Research Journal of Chemical Sciences ______________________________________________ ISSN 2231-606X Vol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 18 A Novel Kinetic Assay for the Examination of Solid-Liquid Extraction of Flavonoids from Plant Material Makris D.P. School of Environment, University of the Aegean, Mitr. Ioakim Street, Myrina, 81400, Lemnos, GREECEAvailable online at: www.isca.in, www.isca.me Received 1st October 2015, revised 20th October 2015, accepted 15th November 2015 AbstractThe medicinal plant Artemisia inculta, endemic to the island of Crete (southern Greece), was chosen as a matrix for the examination of solid-liquid extraction of total flavonoids, using a novel kinetic model based on a rectangular 2-parameter hyberbola function, which was established employing non-linear regression. The determination of basic kinetic parameters and the critical comparison with the well-studied and widely used second-order model, proved unequivocally that the model proposed can be implemented as an alternative means of describing solid-liquid extraction kinetics, with high reliability. The advantage over other models used in similar processes is the simplicity of the determination of the basic kinetic parameters, by utilising a “double reciprocal” linear mathematical transformation Keywords: Flavonoids, kinetics, solid-liquid extraction. Introduction The solid-liquid extraction processes usually follow a kinetic curve which illustrates that the rate of mass transfer is not constant. The extraction curves usually consist of three distinct phases: constant extraction rate period, falling extraction rate period, and diffusion controlled period. During the first period, the easily accessible solute that surrounds the particle passes into the liquid phase with an approximately constant rate. Throughout the second period, the extraction rate decreases rapidly as a result of the decrease of the effective mass transfer area and diffusion becomes the governing phenomenon. Finally, in the last period the easily accessible solute layer is depleted and thus the extraction rate is determined exclusively by the diffusion rate of the solvent into the solid particle and of the solvent and solute from the solid particle to the bulk solvent. This mechanism characterizes a slow stage of the extraction process. The solid-liquid extraction of polyphenolic substances has been dealt with by several examinations, which proposed various empirical mathematical models, with the aim of providing the appropriate practical tools for a reliable description of the extraction process. Conventional extraction methodologies are usually based on Fick’s law derivatives models, as the parameters estimated contain physical meanings which can be used for further interpretations. However, various empirical models have also been developed from the fundamental or adapted models, as they are considered to be more suitable for non-conventional extraction processes (microwave, ultrasound-assisted etc.). Similar models have been used in several studies pertaining to adsorption3-7, which is considered the inverse phenomenon of diffusion. To this purpose, this study was performed to examine the kinetics of batch stirred-tank flavonoid extraction from a flavonoid-rich medicinal herb, Artemisia inculta, using second-order and a single rectangular 2-parameter hyberbola model. The test employed to assess the total flavonoid yield is more specific compared with the widely used Folin-Ciocalteu for total polyphenol determination, thus providing more accurate data regarding the extraction kinetics. The efficiency of each model was checked by comparing parameters including the extraction rate constant, the yield at saturation, the time required for the extraction to enter the regular regime and the initial extraction rate. Kinetics was performed at increasing temperatures, to ascertain that thermal degradation of flavonoids is not implicated, a fact that could complicate the apparent kinetic values and produce misleading results. Material and Methods Chemicals and reagents: Rutin (quercetin 3--rutinoside) was from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Aluminium chloride (AlCl) was from Merck (Darmstad, Germany). Glycerol was from Fisher Scientific (New Jersey, U.S.A.). Plant material: The material used consisted of the aerial parts of Artemisia inculta and it was provided by the Mediterranean Plant Conservation Center (Chania, Crete), where a voucher specimen was deposited (9493 MAIC). Following collection, the tissue was air-dried at room temperature in a dry and dark chamber for 7 days, and then ground to a fine powder (approximate mean particle diameter 1 mm) using a domestic blender (Bosch MMB 112R, Stuttgart, Germany). The ground material was placed in plastic screw-cup tubes at 4°C, until used. Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 19 Batch extraction procedure: The optimised method previously described was used. Amount of 1 g of the pulverised material was added in 100 mL of 90% (w/v) aqueous glycerol in a plastic container and extracted under stirring at 600 rpm, with a teflon-coated magnetic stirrer, for 160 min. Heating was accomplished by an oil-bath placed on a hot plate (Yellow Line MST Basic C, Richmond, VA, U.S.A.), set at 50, 70 and 80±1.5°C. Temperature was controlled by a Yellow Line TC1 thermostat. Sampling was carried out at predetermined intervals (5 - 160 min) and all samples were centrifuged in a table centrifugator (Hermle, Wehingen, Germany) at 10,000 rpm for 10 min and filtered through 0.45-µm syringe filters prior to determinations. Determination of total flavonoid yield (YTFn): A previously published protocol was used, with modifications. A volume of 0.25 mL sample was combined with 0.75 mL AlCl reagent [0.16% (w/v) AlCl in 5% (v/v) acetic acid in methanol] and allowed to react for 30 min, at ambient temperature. The absorbance was recorded at 415 nm (A415) and the total flavonoid concentration (TFn) was estimated from a calibration curve, constructed with rutin (quercetin 3--rutinoside) as standard. Yield in total flavonoids (YTFn), expressed as mg rutin equivalents (RtE) per g of dry weight, was determined using the following equation:  \n  \r (1) Where is the volume of the extraction medium (L) and the dry weight of the plant material (g). Statistical analyses and kinetics: All treatments were performed in duplicate. Determinations were carried out in triplicate and values were averaged. Kinetics was established by non-linear regression between YTFn and . Non-linear and linear regressions were performed with Sigma Plot™ 12.0, at least at a 95% significance level. Results and Discussion Second-order model: The model fitted to the extraction kinetics using non-linear regression between extraction yield in TFn (YTFn,) values and t (figure-1, upper plot), was a modified hyperbola described by the equation: \r (2) For all temperatures tested, fitting was statistically significant (table-1), suggesting that extraction yield as a function of can be adequately predicted by the equation-2. This equation actually describes a second-order extraction kinetics, as previously reported10,11, when the boundary conditions = 0 to and YTP(t) = 0 to YTP(t) are considered:  "\r " (3) Table-1Statistical parameters calculated after non-linear regression between YTFn and values T (°C) Rp Second-order 50 0.9857 0.0001 70 0.9964 0.0001 80 0.9938 0.0001 Double reciprocal 50 0.9857 0.0001 70 0.9964 0.0001 80 0.9938 0.0001 TFn(s) and are the TFn yield at saturation and the extraction rate constant, respectively. Transformation of equation-3 yields the following linear expression: # "  $  (4) When approaches 0, the initial extraction rate, , given as TFn(, may be expressed as: %&'(5) If /YTFn() is plotted against , there would be a straight line with a slope = 1/YTFn(s) and intercept = 1/ (figure-2, upper plot). Therefore, for every extraction performed, YTFn(s), and could be determined graphically. In all cases examined, the correlations between /YTFn( and were very high and statistically significant (table-2). The course of extraction, from fast to slow, occurs when approximately 80% of the extraction efficiency has been reached12. This period is called “regular regime” and characterises the second, slow phase of the extraction, where small increases in the extraction rate are recorded within a significant period of time. The time required for the extraction to achieve the regular regime, , can be determined as follows13: = "  (6) Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 20 Figure-1 Non-linear regression between YTFn and , during extraction of TFn from A. inculta, using 90% (w/v) glycerol. Extractions were carried out under continuous stirring at 600 rpm, with RL/S = 100 mL g-1. Upper and lower plot display model fitting according to the equation-2 and 7, respectively (min) 020406080100120140160180 TFn (mg RtE g-1 dw) 1015202530 50 70 80 Fitted model (min) 020406080100120140160180 TFn (mg RtE g-1 dw) 1015202530 50 70 80 Fitted model Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 21 Figure-2 Second-order (upper plot) and double reciprocal (lower plot) kinetic model of TFn extraction from A. inculta using 90% (w/v) glycerol. Extractions were carried out under continuous stirring at 600 rpm, with RL/S = 100 mL g-1 (min) 020406080100120140160180 t / YTFn (min g mg-1 101214 50 70 80 Fitted line 1/ (min-1 0.000.020.040.060.080.100.12 1/YTFn (g mg-1 0.00.10.20.30.40.5 50 70 80 Fitted line Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 22 Table-2 Statistical parameters calculated after linear regression between /YTFn and values (second-order model) or1/YTFn and 1/ values (double reciprocal) T (°C) Rp Second-order 50 0.9971 0.0001 70 0.9991 0.0001 80 0.9998 0.0001 Double reciprocal 50 0.9716 0.0001 70 0.9960 0.0001 80 0.9944 0.0001 Double reciprocal: In this case fitting of the data obtained was performed using single rectangular 2-parameter hyberbola, described by the following equation:  (7) As for the second-order model, fitting was highly significant, giving identical statistical parameters (table-1). Based on the previous considerations for the second-order model, equation-7 would correspond to a kinetic model, as given below:  )*+ (8) The term of the equation-7 can be assigned to a parameter termed as 0.5, as it should be called from now and on, because it has dimensions of time and when = 0.5, then  . Therefore, 0.5 represents the time where 50% of the YTFn(s) has been attained. The linearized expression of equation-8 would be: # )*+  \r $  (9) Equation-9 dictates that a plot of 1/YTFn as a function of 1/would yield a straight line (figure-2, lower plot), with the intercept on the -axis being equal with 1/YTFn(s) and the intercept on the -axis with -1/0.5. Thus both t0.5 and YTFn(s) can be estimated graphically. The initial rate of the extraction, , and the second-order extraction rate, , could then be determined as follows: % )*+ (10) &)*+ (11) Table-3 Displays the kinetic parameters, as determined with the two models. It can be seen that in all cases, the two models provided almost identical values. This outcome was unequivocal proof that the alternative approach proposed can describe the extraction kinetics adequately and with high reliability, as judged by comparison with the second-order model. Table-3 Kinetic parameters of the extraction process, calculated after implementation of the two kinetic modelsT (°C) Kinetic parameters (× 10-3) (g mg-1 min-1) H (mg g-1 min-1) R (min) YTFn(s) (mg RtE g-1) Second-order 50 0.750 0.243 74.07 18.00 70 1.140 0.678 35.97 24.39 80 2.967 2.035 12.87 26.19 Double reciprocal 50 0.749 0.243 74.08 18.02 70 1.144 0.678 35.92 24.34 80 2.970 2.036 12.86 26.18 Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(11), 18-23, November (2015) Res. J. Chem. Sci. International Science Congress Association 23 Conclusion The kinetics of the solid-liquid extraction of total flavonoids from A. inculta was determined using second-order model, but also a novel approach based on a rectangular 2-parameter hyberbola function. The kinetic parameters determined with the two models were identical, a fact that proved the validity of the alternative model proposed. The simple mathematical transformation into a linear expression allows for the kinetics to be estimated by experimental data obtained within relatively short assay periods. It is to be said, however, that the model proposed is empirical and its general applicability is to be demonstrated by further studies. Thus implementation of this model requires case experimentation, the appropriate fitting of data provided. Nomenclature TFn : total flavonoid concentration (mg RtE L-1) h : initial extraction rate (mg g-1 min-1) k : second-order extraction rate constant (g mg-1 min-1) L/S : liquid-to-solid ratio (mL g-1) t : time (min) T : temperature (°C or K) TFn : yield in total flavonoids (mg RtE g-1) TFn(s) : yield in total flavonoids at saturation (mg RtE g-1) References 1.Palma M., Barbero G.F., Piñeiro Z., Liazid A., Barroso C.G., Rostagno M.A., Prado J.M. and Meireles M.A.A., Extraction of natural products: principles and fundamental aspects. In Natural Product Extraction: Principles and Applications, Rostagno, M.A. and Prado J.M. Ed., The Royal Society of Chemistry, U.K., 58-88 (2013)2.Chan C.-H., Yusoff R. and Ngoh G.C. Modeling and kinetics study of conventional and assisted batch solvent extraction, Chem. Eng. Res. Des, (92), 1169-1186 (2014)3.Mohammed M.A., Shitu A. and Ibrahim A, Removal of methylene blue using low cost adsorbent: A review, Res. J. Chem. Sci, 4(1), 42-49 (2014)4.Dina D.J.D., Ntieche A.R., Ndi J.N. and Ketcha M badcam J., Adsorption of acetic acid onto activated carbons obtained from maize cobs by chemical activation with zinc chloride (ZnCl), Res. J. Chem. Sci, 2(9), 42-49 (2012)5.Malarvizhi T.S., Santhi T. And Manonmani S., A comparative study of modified lignite fly ash for the adsorption of nickel from aqueous solution by column and batch mode study, Res. J. Chem. Sci, 3(2), 44-53 (2013)6.Shubha D., Prasenjit M. and Chandrajit B., Removal of fluoride using Citrus limetta in batch reactor: kinetics and equilibrium studies, Res. J. Chem. Sci, 4(1), 50-58 (2014)7.Naveen D., Chandrajit B. and Prasenjit M., Studies of kinetic and equilibrium isotherm models for the sorption of cyanide ion on to almond shell, Res. J. Chem. Sci, 4(2), 20-24 (2014)8.Shehata E., Grigorakis S., Loupassaki S. and Makris D.P., Extraction optimisation using water/glycerol for the efficient recovery of polyphenolic antioxidants from two Artemisia species, Separ. Purif. Technol, (149), 462-469 (2015)9.Karvela E., Makris D.P, Kalogeropoulos N. and Karathanos V.T., Deployment of response surface methodology to optimise recovery of grape Vitisvinifera) stem polyphenols, Talanta, (79), 1311-1321 (2009)10.Cavdarova M. and Makris D.P., Extraction kinetics of phenolics from carob (Ceratoniasiliqua L.) kibbles using environmentally benign solvents, Waste Biomass Valor, (5), 773-779 (2014)11.Tzima K., Kallithraka S., Kotseridis Y. and Makris D.P., Kinetic modelling for flavonoid recovery from red grape Vitisvinifera) pomace with aqueous lactic acid., Processes, (2), 901-911 (2014)12.Seikova I., Simeonov E. and Ivanova E., Protein leaching from tomato seed: Experimental kinetics and prediction of effective diffusivity, J. Food Eng, (61), 165–171 (2004)13.Rakotondramasy-Rabesiaka L., Havet J.-L., Porte C. and Fauduet H., Estimation of effective diffusion and transfer rate during the protopine extraction process from Fumariaofficinalis L, Separ. Purif. Technol,(76), 126–131 (2010)