View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Available online at www.sciencedirect.com ScienceDirect

Procedia Chemistry 14 ( 2015 ) 47 – 55

2nd Humboldt Kolleg in conjunction with International Conference on Natural Sciences, HK-ICONS 2014 Microwave Assisted Extraction of Dioscorin from Gadung ( hispida Dennst) Tuber Flour

Andri Cahyo Kumoroa,b*, Indah Hartatic

aDepartment of Chemical Engineering, Faculty of Engineering, bCentre for Natural Medicine Research Diponegoro University, Jl. Prof. H. Soedarto, SH Semarang 50275 Indonesia cDepartment of Chemical Engineering, Faculty of Engineering, Wahid Hasyim University, Jl. Menoreh Tengah X/22 Sampangan Semarang 50236, Indonesia

Abstract

Gadung (Dioscorea hispida Dennst) tuber contains dioscorin, an alkaloid with high angiotensin converting enzyme-inhibitory capacity. In this research, dioscorin was extracted using microwave assisted extraction (MAE). The objectives of this research were to investigate the influence of ethanol concentration (75 % to 96 % w/w), solvent-material ratio (10 : 1 to 20 : 1) and microwave power (100 W to 400 W) on the extraction yield, and to develop a mathematic model to represent that process. The optimum condition for this process was extraction using 85 % aqueous ethanol at solvent-material ratio of 12.5 : 1, and microwave power of 100 W for 20 min. The mathematic model agreed well with the experimental data with average error of 2.96 %.

©© 2015 2015 The A.C Authors.. Kumoro, Published I. Hartati. by Elsevier Published B.V. byThis Elsevier is an open B.V. access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Scientific Committee of HK-ICONS 2014. Peer-review under responsibility of the Scientifi c Committee of HK-ICONS 2014 Keywords: Dioscrin; Dioscoreae hispida Dennst; extraction; gadung; microwave; model

*Corresponding author. Tel.: +62 24 7460 058; cell phone : .+62 818 0589 9868 E-mail address: [email protected]

1876-6196 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Scientifi c Committee of HK-ICONS 2014 doi: 10.1016/j.proche.2015.03.009 48 Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55

Nomenclature

RSD relative standard deviation S slope of regression line in dioscorin analysis SD intercept of the standard deviation of regression line in dioscorin analysis TLC thin layer chromatography v/v volume per volume expression of concentration of a solution w/w weight per weight expression of concentration of a solution

1. Introduction

Bitter Yam (Dioscoreae hispida Dennst) is one of the most economically important yam species, which serves as a staple food for millions of people in the tropical and subtropical countries1, and is classified as a wild creeping and climbing which can grow up to 20 m in height. D. hispida is commonly found in secondary forests and grows under shaded areas or near streams, which is locally known as Umbi Gadung2. The dormant underground tubers of both wild and cultivated varieties of this yam are harvested as a good starch and essential dietary nutrients source. In addition, this yam tuber exhibits nutritional superiority when compared with other tropical root crops. This property is related to the presence of resistant starch, which plays role in creating a slow digestion in the lower parts of the human gastrointestinal tract, resulting in slow liberation and absorption of glucose3. This tuber’s digestive property has suggested its utilization in reducing the risk of obesity, diabetes and other related diseases4. However, some studies have also pointed out that this yam tuber also contains some toxic compounds, which can impart serious health complications. Among the yam species, D. hispida, was considered as one of the most underutilized species because of the presence of poisonous alkaloids known as dioscorin5. To prevent any cases of intoxications, local people prepare the tubers before consumption to make them edible. The treatment practices vary from region to region, but commonly involve peeling, followed by leaching of the sliced tubers in running water, overnight, washing and drying. The resultant sliced tubers are then cooked before being eaten. In order to find new ways to stay healthy or fight diseases, people have embraced the versatile functions of herbal medicines. Amongst thousands of folk medicinal sources, the tuber of D. hispida is well known for its multifaceted biological functions6. Dioscorin, the storage protein accounted for about 80 % of extractable water-soluble proteins from D. hispida tubers7. It belongs to the family of pyrans, compounds containing a pyran ring, which is a six- member heterocyclic, non-aromatic ring with five carbon atoms, one oxygen atom and two ring double bonds. Various dioscorin contents of D. hispida dried tuber have been reported in the literatures. Sasiwatpaisit et al.8, determined dioscorin in D. hispida dried tuber by TLC-densitometry and TLC image analysis and reported the values were 0.72 % w/w and 0.66 % w/w respectively. Also employing TLC method however, Webster and his coworkers reported that the content of dioscorin in D. hispida dried tuber was only 0.12 % w/w)9. Although dioscorin is considered as a poisonous substance, it has been demonstrated to inhibit angiotensin-converting enzyme activity, which is a target for pharmacological agents used in the treatment of hypertension10. Lu et al7 also reported that dioscorin exhibited biological activities both in vitro and in vivo, including the enzymatic, antioxidant, antihypertensive, immunomodulatory, lectin activities and the protecting role on airway epithelial cells against dust mite allergen destruction allergens in vitro. With the purpose of taking the advantage of both starch and alkaloid of D. hispida , it is necessary to separate the alkaloid through a solid-liquid extraction process. Conventional extraction methods are often time consuming, little through-put and require large amount of organic solvent. Thus, they render to additional cost mostly associated with the purchase and disposal of toxic solvents and dealing with the environmental hazard. Microwave assisted extraction (MAE) offers many advantages over the conventional extraction methods, including shorter extraction time, less solvent consumption, higher extraction rate and better product quality at lower cost because the microwaves directly heat up the solvent or the mixture11. One of the factors that may affect the MAE process is the selection of solvent. A correct choice of solvent is fundamental for obtaining an optimal extraction process12. Solvent selection for MAE is determined by (i) the solubility of the target analyte in the solvent, (ii) the interaction between solvent and plant matrix, and (iii) the Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55 49 microwave absorbing properties of the solvent. The preferred solvent is the solvent that have a high selectivity towards the analyte of interest excluding unwanted matrix components11. Dioscorin is soluble in water, alcohol, aceton and chloroform, and slightly soluble in ether, benzene and petroleum ether. The solubility of dioscorin in water is 23.3 g · L–1 13. According to Hartati and Kumoro14, the Hildebrand solubility parameter of dioscorin is about 28 MPa1/2 or between ethanol and methanol, which are 26.1 and 29.7 MPa1/2, respectively. Since the Hildebrand solubility parameter of water is 48 MPa1/2, aqueous ethanol will be the most suitable for dioscorin extraction. The microwave absorbing properties of the solvent is represented by its dielectric constant. The higher the dielectric constant of a solvent will result in a higher heating rate of the solvent under the influence of microwave irradiation. Considering all of the above mentioned facts, the purpose of the present study was to investigate the influence of ethanol concentration (75 % to 96 % w/w solvent-material ratio (10 : 1 to 20 : 1) and microwave power (100 W to 400 W) on the extraction yield, and to develop a mathematic model to represent that extraction process.

2. Material and methods

2.1. Plant materials

Tubers of Dioscorea hispida Dennst were collected from a secondary forest in the vicinity of Diponegoro University Campus, Semarang-Indonesia. All tubers were washed to remove dirt, peeled, sliced, air–dried and milled to obtain 100 mesh size raw flour. Voucher specimens were then deposited in a storage container at 4 oC for further use in the extraction at Food Processing Laboratory, Department of Chemical Engineering, Faculty of Engineering, Diponegoro University, Indonesia.

7

5 (c) 1. Microwave oven (a)(a) 2. Extraction flask 3. Extraction mixture 6 4. Thermometer 4 5. Liebig condenser 6. Cooling water inlet 7. Cooling water outlet N (b)

CH3 2 O 3CH 3 1

O

Fig. 1. (a) gadung flour; (b) molecular structure of dioscorin; (c) the experimental set-up

50 Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55

2.2. Microwave assisted extraction of dioscorin

MAE was carried out in a domestic microwave oven with adjustable nominal power output (100 W to 400 W) as the applied microwave power. The microwave oven was equipped with fiber optic Luxtron I652 thermometer to record the real time temperature profile during extraction. The D. hispida tuber flour (15 g to 30 g) was mixed with 300 mL aqueous ethanol of (75 % to 96 % w/w) to make up a desirable predetermined solvent to feed ratio in a 500 mL Duran flask equipped with Liebig condenser and was irradiated in the microwave oven (Fig. 1) at predetermined time duration without stirring. After extraction, the mixture was allowed to cool to room temperature by using water bath. The mixture was then subjected to filtration for removal of the flour residue. The solvents were completely removed from the filtrate in vacuo from which dry extracts were obtained. They were then weighed and kept in the desiccator prior to TLC analysis for dioscorin determination. The MAE curve was constructed through series of extractions at the same operating conditions but with different extraction time using fresh sample. New batch of sample was taken from the storage container for the construction of each extraction curve.

2.3. Dioscorin determination

TLC analysis was performed on TLC aluminium oxide 60 F254 neutral (Merck, Germany) developed in chloroform – ethanol – ammonia, 100 : 10 : 0.5 following the method of Sasiwatpaisit et al.8. Five microliters of dioscorin standard solution (0.5 mg · ml–1 to 2.5 mg · ml–1) and all sample solutions (5 mg · ml-1 to 20 mg · ml-1) were spotted as 6.0 mm bands in length onto a same TLC plate using a Camag Linomate V syringe. A distance between each spot was 9.4 mm. The plate was then developed to a distance of 8.0 cm in a TLC chamber previously saturated with methanol – chloroform (3 : 97, v/v) for at least 30 min. After development, TLC plate was scanned by a CAMAG TLC scanner III with CAMAG winCATS software in the absorbance mode at 254 nm. The slit dimension was 4.00 mm × 0.30 mm and the scanning speed was 20 mm · s–1. Limit of detection (LOD) and limit of quantitation (LOQ) were determined from the calibration curve as 3.3SD/S and 10SD/S respectively where SD was the y-intercept standard deviation of regression line and S was the slope of regression line. The repeatability and intermediate precision were determined by analyzing six replicates of five different concentrations [(2.5, 5.0, 7.5, 10.0 and 12.5) μg · spot–1)] of dioscorin and expressed as percent relative standard deviation (% RSD). D. hispida extract samples were spiked with dioscorin and the accuracy was expressed as percentage of recovery. The dioscorine content of D. hispida dried tuber used in this study was found to be 0.63 ± 0.07 % w/w. This value is slightly lower than dioscorin contents of D. hispida dried tuber originated from Thailand that reported by Sasiwatpaisit et al.8, but higher than that studied by Webster and his coworkers9, which were 0.72 ± 0.07 % w/w and 0.12 % w/w, respectively.

2.4. Mathematical modeling

The extraction profile of MAE would exhibit similar trends as those observed in typical dynamic extraction profiles (yield vs. time), where two distinct phases are observed, i.e. washing and diffusion steps, respectively15. Two models will be verified their accuracy against the experimental data. The first model is developed based on the original film theory. This model is usually employed for the modeling of typical extraction curve (yields vs. extraction time) can be written as follows16:

CA  1(1  ).eb  tk ).( (1) CAs where CA is the concentration of analyte in the solvent (mg · L–1), CAs is the equilibrium concentration of analyte in the solvent (mg · L–1), b characterizes the washing step, k characterizes the diffusion step (min–1), while t is the extraction time (min). The second model is developed from overall mass transfer phenomena, which describe the dynamic transfer of analyte from the solid matrix to the bulk solvent:

Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55 51

dCA NA K  CACA )( (2) dt sa s

From mass balance of analyte in the solvent, the extraction yield can be calculated using equation:

CA  1  e  sa tK ).( (3) CAs where Ksa is the overall mas transfer coefficient of analyte from solid matrix to solvent (min–1). The values of all modeling parameters were evaluated using solver available in Microsoft Excel 2010.

3. Result and discussion

The results of this work and their comparisons with results obtained by other researchers are presented in the following sections.

3.1. Effect of ethanol concentration

To study the effect of ethanol concentration, a set of MAE experiments were conducted using 50 g gadung flour, solvent volume to flour mass ratio of 10 : 1, microwave power 100 W and ethanol concentrations of (75 to 96) % w/w. Fig. 2 (a) depicts the effect of ethanol concentration as solvent on the extraction yield. The figure shows that the maximum extraction yield (85.37 %) was obtained when 85 % w/w ethanol used as solvent. The presence of water in the 85 % w/w ethanol increased ethanol polarity which accelerated solvation of dioscorin. In addition, water also caused gadung flours swelling which allows ethanol to penetrate into the cell structure of solid particles more easily17. Similar results were also reported by Chen et al.18, who extracted triterpenoids from Ganoderma atrum. They reported that extraction yield increased with the increased of ethanol concentration. Unfortunately, the extraction yield decreased when ethanol concentration exceeded 85 % w/w. At these concentrations, the Hildebrand solubility parameters of the aqueous ethanol are lower than that of dioscorin (28 MPa1/2). Therefore, these solvents only performed limited ability to dissolve dioscorin.

100 100

90 (a) 90 (b)

80 80 eld, % eld, ield, % ield,

yi

70 70 Extraction y Extraction

Extraction Extraction 60 60

50 50 70 75 80 85 90 95 100 7.5 10 12.5 15 17.5 20 22.5 Ethanol concentration, % Solvent : Flour Ratio

Fig. 2. (a) effect of ethanol concentration on extraction yield; (b) effect of solvent volume to flour mass ratio on extraction yield 52 Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55

However, excess water may also induce excessive thermal stress. Too much water means more microwave energy absorbed by gadung flour particles. The value of dielectric constant of a solvent reflects its microwave absorbing capacity. Since water has high dielectric constant, it has a higher heating rate under the influence of microwave irradiation11. As shown in Fig. 2 (a), lower extraction yield was obtained at low ethanol concentration due to faster dioscorin decomposition rate as a result of thermal stress experienced by gadung flour. Lower ethanol concentration also reduced the extraction selectivity. This is because some glucosides, proteins and starch are also easier to dissolve. These findings agreed well with those reported by Xiang et al.19 who extracted sapogenins from tea leaves.

3.2. Effect of solvent volume : flour mass ratio

Solvent volume is one of important parameter in extraction process. Solvent volume has to be adequate to ensure that raw material is completely immersed in the solvent11. The use of large volume of solvent may result in reduction of extraction yield. In addition, it is also uneconomical due to high solvent and extract purification costs. To study the effect of solvent volume : flour mass ratio on the extraction yield, a set of experiments which employ 85 % v/v ethanol as solvent with solvent volume : flour mass ratios between 10 to 20 were conducted. The experimental results are presented in Fig. 2 (b). Fig. 2 (b) shows that highest extraction yield (86.66 %) was achieved when the extraction was performed at solvent volume : flour mass ratio of 12.5 : 1. Implementations of ratio higher than 12.5:1 resulted in lower extraction yield. This is because at higher solvent volume : flour mass ratio, there will be more water present in the extraction system. Solvent with high dielectric constant will absorb more microwave energy. Too much water means more microwave energy absorbed by gadung flour. From this point of view, water will experience a higher heating rate under the influence of microwave irradiation11. Excess of water may induce excessive thermal stress and gadung flour swelling. Dioscorin may experience decomposition as a result of thermal stress experienced by gadung flour. This result is in good agreement with Chen et al.18 who found that optimum solvent:raw material ratio 25 : 1 for the extraction of triterpenoid saponins from Ganoderma atrum using 75 % w/w ethanol. Wang et al.17 also reported that the extraction yield of Radix puerariae tended to decrease as solvent : raw material ratio higher than 30 : 1.

3.3. Effect of microwave power

Microwave power and extraction time are two factors to be considered to prevent thermal degradation of extracts in microwave assisted extraction11. In general, higher extraction efficiency can be achieved at higher microwave power from 30 W to 150 W20. The effect of microwave power was studied on extraction of dioscorin from 50 g gadung flour in 85 % w/w ethanol with solvent volume : flour mass ratio of 12.5 : 1 for 20 min in a microwave reactor. The temperature inside the microwave reactor reached up to boiling point of the solvent. The effects of microwave power on extraction yield obtained in this study are depicted in Fig. 3. It is clear that highest extraction yield was achieved at lowest microwave power (100 W). This result agrees well with that reported by Fernandes et al.22 on the extraction of organochlorine compounds from sea sediment21. Similar results of decrease in extraction yield of astragalosides from Radix astragali at high power due to disorderly molecular interactions have been reported in the optimization study of MAE of four main astragalosides in Radix astragali. As the experiments are conducted in dry matter, as is usually the case11, chances of degradation due to drying or evaporation at higher microwave power intensity are ruled out. They further observed that microwave energy affected the solvation of the analytes and finally the rate of extraction. Reduction of extraction yield during implementation of high microwave power may be due to decomposition of analyte at higher microwave power range, temperature or decrease in solubility23. Mandal et al.11 found that most phytochemicals are thermally labile. Therefore, they are easily decomposed when exposed to heat or high temperature, specifically at longer exposure time. Chen and Lin24 also reported reduction of concentrations and stability of alkaloids of some dioscorea species when heated at temperature between 500 oC to 800 oC. Further heating up to 900 oC caused decomposition of the alkaloids. In this study, less extracts with lower dioscorin concentration were obtained when higher microwave power employed for the extraction. Therefore, it can be concluded that at higher microwave power, dioscorin experiences both solubility decrease and decomposition. Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55 53

100 90

80 (%) (%)

ield 70 y 60

Extraction Extraction 50 40

30 0 100 200 300 400 500

Microwave power (Watt)

Fig. 3. Effect of microwave power on extraction yield

3.4. Comparison of mathematical model

In the present study an attempt was made to predict the extraction yield profile calculated using original film theory model (Equation (1)) and overall mass transfer model (Equation (3)). Equation (3) represents an approach of analyte mass transfer via steady state diffusion through combined solid-liquid film mechanism. The overall mass transfer coefficient (Ksa) was evaluated by minimization of the sum of the square of errors (SSE ) between extraction yields of Equation (3) and experimental data. Similarly, diffusion step (k) and washing (b) step constants were evaluated from Equation (1). Fig. 4 presents the comparison of extraction yields obtained from MAE experiments using 85 % and 96 % ethanol, and those obtained from calculations using original film theory and overall mass transfer models. While the values of all modeling parameters are tabulated in Table 1.

Table 1. Optimized model parameters

–1 –1 Model Ethanol conc. (% v/v) Ksa (min ) k (min ) b (-) Error (%) Original film theory 85 - 0.1810 0.8293 13.27 96 - 0.1605 0.8306 13.74 Overall mass transfer 85 0.1213 - - 3.23 96 0.1136 - - 2.69

Overall mass transfer coefficients (Ksa) obtained from the mass transfer model and diffusion step constant (k) obtained from the film theory model were higher for MAE of dioscorin from gadung flour using 85 % w/w ethanol than that of using 96 % w/w ethanol. These modeling results agree well with the experimental data shown by Fig. 4 where the slope of the extraction curve for extraction using 85 % w/w ethanol is slightly steeper than that of using 96 % w/w ethanol. However, the washing step constants were almost not affected by ethanol concentration. The overall mass transfer model performs it superiority over the original film theory model for the extraction yield prediction as the former needs only one adjustable parameter but results lower error. The overall mass transfer model also offers its simplicity for practical applications.

54 Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55

110

100

90

80 70 ield, %

y 60 Y 96% Y96% MT 50 Y96% FT Y 85 % 40 Y85% MT

Extraction Y85% FT 30

20 10 0 0510 15 20 25 30 35 40 Time, min Fig. 4. Comparison of extraction yield data and predictions using film theory and mas transfer models

4. Conclusion

Ethanol concentration and solvent volume : flour mass ratio affected the MAE of dioscorin, and show certain optimum values. On the other hand, microwave power inversely affected the extraction yield. As a conclusion, the optimum condition for MAE of dioscorin from gadung flour was extraction using 85 % w/w ethanol with solvent volume : flour mass ratio of 12.5 : 1, and microwave power of 100 W for 20 min. The overall mass transfer model performed better extraction yield predictions than the original film theory model and agreed well with the experimental data with average error of 2.96 %.

Acknowledgements

The authors would like to acknowledge the organizing committee for awarding a Travel Grant of 2nd Humboldt Kolleg-ICONS 2014 that make their participation in the event becomes possible.

References

1. Udensi EA, Oselebe HO, Iweala OO. The investigation of chemical composition and functional properties of water yam (): Effect of varietal differences. Pak J Nutr 2008;7(2):324–344. 2. Nashriyah M, Nornasuha Y, Salmah T, Norhayati N, Rohaizad M. Dioscorea hispida Dennst. (): An overview. Bull UniSZA 2010; 4: 12–13. 3. Kumoro, AC, Retnowati DS, Budiyati CS. Removal of cyanides from gadung (Dioscorea hispida Dennst.) tuber chips using leaching and steaming techniques. J Appl Sci Res 2011;7(12):2140–2146. 4. Aprianita A., Purwandari U, Watson B, Vasiljevic T. Physico-chemical properties of flours and starches from selected commercial tubers available in Australia. Int Food Res J 2009;16:507–520. 5. Bhattacharyya SS, Mandal SK. In vitro studies demonstrate anti-cancer activity of an alkaloid of a plant (Gelsemiun sempervirens). Exp Biol Med 2008;233(12):1591–601. 6. Wang CS, Chen MJ, Tseng CY, Wei KM, Chen W. Structural characterization of functional compositions isolated from Dioscorea purpurea (cultivar of Ming-Chien) by raman spectroscopy, Nat Prod Chem Res 2013;1(3). 7. Lu YL, Chia CY, Liu YW, Hou WC. Biological activities and applications of dioscorins, the major tuber storage proteins of yam. J Tradit Complement Med 2012;2(1):41–46. 8. Sasiwatpaisit N, Thitikornpong W, Palanuvej C, Ruangrungsi N. Dioscorine content in Dioscorea hispida dried tubers in Thailand by TLC- densitometry and TLC image analysis. J Chem Pharm Res 2014;6(4):803–806. 9. Webster J, Beck W, Ternai B. Toxicity and bitterness in Australian Dioscorea bulbifera L. and Dioscorea hispida Dennst. from Thailand. J Agric Food Chem 1984; 32( 5): 1087–1090. 10. Hsu FL, Lin YH, Lee MH, Lin CL, Hou WC. Both dioscorin, the tuber storage protein of yam (Dioscorea alata cv. Tainong No. 1), and its peptic hydrolysates exhibited angiotensin converting enzyme inhibitory activities. J Agric Food Chem 2002;50:6109–6113. Andri Cahyo Kumoro and Indah Hartati / Procedia Chemistry 14 ( 2015 ) 47 – 55 55

11. Mandal,V, Mohan,Y, Hemalatha S. Microwave Assisted Extraction-An Innovative and Promissing Extraction Tool for Medicinal Plant Research. Pharmacogn Rev 2007;1(1):7–18. 12. Mandal,V, Dewanje S, Mandal, SC. Microwave assisted extraction of total bioactive saponin fraction from Gymnema sylvestre with reference to Gymnemagenin. Phytochem Anal 2009:491–497. 13. Merck. Dioscorine. Whitehouse Station, New York: Merck and Co., Inc.; 1999 14. Hartati I, Kumoro, AC. Solvent selection for microwave assisted extraction of dioscorin from Dioscorea hispida Dennst. flour, In: Robandi, editor. Proceedings of International Seminar on Applied Technology, Science, and Arts (2nd APTECS), Surabaya, 21-22 Dec. 2010. 15. Franco D, Sineiro J, Pinelo M, Núñez MJ. Ethanolic extraction of Rosa rubiginosa soluble substances: Oil solubility equilibria and kinetic studies. J Food Eng 2007;9(1):150–157. 16. Velickovic DT, Milenovic DM, Ristic MS, Veljkovic VB.Ultrasonic extraction of waste solid residues from the Salvia sp. essential oil hydrodistillation. Biochem Eng J 2008;42(1): 97–104. 17. Wang YL, Xi GS, Zheng YC, Miao FS. Microwave assisted extraction from Chinese herb Radix puerariae. J Medicinal Plant Res 2010;4(4):304–308. 18. Chen Y, Xie MY, Gong XF. Microwave assisted extraction used for the isolation of total triterpenoid saponins from Ganoderma atrum. J Food Eng 2007;81:172–170. 19. Xiang L, Zhong MJ, Jiang X, Dhong SY. A study on the extraction and purification of tea sapogenin. Afr J Biotechnol 2010; 9(18):2691– 2696. 20. Shu YY, Ko MY. Microwave assisted extraction of ginsenosides from ginseng root. Microchem J 2003;74:131–139. 21. Fernandes AE, Ferrera ZS, Rodriquez JJS. MAE of organochlorine compounds in marine sediments with organized molecular systems. Chromatographya 2001;53:357–379. 22. Yan MM, Liu W, Fu YJ, Zu YG. Optimisation of the microwave-assisted extraction process for four main astragalosides in Radix astragali. Food Chem 2010;119:1663–1670. 23. Kullu J, Dutta A, Constales D, Chaudhuri S, Dutta D. Experimental and modeling studies on microwave-assisted extraction of mangiferin from Curcuma amada. Biotech 2014;4:107–120. 24. Chen YT, Lin KW. Effect of heating temperature on the total phenolic compound, antioxodative and stability of dioscorin of various yam cultivars. Food Chem 2007;101:955–963.