African Journal of Pure and Applied Chemistry Vol. 5(1), pp.
1-8, January 2011 A
vailable online at http://www.academicjournals.org/AJPAC ISSN 1996 - 0840 2011 Ac ademic Journals Full Length Research Paper Characterization of caffeine and determination of caffeine in tea leaves using u v-visible spectrometer Tadelech Atomssa* and A.V. Gholap Department of Physics, Faculty of Science, Addis Ababa University, P.O. Box 1176 , Addis Ababa, Ethiopia. Accepted 4 November, 2010 In this research, characterization of pure caffeine and the method for determina tion of caffeine content in twelve commercial tea leaves have been reported usin g UV-visible spectrometer. The optical transition properties of caffeine were me asured in different solvents (dichloromethane, water, chloroform and ethyl aceta te). The results show that, caffeine has highest optical transitions in dichloro methane than the other solvents. After characterization of the electron transiti on, the content of caffeine in 6 Ethiopian black and green tea leaves and 5 diff erent black tea leaves from South Africa, Kenya, 2 India, Zimbabwe, and Sri Lank a green teas, were analyzed at 30 Four Ethiopian tea samples C. were also analyze d at the boiling temperature using smaller extraction time. The result of the ex periment at both temperatures indicates that, the caffeine content of Ethiopian green tea leaves is greatest of all the Ethiopian teas while black lion tea has the least. Of all the samples tested, Kenyan black tea had the highest caffeine content and Sri Lankan green teas had the least caffeine. The result also indica tes that, caffeine can be extracted more at the boiling temperature than at 30 Fu rthermore, C. the absorption range of cinnamon tea was greater than the others b y 16 nm. Key words: Tea leaves, caffeine, extraction, UV-visible spectrometer, o ptical transition properties. INTRODUCTION Tea refers to the agricultural produc ts of the leaves, leaf buds and internodes of the Camellia sinensis plant. It ha s been consumed as a beverage for almost 2,000 years starting in China. It is th e most widely consumed beverage after water (Alan and Iris, 2004). Teas from man y areas may be blended. The aim is to obtain better taste, higher price or both, as a more expensive, better tasting tea may cover the inferior taste of cheaper varieties. There are two major kinds of tea, black tea and green tea. Both cont ain caffeine (1 to 5) % of its dry weight (Amro et al., 2006) depending on type, brand (Bennett and Bonnie, 2001) and brewing method (Hicks et al., 1996). This is why the reported values in the literature are so variable. Caffeine belongs t o a family of naturally occurring components known as xanthines. The xanthines w hich come from plants are possibly the oldest known stimulants. Caffeine is the most powerful xanthine, in its ability to increase alertness, put off sleep and to improve attention in study (Boltonad, 1981), caffeine is a vasodilator (relax es the blood vessels) as well as a diuretic (increase urination). On the other h and, sever restlessness and excitement, leading to mild delirium, muscular tensi on and twisting and cardiovascular disturbances such as tachycardia, are negativ e effects of caffeine at large doses (Boltonad, 1981). The spinal cord is stimul ated at higher doses, convulsions and death may result (Bolton and Null, 1981). Many consumers prefer to avoid caffeine partially or altogether, due to its stim ulant effects and others, still on health concern. This makes decaffeination of tea an important industrial process. In addition, caffeine has a slightly bitter flavor. As a result, decaffeinating coffee beans and tea leaves will leave the flavor slightly changed, even if no other components are lost. It should be note d that, decaffeinated coffee and tea are not caffeine free. Decaffeination is do ne by the use of solvent, which extract caffeine. For this purpose, the currentl y available solvents are chloroform, dichloromethane, ethyl acetate, super criti cal carbondioxide, etc. Dichloromethane is used to decaffeinate a high proportio n of conventional *Corresponding author. E-mail: tadelechad@yahoo.com. 2 Afr. J. Pure Appl. Chem. teas. It is also a relatively non-toxic solvent and is often used in place of ch loroform. Ethyl acetate also removes caffeine from tea leaves effectively; it ca n also extract other chemical components as well (Abdul Mumin et al., 2006). Stu dies on green tea, decaffeinated with ethyl acetate, have shown a potential of u p to 30% of epigallocatechin gallate (EGCG, considered to be the beneficial comp onent in green tea) and other beneficial antioxidant compounds, to be extracted along with caffeine (Kirmer, 1988; Johnson, 1988; Jefferson, 1998). Due to the w ide spread consumption of caffeine, it is important to collect precise informati on of their content in foods. Most research activities have been focused on chro matographic methods; however, spectrophotometric determination is preferred beca use of its rapidity, high accuracy and reproducibility (Komes et al., 2009). Fur thermore, UV-visible spectrophotometer is cheap and it is found in many laborato ries. But caffeine content in tea leaves cannot be determined directly using UV visible spectrometer, due to the matrix effect of UV absorbing substances (Guzin , 2002). In this study, we have developed a method to use UVvisible spectrometer for the determination of caffeine in tea leaves. That is, dissolving tea leaves first in distilled water and then extracting caffeine by dichloromethane and an alyzing it using UV-visible spectrophotometer, to define the caffeine content of different Ethiopian tea leaves and to examine the extraction efficiency of caff eine by hot water, with respect to extraction time and temperature. For the comp arison purpose, caffeine content of Sri Lankan green tea, Kenyan, Indian, South African, and Zimbabwean black tea leaves were also analyzed. The method includes characterizing pure caffeine in water, dichloromethane, chloroform and ethyl ac etate. MATERIALS AND METHODS Samples A set of twelve commercial tea samples made up of, five Ethiopian black teas (Addis, Black lion, Cinnamon, Tosign and Ginger), one Ethiopian green tea, one Sri Lankan green tea, two Indian (Tatagold, Taj Mahal) , one South African, one Kenyan and one Zimbabwean (Taganda )black tea. All Ethi opian tea samples and Sri Lankan green tea were from supermarkets in Addis Ababa . The other tea leaves were purchased from their respective countries. These sam ples were from tea bags except black lion and Indian Tatagold black tea. Cinnamo n, tosign, and ginger tea leaves are blended teas with cinnamon, tosign and ging er respectively. Pyrex, England), beakers, thermometer, 1 cm quartz cuvette, separatory funnel, fun nel, ice bath, distilled water, dichloromethane (assay: 99.6%, Aldrich Germany), chloroform (assay: 99.8%, European Union), ethyl acetate (assay: 99%, Indian ), caffeine (M.W. 194.19 g/mol, Aldrich Germany). Preparation of samples and stand ard solution Standard solution preparation To characterize caffeine in different solvents and to determine caffeine content in different samples of tea, solutio ns of caffeine in different solvents were prepared. 0.46 mg, 1.40 mg, 0.85 mg an d 0.80 mg of pure caffeine was immersed into 25 ml, 50 ml, 50 ml and 40 ml of di chloromethane, water, chloroform and ethyl acetate respectively and stirred by m agnetic stirrer for 30 min. The absorbance versus wavelength of the solution was measured using UV-visible spectrometer. These procedures were repeated ten time s for different concentrations. Finally, the molar decadic absorption coefficien t, transitional dipole moment and integrated absorption coefficient were calcula ted. Tea sample preparation For 12 tea samples (all the tea samples listed in th e samples above) when distilled water for dissolving tea leaves was at 30 C: 30 m l of distilled water was heated to 30 50 mg of tea leaves C. were immersed into t he hot water and stirred for 30 min with a magnetic stirrer. It was then filtere d through a glass filter and cooled to room temperature (21C). 40 ml of dichlorom ethane was poured into the tea infusion and stirred for 10 min with a magnetic s tirrer. Since caffeine is more soluble in dichloromethane (140 mg/ml) than it is in water (22 mg/ml), it readily dissolves in dichloromethane. The water phase a nd the dichloromethane phase were then separated by a separatory funnel. The vol ume of dichloromethane phase was measured. This solution was poured into 1 cm qu artz cuvette and placed into the sample holder of the spectrometer and the spect rum was then taken. These procedures were repeated four times .The total content of caffeine in tea samples were found from the spectrum of combinations of the four extractions. The contents of caffeine in the 5th extract were found negligi ble. For four tea samples of only Ethiopian origin (Addis, black lion and cinnam on black teas and green tea) at the boiling temperature (94 50 mg tea leaves were immersed into 30 ml boiled distilled C): water and stirred for 3 min and filter ed through a glass filter and cooled to room temperature using ice bath. 40 ml o f dichloromethane was poured into the tea infusion and stirred for 5 min by a ma gnetic stirrer. The dichloromethane phase and water phase were separated by a se paratory funnel. Volume of dichloromethane phase was measured and analyzed by UV -visible spectrometer. These procedures were repeated up to the fourth extractio n, similar to that of 1 and the percentage of caffeine was determined. Note that , back ground correction using dichloromethane was taken before taking spectrum of the above measurements and the cuvette was rinsed four times and dried before taking the next measurement. To make the result more reliable, three independen t measurements were taken for each sample and the average values were taken. Materials and chemicals UV-visible spectrometer (Perkin-Elmer Lambda 19, wavelen gth range 170 nm to 3200 nm, double monochromator), microbalance (Mettler At250, 0.1/0.01 mg) for measuring mass of caffeine, balance (BP 210S, d = 0.1 mg, Max 210 g) for measuring mass of tea leaves, measuring cylinder, magnetic stirrer, g lass filter (2 RESULTS AND DISCUSSION The absorption spectra of pure caffeine and caffeine in t ea leaves were measured using UV-visible spectrometer Atomssa and Gholap 3 3 1 0 250 300 350 400 450 500 Absorbance Absorbance 2 2 0 250 300 350 400 450 500 Wavelength (nm) Wavelength (nm) Figure 1. Absorption spectrum of caffeine in dichloromethane. Figure 3. Absorption spectrum of caffeine in chloroform. 3 Absorbance 2 1 0 250 300 350 400 450 500 Wavelength (nm) Absorbance 2 0 250 300 350 400 450 500 Figure 2. Absorption spectrum of caffeine in water. Wavelength(nm) in the spectral range of 200 to 500 nm. The spectra were recorded at a spectral bandwidth of 2 nm and scan speed of 480 nm per min. Each data in these experimen tal activities was found from computerized recorder interfaced with the spectrom eter. It was then analyzed using origin 6.1 software. UV-visible absorbance of c affeine in dichloromethane, water, chloroform, and ethyl acetate To determine th e molar decadic absorption coefficient, integrated absorption cross-section and transitional dipole moments of caffeine in dichloromethane, water, chloroform an d ethyl acetate, a series of solutions each Figure 4. Absorption spectrum of caffeine in ethyl acetate. with a known concentration was prepared. The absorbance versus wavelength was me asured using UVvisible spectrometer and the results were presented in Figures 1, 2,3,4. From the spectra, it can be observed that caffeine absorbs in the spectra l range between 244 nm to 300 nm in dichloromethane and water with max at 276 nm and 272.8 nm respective y, 250 nm to 300 nm at nm in ch oroform and 258 nm to 300 nm at nm in ethy acetate. max of 274.4 max of 276 4 Afr. J. Pure App . Chem. 1.0 Absorption coefficient/cm-1 coefficient measures the intensity of optica absorption at a given wave ength ( Angew, 1969) was ca cu ated from Equation (3) at max for caffeine in the four so vents. From Equation (1), the absorption coefficient is given by 0.5 1 I a = n 0 I (4) The integrated absorption coefficient at which is the sum of absorption coeffici ent for a frequencies in the band is expressed as: a t = a d , 0.0 30000 35000 40000 -1 (5) Waveumber/cm Figure 5. a versus of caffeie i dichloromethae. where is the frequecy. The itegrated absorptio coefficiet is idepedet of lie fuctio, which may vary with parameters like pressure, temperature ad sol ute solvet iteractio. It thus provides a measure of the iheret absorbig st regth of atoms. The itegrated absorptio crosssectio t , can be calculate by the following equation: Optical transition properties of caffeine in ichloromethane, water, chloroform an ethyl acetate From UV-visible absorption spectra, the optical transition pro perties of caffeine were calculate in solvents to compare the strength of trans ition. For incient light intensity I0, propagating a istance l in the absorbin g meium, the transmitte light intensity I can be escribe as: I = I0 e where a t = 1 a , N (6) , is the absorptio coefficiet. (1) a The BeerLambert aw resu ts irect y from Equatio (1) (Guter a Tua, 2003). ( )cl I = I0 , whr N is th umbr dsity of th molculs. Th itgratd absorptio cross- sctio of caffi ca b foud by rcalculatig th absorbac vrsus wav um br usig origi 6.1 softwar. Th itgratd ara udr th curv was obtaid by itgratig from -1 32,981.5303 to 40,849.6732 cm . Figur 5 shows th spctr a of absorptio cofficit vrsus wavumbr of caffi i dichloromtha. Th trasitioal dipol momt of th dissolvd molcul, which is rlatd to th molar dcadic absorptio cofficit by th itgral absorptio cofficit, is calculatd by th quatio (Liptay, 1969 ad Michal, 1999): IA = bad (2) ( ) 1 2 h 2 Na d = fi 3 l (10 )c 0 0 -1 2 = 2 1 S fi , 3 (7) whr is th molar dcadic absortio cofficit ad c is th coctratio of th absorbig comoud. Th Br-Lambrt law is usually xrssd i its logarit hmic form: ( ) whr S = 2.9352 1060 c-2mol . Th trasitioal diol momt of caffi was fo ud by rcalculatig th absorbac vrsus wavlgth ito l I0 = A = ( ) cl, I (3) usig orgi 6.1 softwar alog with Equatio 7. Figur 6 shows th sctra of mo lar dcadic absortio cofficit ovr wavumbr vrsus wav umbr of caffi i dichloromtha. ( ) vrsus
whr A is th dimsiolss quatity calld absorbac, ad I/I0 trasmittac
(T). Th molar dcadic absortio ( ) vrsus Atomssa ad Ghola 5 Tabl 1. Otical trasitio rortis of caffi i dichloromtha, watr, ch loroform ad thyl actat. Solvt Dichloromtha Watr Chloroform Ethyl actat Molar 2 -1 cofficit( max ) i m mol 12240.689 9200.85 6581.70 10082.72 dcadic absortio Itgratd absortio -1 cross sctio i cm mol (86.650.47)10 -15 (65.040.11)10 - 15 (43.910.10)10 -15 (56.810.61)10 -15 Trasitioal diol momt i Cm (12.070.06)10 -30 (10.390.01)10 -30 (8.580.01)10 - 30 (9.740.03)10 -30 -1 Molar dc.abs. coff./wavumbr/cm mol 500 3 1.4 1.2 .Ex. data ____Liar fit 400 1.0 300 200 Absorbac 35000 40000 -1 0.8 0.6 0.4 0.2 100 0 30000 Wavumbr/cm Figur 6. ( ) 0.0 0.00 0.02 0.04 0.06 3 0.08 0.10 vrsus of caffi i dichloromtha. Coctratio(mol/m ) Figur 7. Absorbac vrsus coctratio grah for caffi i dichloromtha. . Thus, for ur caffi th lctroic ty * trasitio is ad this trasitio is th caus for * is I lctroic sctroscoy scially i orgaic molculs, th trasitio obsr vd i UV-visibl rgio or i othr wards Br-Lambrt law is valid. Dtrmiat io of caffi i ta lavs Caffi cott i ta lavs caot b dtrmi d dirctly usig UV-visibl sctromtr, du to th matrix ffct of UV absorbi g substacs (Guzi, 2002). This ffct ca b obsrvd i th sctral bads o f ta lavs i watr as show i Figur 8; as s from th Figur, caffi s ctra itrfac with othr comouds i ta lavs. Thrfor, it is cssary t o dvlo a mthod to ovrcom this difficulty. Th mthod dvlod i ths x rimtal activitis is to first dissolv caffi i watr ad th xtract it , usig dichloromtha as mtiod i th rocdur art. I Figur 9, th abs ortio sctra of ur caffi ad caffi xtractd by dichloromtha xact ly ovrlas. absortio. Usig th abov quatios, th otical trasitio rortis of caff i i diffrt solvts ar rstd i Tabl 1. Validatio of th mthod Th calibratio grah corrlatig to th absorbac ad coctratio of ur ca ffi i dichloromtha was costructd at th highst ak for coctratio rag -8 -3 of (1.16 to 9.47)10 molcm (Figur 7). Th stadard dviatio is 0.00 84 ad th liar rgrssio cofficit is 0.99972. From Figur 7, good liar rlatioshis wr obsrvd for a wid coctratio rag. This idicats that absorbac is dirctly roortioal to coctratio 6 Afr. J. Pur Al. Chm. 3 Absorbac 2 1 Absorbac 2 B C D E F 0 250 300 350 400 450 500 0 250 300 350 400 450 500 Wavlgth(m) Figur 10. Th ovr lad sctra of caffi for diffrt roud of xtractio (B) for first roud, (C) for scod roud, (D) for third roud, (E) for fourth roud, ad (F) for fifth roud. Wavlgth (m) Figur 8. Absortio sctrum of ta lavs i watr. 3 __Absortio sctrum of ur caffi __Absortio sctrum of caffi xtract d from ta lavs Absorbac 2 1 0 250 300 350 400 450 500 Wavlgth (m) Figur 9. Absortio sctra of ur caffi ad caffi xtractd from ta l avs. This idicats that dichloromtha was a good solvt usd to xtract caffi from ta lavs. To dtrmi caffi from ta lavs, th xtractios wr r atd four tims util th sctrum of caffi bcoms flat, wh s udr UV -visibl sctromtr. I th fifth xtractio, th cott of caffi was fou d gligibl (Figur 10). Tabl 2 rsts th xrimtal rsult for th dtrmiatio of caffi at tw o diffrt tmraturs (30 C ad 94 C or boilig tmratur). For covic, w usd to rrst ta samls by symbols i Tabl 2. Th rrstatios ar as follows, Addis black ta (A) , Ethioia gr ta (B), black lio ta (C), c iamo black ta (D), tosig black ta (E), gigr black ta (F), Sri Laka gr ta (G), Idia Tatagold black ta (H), Idia Taj Mahal ta (I), Kya bla ck ta (J), South Africa black ta (K), ad Zimbabwa Tagada black ta (L). T h rsult of th xrimt idicats that (ot idicatd i th Tabl 2), most caffi was xtractd durig th first xtractio (about 85% for 30 ad 65% C fo r 94 of th total caffi. Th rsult i Tabl 2 C) idicats that at 30 th rc tag of caffi i C, Ethioia gr ta lavs is th gratst, of all th othr Ethioia ta lavs studid. O th othr had, Ethioia black lio ta has th last caffi cott comard to th othr Ethioia ta lavs. As ca b s from th Tabl 2, caffi cott of Kya black ta lavs is th g ratst of all th othr ta lavs. Nxt to Kya, Idia Taj Mahal ta has th xt highr cott of caffi .To th othr d, Sri Laka gr ta, which is urchasd i Ethioia sur markts, has th last caffi cott of all th abov tas. From th ta lavs tstd at th boilig tmratur (94 C), sim ilar to th rsult at 30 th cott of C, caffi i Ethioia gr ta lavs is th gratst of th othr ta lavs ad Ethioia black lio ta has th l ast caffi cott. Wh w comar th cott of Atomssa ad Ghola 7 Tabl 2. Exrimtal rsult of dtrmiatio of caffi i ta lavs at two d iffrt tmraturs (30 ad 94 ( = 3). C C) 30 C Saml A B C D E F G H I J K L 94 C Prctag of caffi 2.010.04 2.080.01 1.500.00 1.980.09 1.940.04 1.630.03 1. 04 1.870.09 2.300.04 2.360.00 1.700.05 1.520.01 max (m) 274.800.28 275.20.00 274.40.00 276.80.00 275.20.00 275.20.00 275.60.28 275.60.28 27 0 275.60.28 275.20.56 275.60.28 Amax 0.4680.011 0.4780.004 0.3460.004 0.4620.025 0.4580.009 0.3820.006 0.3170.013 0. .040 0.5380.001 0.5470.006 0.4010.011 0.3570.005 max (m) 275.50.28 275.20.00 275.50.21 276.00.00 Amax 0.4860.010 0.4940.028 0.3660.005 0.4220.019 Percetage of caffeie 2.110.04 2.150.01 1.600.01 1.870.08 Absorbace 2 ____Ethiopia gree tea Ciamo tea 0 250 300 350 400 450 500 Wave egth(m) Figure 11. Over appe spectra of caffeie extracte from Ethiopia gree tea a ciamo tea. caffeie at 30 a boi ig temperature, the percetage C of caffeie at the boi i g temperature is greater. Caffeie cotet i tea eaves reporte i this resea rch is i the rage of va ues reporte (1 to 5%). As it ca be see i this rese arch, the maximum cotet of caffeie is 2.36% which is a most simi ar to the re su ts obtaie for the caffeie cotets of markete tea by Misra et a . (2009), usig HPLC i which the maximum cocetratio of caffeie i tea samp es of mar kete tea was 2.145%. Abu et a . (2006) a so reporte usig HPLC that, the caf feie cotet for b ack tea was 3.34% a gree tea was 2.44%. Precisio of the metho was etermie by three rep icatios of each samp e. The precisio (%RSD) of the rep icatios was fou to be ess tha two which is iicative of a precise meth o. The UVvisib e absorbace rage of caffeie is from 243 m to 302 m. But fo r ciamo tea eaves the rage is from 243 m to 318 m (Figure 11). This iic ates that the aitio of ciamo to tea eaves icreases the wave egth rage by 16 m. Coc usio A terative methos were eve ope to etermie caffeie co tet i rea tea eaves usig UV visib e spectrometer. I this research, optica trasitio 8 Afr. J. Pure App . Chem. properties of pure caffeie a caffeie cotet of rea tea eaves at two iffe ret temperatures were aa yze a the resu t agrees with the iterature va ues of other aa ytica methos. To make the resu t more re iab e, the experimets were repeate three times a the average va ues were take. The resu ts at two ifferet temperatures revea that, caffeie is better extracte at the boi ig temperature tha that of 30 a more C extractio time. But the iffereces were very sma . Therefore, it ca be sai that, by takig more extractio time a a t higher temperatures, more caffeie ca be extracte tha at boi ig temperatur e a a ess extractio time. The resu t of the experimet iicates that for ci amo tea, the wave egth rage icreases to the oger wave egth. This iica tes that, aitio of ciamo to tea eaves icreases the wave egth rage. The broaeig of absorptio spectrum whe tea is b ee with ciamo ca be exp aie by the fact that i this tea, caffeie a compous i ciamo wi have ifferet atura frequecies. REFERENCES Abu M, Kazi FA, Zaia A, Zakir H (2006). Determiatio a charact erizatio of caffeie i tea, coffee, a soft riks by so i phase extractio a high performace iqui chromatography (SPE HPLC). Malaysia J. Chm., 8(0): 1, 045051. Ala M, Iris M (2004). Th Empir of Ta. Th ovrlook prss. ISBN 1 58567 493- 1, p. 32. Amra P, Mojca S, Zljko K, Brd W, Frak O, Sabi G (2006 ). Extractio of activ igrdits from gr ta (camllia sisis ): Extract io fficicy of major catchis ad caffi. Food Chm., 96: 597 605. Btt AW, Boi KB (2001). Th world of caffi: Th scic ad cultur of th Wor lds Most Popular Drug. Roultldg, ISBN 0415927226, p. 228. Bolto S, Null G (1981). Caffi, psychological ffcts, us ad abus. Orthomo l. Psychiatr., 10(3): 202 211. Koms D, Horzic D, Blscak A, Kovacvic GK, Boljak A (2009). Dtrmiatio of caffi cotts i ta ad mat ta by usig diff rt mthods. Czch J. Food Sci., 27: 69. Gutr G, Tua V-D (2003). Hadbook of spctroscopy. Wily-VCH Vrlag GmbH ad CO. KGa A, Wihim ISBN 3 527- 29782, 1: 0, 69. Guzi A (2002). Drivativ spctrophotomtr dtrmiatio of caffi i som Bvrags. Turk. J. Chm., 26: 295302. Hicks MB, Hsih Y-HP, Bll LN (19 96). Ta prparatio ad its ifluc o mthylxathi coctratio. Food Rs . It., 29(3-4): 325330. Jffrso JW (1998). Lithium trmor ad caffi itak: Two cass of drikig lss ad shakig mor, J. Cli. Psychiatr., 49: 72. Johs o GD, Fatis M, Sok D (1988). A survy of caffi us ad associatd sid f fcts i a collg populatio. J. Drug Educ., 18(3): 211. Kirmr DA (1988). Caff i us ad abus i psychiatric clits. J. Psychosoc. Nurs. Mt. Halth Srv ., 26: 20. Liptay W (1969). Elctrochromism ad solvatochromism. Agwadt Chm i, It. Ed., 8(3): 177-187. Michal JL (1999). Molcular spctroscopy. USA: Pr tic- Hall, Ic., pp. 160-161. Misra H, Mhta D, Mhta BK, Soi M, Jai DC (200 9). Study of xtractio ad HPTLC UV mthod for stimatio of caffi i markt d ta grauls. It. J. Gr Pharm., 3(1): 4751.