Supercritical CO2 extraction of essential oils from Thymus vulgaris

SUPERCRITICAL CO₂ EXTRACTION OF ESSENTIAL OILS FROM Thymus vulgaris

S.A.B. Vieira de Melo^1*, G.M.N. Costa², R. Garau³, A. Casula³ and B. Pittau3

1ITP - UNIT - Tiradentes University, Av. Ivo do Prado, 1162, Bairro 13 de Julho, 49015-070,

Fax +(55) (79) 211-4033, Aracaju - SE, Brazil E-mail: [e-mail]

²DEQ - EP - Federal University of Bahia, R. Aristides Novis 2, Federação,

40210-630, Salvador - BA, Brazil.

³Florys SpA, trav. 2^a strada est Assemini (Ca), Italy.

(Received: February 10, 2000 ; Accepted: May 30, 2000)

Abstract - Supercritical CO₂ extraction of essential oil from Thymus vulgaris leaves was studied using experimental data recently obtained in the Florys S.p.A. laboratory. Mass transfer coefficients in the supercritical and solid phases from extraction curves at 40°C and 20 MPa were evaluated using a mathematical model based on the local adsorption equilibrium of essential oil on lipid in leaves. The adsorption equilibrium constant was fitted to these experimental data, and internal and external mass transfer resistances were calculated, allowing identification of the mechanism controlling the extraction process.

Keywords: mathematical modeling, supercritical CO₂ extraction, essential oils, Thymus vulgaris

INTRODUCTION

Essential oils are widely used as active principle and flavoring agent in pharmaceutical, cosmetic and food industries. Traditional methods for isolating essential oils from plant materials, such as steam distillation and solvent extraction, have some drawbacks due to the heat instability of essential oil and the presence of residual organic solvent in the extract. Thus, the use of supercritical fluids for extraction of essential oils has received increasing attention as an alternative to these traditional techniques (Stahl et al., 1987).

Supercritical carbon dioxide has such attractive properties (it is nontoxic, inexpensive, odorless, colorless, nonflammable and has near ambient critical temperature, low viscosity and high diffusivity compared to liquids) that it has become the preferred solvent in the processing natural materials (McHugh & Krukonis, 1994).

Thymus vulgaris (thyme), as vegetable matter or as extract, is widely used in pharmaceutical, cosmetics and food industries. It is also highly regarded as a medicinal herb with antispasmodic, expectorant and flatulence-reducing functions. In western medicine the main application has been in the treatment of digestive complaints and respiratory problems and in the prevention and treatment of infection.

The biological activity of thyme essential oil is related to its major constituents, namely thymol and carvacrol. Thymol has been shown to have antibacterial, antifungal and anthelmintic effects (Neeman et al.,1995), while carvacrol has been investigated for its bactericidal effect (Ultee et al., 1998). Moreover, due to the phenolic structures of the two principal constituents, the essential oil has shown significant evidence of antioxidant function (Daphevicius et al., 1998).

Extraction of materials from solid natural matrixes involves an internal mass transfer, i.e., dissolution and diffusion of the solute in cellular material, and an external mass transfer around the solid particle, i.e., diffusion of the solute through the supercritical fluid-rich phase to be carried away by bulk flow. Several attempts have been made to model the supercritical extraction process (Sovová et al., 1994; Reverchon & Polleto, 1996; Roy et al., 1996; Reverchon, 1996; Akgerman & Madras, 1994; Esquivel et al., 1996), and a comprehensive discussion of modeling aspects is presented elsewhere (Reverchon, 1997).

In this paper, a mathematical model developed to describe supercritical CO₂ extraction of essential oil from peppermint leaves (Goto et al., 1993) was applied to study extraction of Thymus essential oil with supercritical carbon dioxide. Extraction curves were obtained from experiments carried out in the Florys S.p.A. laboratory at 40°C and 20 MPa, for three different particle sizes and three different flow rates. The adsorption equilibrium constant was fitted to these experimental data and internal and external mass transfer resistances were calculated. Comparison between experimental data and the calculated results showed that the model is well able to describe SC-CO₂ extraction of thyme oil. Intraparticle diffusion was suggested as the mechanism controlling the extraction process.

EXPERIMENTAL PROCEDURE

Thymus vulgaris was collected in the Northeast of Italy and was supplied as leaves and blossoms dried at ambient temperature. According to the official Italian Pharmacopoeia (F.U. IX), the dried samples as supplied were preliminarily characterized in terms of water content by weight loss in an oven at 105°C, showing a moisture percentage (wt/wt) of 14.2 %.

To eliminate the effect of water content in the extractions from the plant matter, fractions of thyme samples were treated in a freeze-drier. The lyophilizing process was carried out using a Edwards Alto Vuoto S.p.A. mod. Minitast 2000, treating 100 g samples for 90 min at temperatures ranging from –30 to 5 °C and 38.9 m B vacuum. The vegetable materials obtained, with a 10.5 % residual water content, were used both without further treatment and after sifting. The latter procedure allowed collecting of fractions characterized by particle size. Thyme extracts were obtained on laboratory scale by supercritical CO₂ extraction (SCF-CO₂) by means of a commercial Suprex Corporation (Pittsburg, PA USA) apparatus, provided with a 50 ml volume extraction vessel. In semi-batch processing the fluid leaving the extractor is expanded to room pressure by a variable flow thermostat valve and the solute is collected in a cryogenic trap. Each extraction carried out on laboratory scale, was performed with 10–12 g of dried material placed in the extractor vessel. As shown in Table 1, the experiments were run setting SCF CO₂ at 200 bar and 40 °C; the flow rate was set ranging from 0.5 to 2.0 ml for extraction times varying between from 90 and 240 min. The extracts were weighed and diluted in 4 ml of ethanol. Yields in this work were reported as the weight of dried drugs, as determined at 105°C.

MATHEMATICAL MODEL

The model used in this work to describe supercritical carbon dioxide extraction of Thymus essential oil [13] considers Thymus leaves to be a porous solid containing essential oils associated with lipids located in vacuoles. Internal mass transfer of essential oils occurs by desorption from these vacuoles and diffusion of essential oils dissolved in supercritical CO₂ to the external surface of the leaf. External mass transfer of essential oil consists of diffusion through the supercritical rich phase to be carried away by bulk flow. Negligible axial dispersion, constant solvent density and constant flow rate along the bed are assumed in order to model the essential oil as a single component. Essential oil Thymol is considered the major component of Thymus leaves.

A detailed development of this model is presented elsewhere (Goto et al, 1993). In terms of dimensionless variables, the mass balance for the solute in the bulk solvent and the mass balance for the solute in the pores are given, respectively, by

(1)

and

(2)

where X is the dimensionless solute concentration in the effluent, defined as the ratio of solute concentration in CO₂ in bed void volume to total solute concentration; X_s is the dimensionless solute concentration in the pores defined as the ratio of solute concentration in the leaves to total solute concentration; q is dimensionless time, defined as the ratio of time to residence time (residence time is defined as the total bed volume divided by the volumetric flow rate of supercritical fluid under extraction conditions); f is the dimensionless mass transfer coefficient, defined as the combined mass-transfer coefficient times residence time; a and b are the void fractions in bed and leaf porosity, respectively; and K is the equilibrium adsorption constant, defined as the fitting parameter for the model.

Initial conditions for Eq. (1) and (2) are

(3)

and

(4)

Analytical solution of Eq. (1)-(4) gives

(5)

where

(6)

(7)

(8)

(9)

(10)

The reduced extraction yield (defined as the cumulative fraction of solute extracted) up to dimensionless time q is given by

(11)

RESULTS AND DISCUSSION

As stated previously, several mathematical approaches proposed in the literature consider the leaf a porous medium, but without experimental data it is impossible to determine which mass transfer mechanism predominates.

The mass transfer model described above was used to simulate experimental conditions obtained for SC-CO₂ extraction of thyme essential oil at 200 bar and 40°C, as shown in Table 1. Equilibrium adsorption parameter K should be fitted to the experimental data for times of extraction from 90 to 240 minutes, but experimental observation led to some uncertainty regarding yield at 240 minutes and this value was not included in the fitting procedure. Na extraction time of 180 minutes was considered to be the condition for a yield of 95% , but more experimental data should be collected to confirm this evidence. As seen in Eq. 11, the solution of the model is expressed as reduced yield, i.e., the ratio of yield to maximum yield. In the experimental data yields at 90, 120 and 180 minutes were divided by the maximum yield value (0.202), giving the reduced yield values of 0.52, 0.76 and 0.95, respectively; an experimental dead time of 50 minutes was assumed. Parameter fitting to these data points gave a K value equal to 0.94 for a mean relative deviation in reduced yield of about 2.32%, as shown in Fig. 1. It is important to note that equilibrium adsorption constant K, as originally used by Goto et al. (1993), was fitted to experimental data as a function of pressure and temperature. In the present ongoing work, there has not been sufficient experimental data up to now to fit this parameter under different conditions of pressure and temperature. If one uses 10 as the K value (the value of Goto et al. (1993) at 200 bar and 40°C), the mean relative deviation in reduced yield becomes about 28.7%. Table 1 also illustrates that increasing particle size decreases yield, as expected, because for higher values of leaf size the intraparticle diffusion path becomes longer. It can also be seen that increasing the flow rate increases yield.

CONCLUSIONS

Supercritical carbon dioxide extraction of essential oil from Thymus vulgaris leaves was studied in the Florys S.p.A. laboratory at 40 °C and 20 MPa for three different particle sizes and three different flow rates. A mathematical model, proposed in the literature, based on the local adsorption equilibrium of essential oil on lipid in leaves was used to calculate internal and external mass transfer resistances. The adsorption equilibrium constant was fitted to these experimental data, and preliminary results suggested that the extraction process is controlled by mass transfer resistance due to intraparticle diffusion. Experimental data points are currently being measured under other conditions of pressure and temperature to confirm these conclusions.

Neeman, I., Tabak, M. and Armon, R. usp56472695.

• Akgerman, A. and Madras, G., in Supercritical Fluid Science and Technology, E. Kiran and J.M.H. Levelt Sengers (eds.), Kluwer Academic Publishers, 669 (1994).
• Daphevicius, A., Venskutonis, R., Van Beek, T. and Linssen, J.P.H.; J. Sci. Food Agric, 77, 140 (1998)
• Esquivel, M.M., Sousa, C.L., Ribeiro, M.A. and Bernardo-Gil, M.G., In High Pressure Chemical Engineering, Ph.R. Rohr and Ch. Trepp (eds.), Elsevier, 525 (1996)
• Goto, M., Sato, M. and Hirose, T., J. Chem. Eng. Japan, 26 (4), 401 (1993).
• Mchugh, M.A. and KRUKONIS, V., Supercritical Fluid Extraction Principles and Applications, 2^nd Ed; Butterworths, Boston, (1994)
• Reverchon, E., AIChE Journal, 42 (6), 1765 (1996).
• Reverchon, E., J. of Supercritical Fluids, 10, 1 (1997)
• Reverchon, E. and Polleto, M., Chem. Eng. Sci, 51 (15), 3741 (1996).
• Roy, B.C., Goto, M. and Hirose, T., Ind. Eng. Chem. Res, 35, 607 (1996).
• Sovová, H., Komers, R., Kucera, J. and Jez, J., Chem. Eng. Sci, 49 (15), 2499 (1994)
• Stahl, E., Quirin, K.-W. and Gerard, D., Dense Gases for Extraction and Refining, Springer Verlag, Berlin, (1987)
• Ultee, A., Gorris, L.G.M. and Smid, E.J., J. Applied Microbiology, 85 (2), 211 (1998)

^*

To whom correspondence should be addressed

Publication Dates

• Publication in this collection
  18 Oct 2000
• Date of issue
  Sept 2000

History

• Accepted
  30 May 2000
• Received
  10 Feb 2000