8 0.42 0.427428 -0.00743 9 0.08 0.1126279 -0.03263 10 0.45 0.5073153 -0.057315 11 0.05 0.1412667 -0.091267 12 0.25 0.2487616 0.001238 13 0.03 0.134363 -0.104363 14 0.30 0.285641 0.014359 15 0.35 0.364125 -0.01413 16 0.38 0.364125 0.015875 17 0.30 0.364125 -0.064125 18 0.34 0.364125 -0.024125 19 0.45 0.364125 0.085875 20 0.38 0.364125 0.015875 Residual = Observed--Predicted
|Printer friendly Cite/link Email Feedback|
|Publication:||Pakistan Journal of Scientific and Industrial Research Series B: Biological Sciences|
|Date:||Jul 1, 2013|
|Previous Article:||Quantification and detoxification of aflatoxin in food items.|
|Next Article:||Variation in activity of pepsin extracted from buffalo stomach mucosa.|