SciELO - Scientific Electronic Library Online

vol.17 issue3Phase behaviour of sterols and vitamins in supercritical CO2 author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632On-line version ISSN 1678-4383

Braz. J. Chem. Eng. vol.17 n.3 São Paulo Sept. 2000 



M.D.A. Saldaña1, R.S. Mohamed1 and P. Mazzafera2
1School of Chemical Engineering, Universidade Estadual de Campinas,
P. O. Box 6066, 13083-970, Campinas - SP, Brazil,
2Biology Institute, UNICAMP, Campinas (SP), Brazil.


(Received: February 8, 2000 ; Accepted: April 6, 2000)



Abstract - Methylxanthines are alkaloids found in natural products such as tea, coffee and guaraná. These alkaloids are commonly used in cola drinks and pharmaceutical products due principally to their stimulant and diuretic effects on the human organism. In this work, experimental data on the supercritical CO2 extraction of caffeine, theophylline and theobromine from herbal maté tea, a beverage traditionally consumed by the gauchos of southern Brazil, the Argentine, Paraguay and Uruguay, were obtained using high pressure extraction equipment that allows adequate control of temperature and pressure. The continuous extraction/fractionation of maté tea leaves, Ilex paraguariensis in natura using carbon dioxide was carried out at 313.2 and 343.2 K and pressures of 13.8 and 25.5 MPa. Extraction/fractionation curves revealed the large influence of temperature and pressure on extraction yield. CO2 was also found to show a higher selectivity for caffeine than for theophylline and theobromine.
Keywords: Supercritical extraction, Ilex paraguariensis, caffeine, theobromine, theophylline.




Caffeine (1,3,7-trimethylxanthine), theophylline (1,3-dimethylxanthine) and theobromine (3,7-dimethylxanthine) are some of the natural components of maté tea leaves (Suzuki and Waller, 1988; Clifford and Ramirez, 1990; Saldaña et al., 1999), coffee beans (Brunner, 1984; Saldaña, 1997), guaraná seeds (Mehr et al., 1996) and cocoa nuts (Li and Hartland, 1992). In the human organism, these alkaloids stimulate the central nervous, muscular and circular systems (James, 1991). Decaffeination of natural products is economically attractive as it results in higher value decaffeinated products and caffeine, a valuable by-product commonly used in soft drinks and pharmaceutical products (Mazzafera and Carvalho, 1991).

Maté, Ilex paraguariensis, is a native plant of South America, whose taxonomy, chemical composition and other characteristic properties are shown in Table 1. A beverage prepared by the infusion of dry maté tea leaves is traditionally consumed by gauchos of southern Brazil, the Argentine, Paraguay and Uruguay (Alikaridis, 1987; Tormen, 1995). A careful inspection of the chemical constituents of the Ilex species (Table 1) demonstrates the reason for the current and successful use of this natural product as a stimulant, an antirheumatic and a diuretic. Nevertheless, a high intake of maté tea provoke irritability and loss of sleep, can and even may cause cerebral depression, nervous tremors and numbness. Several laboratory studies (Wang et al., 1994; Inagake et al., 1995; Hasegawa et al., 1995; Katiyar et al., 1995) have also shown that polyphenols found in this tea can inhibit the formation and growth of tumors.



Caffeine can be removed from maté tea leaves using organic solvents (dimethyl chloride) or water (Saldaña, 1997). However, the use of chemical solvents involves the risk of leaving toxic residues in the products extracted while the use of water results in a nonselective extraction and the loss of valuable flavor components (Mazzafera and Carvalho, 1991). Carbon dioxide, which has a low critical temperature (31° C) and is nontoxic and relatively inexpensive, has become a universally attractive alternative solvent in the extraction of natural products (McHugh and Krukonis, 1994; Saldaña, 1997).

While there are many patents for the use of supercritical CO2 as a solvent to extract caffeine from coffee beans and Camelia sinensis tea leaves (Brunner, 1984; Peker et al., 1992; McHugh and Krukonis, 1994; Saldaña et al., 1997), little is known about the extraction of methylxanthines from maté tea, Ilex paraguariensis (Saldaña et al., 1999). The objective of this work was to obtain extensive experimental data on the extraction/fractionation of methylxanthines (caffeine, theophylline and theobromine) from maté tea, Ilex paraguariensis in natura using supercritical CO2.




Caffeine, theobromine and theophylline, 99% purity, were purchased from Sigma Chemical Co. (St. Louis, USA). The supercritical fluid, dry carbon dioxide, 99.9% purity was supplied in the liquid phase by White Martins Industrials Gases S.A (Campinas, Brazil). Maté, Ilex paraguariensis tea leaves, in natura were obtained from the Experimental Campus of the Department of Plant Physiology - Biology Institute (UNICAMP-Campinas, Brazil).



(a) Sample Preparation and Analysis

Prior to extraction, whole maté tea leaf samples were separated and classified by hand (4.5cm x 9cm). Moisture content of the samples was determined by drying at 80oC for 24 h.

Compositions of the methylxanthines in the extracted products were determined using a reversed-phase High Performance Liquid Chromatography apparatus (Shimadzu, Japan) with UV monitor set at 280 nm and a Nucleosil C18 column (Supelco, 4 x 150 mm, 5 m m). The isocratic solvent used was a solution of 40% methanol in 0.5% aqueous acetic acid. A flow rate of 1mL min-1 was employed.

(b) Experimental Apparatus

The experimental apparatus used was a semicontinuous-flow, high-pressure system purchased from Autoclave Engineers (Erie, PA, USA), that had been designed for working at pressures up to 37 MPa at 200 oC (Figure 1). This apparatus is the same as that used earlier by Neves et al. (1996) and Saldaña et al. (1999). The major components of this apparatus include positive liquid displacement pumps for solvent delivery, high-pressure extraction vessels and three separator flasks in series. Flow rates and accumulated gas volumes passing through the apparatus were controlled with micrometering valves from Autoclave Engineers (Erie, PA, USA) and measured with a flow computer measuring device from EG&G Instrument Flow Technology (Phoenix, AR, USA). Heating tapes were used to maintain constant temperature in valves in order to prevent freezing of solvents or solid solute precipitation following depressurization and in the extraction section. Pressure in both extractors was monitored with a digital transducer system with a precision of ± 0.03 MPa from Heise Series 901A RTS, Dresser Industries (Stratford, CT, USA). Extractor temperatures were controlled to within ± 0.5 oC.



(c) Determination of Caffeine, Theophylline and Theobromine During Supercritical Carbon Dioxide Extraction From Maté Tea Leaves In Natura

In a typical experiment, liquid CO2 was pumped into the maté tea leaf sample (22.5g with 60% moisture content) in extractor E1 until the specified extraction pressure was reached at the specified temperature. After a three-hour period to reach equilibrium, the saturated supercritical fluid was depressurized whilst pressure was maintained constant in the extractor and a precipitated fraction was collected in the tarred separator flasks placed in a cooling bath maintained at 0 oC by carbon dioxide passing through the apparatus. Collected samples were weighed and analyzed for methylxanthines using HPLC. The three separator flasks were replaced at specified time intervals as fractions were collected. Tubings and valves throughout the apparatus were cleaned with chloroform at the end of each experiment to assure good quantification.



Calibration Curves

To analyze the extracts obtained after fractionation, it was necessary to establish calibration curves for five standard concentrations (0.0001; 0.001; 0.01; 0.1 and 0.2 mg mL-1). This interval covers the entire concentration range found in the extracted samples to be analyzed. Figure 2a shows the retention times (tR) for theobromine, theophylline and caffeine at approximately 3.5, 4.3 and 6 minutes, using a mobile phase 40% methanol in water and Figure 2b shows their chemical structures. The shorter retention times obtained for theobromine can be attributed to the better interaction between the mobile phase and theobromine as a result of the hydrogen radical in position 1 and the two methyl radicals in positions 3 and 7, in comparison to the caffeine molecule which has three methyl radicals in positions 1, 3 and 7 (Figure 2b). Baltassa et al. (1984) used the same mobile phase methanol-water (40-60 v/v) and obtained peaks at approximately 4, 4.7 and 6 min for theobromine, theophylline and caffeine, respectively.





Extraction of Caffeine, Theobromine and Theophylline From Maté Tea Leaves Using Supercritical CO2

Results for the fractionation of maté tea leaves in natura, performed at 13.8 and 25.5 MPa and 313.2 and 343.2 K are shown in Figure 3, where the extraction rate of methylxanthines at a lower temperature and pressure (313.2 K and 13.8 MPa) was found to be almost constant (as shown by the approximately linear behavior). Under these conditions, extraction was controlled by solubility effects with no evidence of mass transfer limitations. The low extraction velocity and yield required the use of 3.74 kg of CO2 to obtain 25.22 mg of caffeine. Extraction/fractionation at 343.2 K and 25.5 MPa presented higher rates of caffeine removal in the first ten fractions. The extraction rates, however, decreased as the cumulative amounts of CO2 increased as shown by the changes in the slope of the extraction curve. The total amount of caffeine extracted at 343.2 K and 25.5 MPa is approximately 2.7 times the amount obtained at 313.2 K and 13.8 MPa.



The three extraction regions, solubility, intermediate and diffusion-controlled, which were first identified by Hedrick et al. (1992), were evident for extractions at 343.2 K and 25.5 MPa (Figure 3). The region dominated by solubility effects was observed for the first 10 fractions. The intermediate region was identified by the eleventh to sixteenth fraction, while the diffusion-controlled extraction was represented by fractions numbered 17 to 40. When the 16th fraction was collected, 84.5% of the caffeine had been extracted. The total quantity of caffeine obtained from the 9g of whole leaves of dry maté tea was 67.27 mg.

A qualitatively similar behavior was observed for the extraction of theobromine and theophylline during the fractionation of maté tea leaves in natura under these very same conditions (Figures 4 and 5). When comparing the extraction curves in Figures 3, 4 and 5, we observe that the last fractions were richest in theobromine and theophylline and increased with the cumulative amount of CO2, showing it to be an interesting scheme for the separation of the three methylxanthines, obtaining fractions with varying concentrations of caffeine, theobromine and theophylline. The selectivity of CO2 for caffeine, compared to those for theobromine and theophylline, is also very evident.





At 13.8 MPa and 313.2 K, the extraction velocity and yield were very low, requiring the use of 3.74 kg of CO2 to obtain merely 0.69 mg of theobromine (Figure 4). At 343.2 K and 25.5 MPa, 70.4% of the theobromine had been extracted when the 16th fraction was collected. The total quantity of theobromine obtained from the 9g of whole leaves of dry maté tea was 2.33 mg, which was approximately 3.4 times the amount of theobromine extracted at 13.8 MPa and 313.2 K.

Supercritical carbon dioxide extraction of maté tea leaves at 343.2 K and 25.5 MPa resulted in the removal of approximately 4.6 times the amount of theophylline obtained under the extraction conditions of 313.2 K and 13.8 MPa (Figure 5).

The total quantity of methylxanthines obtained at 25.5 MPa and 343.2 K from the 9g of whole leaves of dry maté tea were 67.27, 2.33 and 0.27 mg of caffeine, theobromine and theophylline, respectively. These quantities amount to 7474.4, 258.9 and 30 mg of caffeine, theobromine and theophylline per kg of dry maté tea, respectively. The corresponding values reported by Mazzafera (1994) and shown in Table 1 for new leaves from branches with fruit are 8375+251, 768+3 and 209+5 mg of caffeine, theobromine and theophylline per kg of dry maté tea leaves.

Figure 6 shows a comparison between the results of caffeine extraction from two different matrixes of Ilex paraguariensis: in natura, whole maté leaves (this work) and ground commercial maté tea (10g, 10% humidity), reported earlier by Saldaña et al. (1999). The two samples contain the same amount of dry material (9g) with two different moisture contents (10% and 60%). The results shown in Figure 6 clearly reveal that the higher caffeine extraction rate for ground commercial maté tea at the early stages of extraction, as expected due to the absence of mass transfer resistance and plant matrix interference with the extraction, is determined mainly by solubility effects. Figure 6 also shows that higher amounts of caffeine were extracted from whole leaves, which could possibly be attributed to the higher moisture content of the whole leaf sample.



The binary solubilitites, 2047.6, 14.13, 7.15 mg of caffeine, theophylline and theobromine per kg CO2, respectively (Saldaña et al., 1999), are substantially greater than those obtained during extraction at 25.5 MPa and 343.2 K of whole maté tea leaves (48.46, 1.51 and 0.14 mg of caffeine, theobromine and theophylline per kg CO2, respectively). This can be attributed to several factors, including mass transfer limitations caused by the plant matrix, and points to the difficulty of using binary data to predict the behavior of complex plant systems.



(a) Using carbon dioxide, it was possible to extract caffeine, theobromine and theophylline from maté tea, Ilex paraguariensis in natura.

(b) Extractable material was fractionated into several fractions of varying concentrations of methylxanthines.

(c) Extraction curves revealed the existence of three regions: solubility, intermediate and diffusion-controlled, in agreement with the literature.

(d) Rates of extraction from ground material were higher than those from whole leaves observed due to the absence of mass transfer resistance and the interference of the plant matrix.

(e) The higher moisture content of the extracted material resulted in improved total removal of caffeine.

(f) Solubilities in CO2/methylxanthine binary systems were greater than those found during the extraction of maté tea leaves, illustrating the difficulty of using binary data to predict complex natural product systems.



The authors acknowledge the financial assistance of FAPESP and CNPq, Brazilian research funding agencies.



Alikaridis, F., Constituintes naturais do Ilex espécies, Journal of Ethnopharmacology, 20, 121-144 (1987).        [ Links ]

Baltassa, F., Carbour, N. and Ferry, S., Étude du contenu purique de drogues a caféine: I.-Le Maté: Ilex paraguaryensis Lamb., Plantes médicinales et phytothérapie, 18(4), 195-203 (1984).        [ Links ]

Brunner, G., Mass transfer from solid material in gas extraction, Ber. Bunsenges. Phys. Chem., 88, 887-891 (1984).        [ Links ]

Clifford, M. and Ramirez-Martinez, J.R., Chlorogenic acid and purine alkaloids contents of maté (Ilex paraguariensis) leaf and beverage, Food Chemistry, 35, 13-21 (1990).        [ Links ]

Hasegawa, R., Chujo, T., Sai-kato, K., Umemura, T., Tanimura, A. and Kurokawa, Y., preventive effects of green tea against liver oxidative DNA damage and hepatotoxicity in rats treated with 2-nitropropane, Fd. Chem. Toxic., 33 (11), 961-970 (1995).        [ Links ]

Hedrick, J.L., Mulcahey, L.J. and Taylor, L.T., fundamental Review: Supercritical Fluid Extraction, Mikrochim Acta, 108, 115-132 (1992).        [ Links ]

Inagake, M., Yamane, T., Kitao, Y., Oya, K., Matsumoto, H., Kikuoka, N., Nakatani, H., Takahashi, T., Nishimura, H. and Iwashima, a., inhibition of 1,2-dimethylhydrazine-induced oxidative DNA damage by green tea extract in rats, Jpn. J. Cancer Research, 86, 1106-1111 (1995).        [ Links ]

James, J.E., Caffeine and health. Academic Press Inc., San Diego (1991).        [ Links ]

Katiyar, S.D., Elmets, C.A., Agarwal, R. and Mukhtar, H., protection against ultra violet-b radiation-induced local and systemic suppression of contact hypersensitivity and edema responses in C3H/HeN mice by green tea polyphenols, Photochemistry and Photobiology, 62 (5), 855-861 (1995).        [ Links ]

Li, S. and Hartland, S., Influence of co-solvents on solubility and selectivity in extraction of xanthines and cocoa butter from cocoa beans with supercritical CO2, J. Supercritical Fluids, 5, 7-12 (1992).        [ Links ]

Mazzafera, P., Caffeine, theobromine and theophylline distribution in Ilex paraguaryensis, R. Bras. Fisiol. Veg., 6 (2), 149-151 (1994).        [ Links ]

Mazzafera, P., Maté drinking: caffeine and phenolic acid intake, Food Chem., 60, 67-71 (1997).        [ Links ]

Mazzafera, P. and Carvalho, A., A cafeína do café, Documentos IAC, 25, 1-22 (1991).        [ Links ]

McHugh, M.A. and Krukonis, V.J., Supercritical Fluid Extraction: Principles and Practice, Butterworths Publishers, Boston MA (1994).        [ Links ]

Mehr, C.B., Biswal, R.N., Collins, J.L. and Cochran, H.D., Supercritical carbon dioxide extraction of caffeine from guaraná, J. of Supercritical Fluids, 9, 185-191 (1996).        [ Links ]

Neves, G.M., Mohamed, R.S. and Kieckbusch, T.G., Reduction of cholesterol level in dairy products using supercritical carbon dioxide, Proceedings of the International Conference on Solvent Extraction, Melbourne, Australia, 999 (1996).        [ Links ]

Peker, H., Srinivasan, M.P., Smith, J.M. and McCoy, B.J., Caffeine extraction rates from coffee beans with supercritical carbon dioxide, AlChE J., 38(5), 761-770 (1992).        [ Links ]

Saldaña, M.D.A., Extração de cafeína, trigonelina e ácido clorogênico de café com CO2 supercrítico, M.Sc. thesis, Universidade Estadual de Campinas, Campinas, Brazil (1997).        [ Links ]

Saldaña, M.D.A., Mazzafera, P. and Mohamed, R.S., Extraction of caffeine, trigonelline and chlorogenic acid from Brazilian coffee beans with supercritical fluids, The 4th International Symposium on Supercritical Fluids, May 11-14, Sendai, Japan, 219-222 (1997).        [ Links ]

Saldaña, M.D.A., Mazzafera, P. and Mohamed, R.S., Extraction of purine alkaloids from maté plant (Ilex paraguariensis) using supercritical CO2, Journal of Agric. and Food Chemistry, 47(9), 3804-3808 (1999).        [ Links ]

Suzuki, T. and Waller, G., In Modern methods of plant analysis: Analysis of Non Alcoholic Beverages, Linskens, H.F. and Jackson, J.F., Eds., Springer Verlag: Heidelberg, 184-220 (1988).        [ Links ]

Tormen, M.J., Economia ervateira brasileira, In erva maté: biologia e cultura no cone sul, eds H. Winge, A.G. Ferreira, J.E.A. Mariath and L.C. Tarasconi, Editora da Universidade Federal de Rio Grande do Sul, Porto Alegre, Brazil, 27-40 (1995).        [ Links ]

Wang, Z.Y., Huang, M., Lou, Y., Xie, J., Reuhl, K., Newmark, H., Ho, C., Yang, C. and Conney, a., inhibitory effects of black tea, green tea, decaffeinated black tea, and decaffeinated green tea on ultraviolet b -light-induced skin carcinogenesis in 7,12-dimethylbenz[a]anthracene-initiated SKH-1 mice, Cancer Research, 54, 3428-3435 (1994).         [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License