SciELO - Scientific Electronic Library Online

vol.28 issue2Maximum thermodynamic efficiency problem in batch distillationErratum author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632

Braz. J. Chem. Eng. vol.28 no.2 São Paulo Apr./June 2011 



Extraction of citronella (Cymbopogon nardus) essential oil using supercritical co2: experimental data and mathematical modeling



C. F. SilvaI; F. C. MouraI; M. F. MendesII, *; F. L. P. PessoaI

IUFRJ, Departamento de Engenharia Química, Rio de Janeiro - RJ, Brazil
IIUFRRJ, Departamento de Engenharia Química, Seropédica - RJ, Brazil. Universidade Federal Rural do Rio de Janeiro, Phone: + (55) (21)3787-3742, Fax: + (55) (21) 3787-3750, Cidade Universitária, Instituto de Tecnologia, Departamento de Engenharia Química, BR 465, Km 7, CEP 23890-000, Seropédica - Rio de Janeiro - RJ, Brazil. E-mail:




Citronella essential oil has more than eighty components, of which the most important ones are citronellal, geranial and limonene. They are present at high concentrations in the oil and are responsible for the repellent properties of the oil. The oil was extracted using supercritical carbon dioxide due to the high selectivity of the solvent. The operational conditions studied varied from 313.15 to 353.15 K for the temperature and the applied pressures were 6.2, 10.0, 15.0 and 180.0 MPa. Better values of efficiency of the extracted oil were obtained at higher pressure conditions. At constant temperature, the amount of extracted oil increased when the pressure increased, but the opposite occurred when the temperature increased at constant pressure. The composition of the essential oil was complex, although there were several main components in the oil and some waxes were presented in the extracted oils above 10.0 MPa. The results were modeled using a mathematical model in a predictive way, reproducing the extraction curves over the maximum time of the process.

Keywords: Citronella oil; Geraniol; Empirical model; Extraction modes.




Essential oils are concentrated essences extracted from different parts of plants, containing hundreds of substances, but typically with the prevalence of one, two or three of them that really characterize the fragrance (Mendes et al., 2007).

Industrial interest in essential oils is due to their application as fragrances in perfumes, as flavor additives for use in food products or even as pharmaceutical products. In the case of citronella species, for example, the components present in the oil are responsible for the desirable repellent characteristics of the plant against mosquitoes (Katz et al., 2008; Simic et al., 2008). Numerous plants and derived products, in particular essential oils, have been investigated and described as potential natural sources of insect repellent. Trongtokit et al. (2005) compared the repellent efficiency of 38 essential oils against mosquito bites, including the species Aedes aegypti. Among other essential oils, citronella oil was the most effective and provided 2 hours of repellency. Wong et al. (2005) studied five commercial plant extracts, including citronella, and showed that it is effective in deterring the infestation of cartons containing muesli and wheat germ by red flour beetles. Moreover, Olivo et al. (2008) proved that citronella oil has other effects, such as the control of cattle ticks, the most important active principles being citronelal and geraniol. Nakahara et al. (2003) studied the chemical composition of citronella oil and its antifungal activity. The crude essential oil markedly suppressed the growth of several species of Aspergillus, Penicillium and Eurotium. The most active compounds among the 16 volatiles examined, consisting of 6 major constituents of the essential oil and 10 other related monoterpenes, were citronellal and linalool.

Currently, there are plant-based insect repellents on the market that contain essential oils from one or more of the following plants: citronella (Cymbopogon nardus), cedar (Juniper virginiana), eucalyptus (Eucalyptus maculata), geranium (Pelargonium reniforme), lemon-grass (Cymbopogon excavatus), peppermint (Mentha piperita), neem (Azadirachta indica) and soybean (Neonotonia wightii). Most of these essential oil-based repellents tend to give short-lasting protection for less than 2 h (Choochote et al., 2007). Citronella oil has demonstrated good efficacy against 44 mosquitoes in concentrations ranging from 0.05 % to 15 % (w/v) alone or in combination with other natural or commercial insect repellent products (Sakulku et al., 2009 apud Fradin, 1998). Olivo et al. (2008) apud Shasany et al. (2000) confirmed that this characteristic of the oil is due to the presence of four main components, citronelal, eugenol, geraniol and limonene.

Reis et al. (2006) studied the effect of drying at different temperatures (323.15, 333.15 and 343.15 K) on the composition of the essential oil of citronella extracted by hydrodistillation. The results showed a slight change in the composition for some components, but the drying did not influence the composition of the main components. The yield of the extraction was approximately 9.4 %, defined as the ratio between the mass of essential oil extracted and the initial mass of citronella used in the experimental tests.

According to Reverchon and De Marco (2006), the extraction of compounds from natural sources is the most widely studied application of supercritical fluids (SCFs), with several hundreds of published scientific papers. Indeed, supercritical fluid extraction (SFE) has immediate advantages over traditional extraction techniques: it is a flexible process due to the possibility of continuous modulation of the solvent power/selectivity of the SCF and it allows the elimination of polluting organic solvents and of the expensive post-processing cost of solvent elimination from the extracts.

Because of this motivation, this work has as its objective the study of essential oil extraction from citronella species using supercritical carbon dioxide to produce an extract free of solvent and concentrated in the active components of the oil, with attention to the efficiency and the composition of the extracted oil. The extraction with supercritical fluid has potential as an alternative technology with the objective of minimizing energy and the use of organic and pollutant solvents. In this work, the supercritical solvent used is carbon dioxide because of its atoxicity, low cost, volatility and low critical properties.




The experiments were conducted using approximately, 6 g of dried leaves, 0.03 m in size, at ambient temperature,. The leaves were harvested in the Botanical Garden of the Universidade Federal Rural do Rio de Janeiro (UFRuralRJ). The citronella type used was citronella of Ceilão. The CO2 (minimum purity of 99.9 %) was from Linde Gases S.A. (Rio de Janeiro).

Experimental Procedure

The extraction was performed in an experimental apparatus containing a high pressure pump, a stainless steel extractor with a capacity of 42.0 cm3, a micrometric valve for sampling, a thermostatic bath (± 0.1 K, Lab-Line Instrument, Inc.; Melrose Park Illinois) to control the temperature, a manometer and a rotameter to measure the flow rate of CO2. The flow diagram of the experimental apparatus is presented in Figure 1 (from Mendes, 2002).



The experimental procedure starts with the adjustment of the temperature in the heating bath. The experimental procedure was done in the semi-batch mode and carbon dioxide was fed with a volumetric flow rate of 0.155 cm3 s-1 during the extraction time of 3600 seconds. When the pump reached the desired pressure and the system stabilized in a steady - state mode, the extraction was initiated with sampling, using a micrometric valve. The oil was collected in tubes of 15.0 cm3.

A salt and ice bath (temperature equal to 263.15 K) was placed on in the high pressure pump head with the objective of ensuring that the solvent was in the liquid state.

The extracted oil was weighed on analytical balance with a precision of 1 x 10-4 g.

The operational conditions investigated were 313.15, 313.15 and 353.15 K for the temperature and 6.2, 10.0, 15.0 and 18.0 MPa, for the pressure.

Chromatographic Analysis

The extracts obtained from the supercritical fluid extraction were analyzed using gas chromatography (HP - 6890 Series System) coupled to a mass spectrometer (5973 MSD) using the Wiley 275 library (Adams, 1993). A DB-5 capillary column DB-5 (30 m long, 0.25 mm internal diameter, 0.25μm film thickness) was utilized with temperature limits from 213.15 K to 598.15 K. The flow of the carrier gas was 1.0 ml/min and the temperature programming varied from 353.15 K to 573.15 K at 279.15 K/min, with the injector temperature of 563.15 K and detector temperature of 583.15 K.



A total of 12 experiments were performed, varying the conditions of temperature and pressure. The results are presented in Table 1.

The yield (Y) is presented as the ratio between the mass of extracted oil and the mass of leaves fed into the extractor. This formula was the same used by Reis et al. (2006). The best value of the yield was 2.2 % at 353.15K and 18.0 MPa of pressure. Although this is a low value, compared to 9.40 % obtained by Reis et al. (2006), it is important to remember that, in the process using supercritical carbon dioxide, the extract does not have residual solvent, which could justify its application to produce more pure oil. Moreover, the traditional processes using organic solvents extract all of the components without selectivity in relation to the major components. Reis et al. (2006) extracted the essential oil using a Clevenger apparatus (hydrodistillation with extraction with heptane and drying of the organic fraction with magnesium sulfate), using the same methodology described in Radünz et al. (2002).

Figure 2 shows that the quantity of extracted oil increases with increasing pressure, at constant temperature. This reflects the increase in the carbon dioxide density and, consequently, the increase of its solvent power.



As observed with other raw materials (Vargas et al., 2010), practically all the oil was extracted in the first 30 minutes of extraction. This indicates that the majority of the oil was probably present near the surface of the leaves.

The quantity of solvent consumed during the extraction can be calculated and verified as the ratio between the extracted oil mass and the consumed mass of carbon dioxide, ME/MCO2. Chrastill (1982) called this variable the operational solubility because of the similarity of this ratio to the concept of solubility, which is however a thermodynamic property. Table 2 shows how the ratio ME/MCO2 varies as a function of the density of the solvent. Table 2 also shows the extracted mass (ME) and the quantity of CO2 consumed. Figure 3 shows the behavior of the variables presented in Table 2.





In a general manner, Table 2 shows that the extracted mass increased with increasing CO2 density. Therefore, a cross-over effect can be observed in the extraction curves in Figure 3 at 313.15 and 333.15 K, probably due to the increase in the vapor pressure of the solutes, contrasting with the decrease in the CO2 density. Another possibility might be the extraction of other components together with the oil.

At 353.15 K the same behavior was not observed, which probably indicates that other components are being solubilized with increasing temperature. Thus, the extracted mass increases, but in part due to the solubilization of other chemical components that could potentially inhibit the essential oil extraction or change the original characteristics of the oil.

In general, at pressures higher than 10.0 MPa, other components present in the plant are extracted in addition to essential oil. This fact could be verified by the variation of the color of the extract obtained. At greater pressures, the extract is darker. At lower pressures, the color of the raffinate was yellow, while at higher pressures the color was green. When the pressure was increased, the intensity of the color and the amount of extracted material increased. The co-extraction of waxes was also observed by Carlson et al. (2001) in the extraction of lemongrass (Cymbopogon citratus) essential oil with dense carbon dioxide.

Figure 4 shows the chromatogram of the citronella oil obtained at 353.15 K and 18.0 MPa, indicating the relative abundance of the substances and their retention times. The identification of each component in the extract is indicated in Table 3, corresponding to the numbers of each peak in Figure 4. The Wiley library, used to identify the components, only confirms the presence of any component if the spectral match has a probability (P) higher than 70%.



To compare the selectivity of the process, Figure 5 shows the results of the chromatographic analysis of the extract obtained at 313.15K and 62.0MPa. Each identified component is listed in Table 4.



The results indicated a higher selectivity in the essential oil extracted at the higher temperature and pressure, as evidenced by the number of components present in the chromatograms. This behavior could be observed by comparing Figure 4 and Figure 5 and further certified in Tables 3 and 4.

The compositions of the citronella oil extracted under two different operational conditions, presented in Tables 3 and 4, indicate that these oils can be used as repellent with numerous applications because of the presence of citronellal, citronelol, geraniol and eugenol.



A simple model was applied to fit the extraction results of citronella oil. The model was proposed by Naik and Lentz (1988) and it does not consider interactions between the solute and the solid matrix. This model was also applied by Esquível et al. (1999), Silva et al. (2008), and to Jatropha curcas oil (Pereira, 2009). The model is represented by Equation (2) and has only two adjustable parameters, elim and b:

where e is the ratio of the mass of oil extracted at time t (s) to the initial mass of solute-free feed (kg of oil/kg of solute-free feed) and elim is the e value for an infinite extraction time or the maximum amount of solute initially present in the matrix. The ratio elim/b corresponds to the initial slope of the extraction curve versus time.

The model was used to represent the extraction behavior. The value of elim, taken from the conventional hidrodistillation method extraction, was 9.4% of oil in citronela leaves (Reis et al., 2006).

Table 5 shows the empirical model parameter, b, and the calculated efficiency (ecalc - %) for all the operational conditions. Although the b parameter does not have a physical meaning, it is possible to see its variation as a function of the pressure.

The behavior of the efficiency (%) as a function of the time, at 6.2, 10.0, 15.0 and 18.0MPa for all temperatures, confirms the fact that almost all the oil had been extracted by half the time of the process. The presence of the oil on the surface of the leaves favours the convection mechanism as opposed to diffusion. As suggested by Sovová (1994), the extraction process occurs in three stages and they are controlled by either a convective or diffusion mechanisms, with a transition between them.

The results show that 313.15K is the best temperature, with no great differences in the efficiency at 333.15 and 353.15K. The results at 10.0MPa are in accord with the behavior observed in the experimental results. At 15.0MPa, the best efficiency was found at 333.15K, as can also be seen at 18.0MPa. Moreover, there were no great differences between 333.15 and 353.15K in relation to efficiency.



This work had as its objective the use of supercritical carbon dioxide to extract the essential oil of citronella. Different operational conditions were investigated using 6.0 x 10-3 kg of the plant in the extractor. The best condition was found to be 353.15K and 18.0MPa with the maximum efficiency of the process in relation to the quantity of extracted mass. Besides the better value of efficiency, this operational condition showed good selectivity compared to other conditions. A high selectivity was obtained at 353.15K and 18.0MPa, with a more pure essential oil concentrated in the active main components. At 18.0MPa and 353.15K, seven components were presented, while at 62.0MPa and 313.15K there were seventeen components. The model used to predict the extraction behaviour showed a similar behaviour for all experimental conditions. This result suggests that the process is controled by the convection mechanism. The composition of the essential oil obtained using supercritical carbon dioxide indicates that the oil can be used for its antimicrobial, antifungal and repellency activities.



ME Extracted mass kg

Eff. Efficiency of extraction as a function of temperature %

T Temperature K

P Pressure MPa

D Density kg.m-3

MCO2 consumed mass of CO2 kg

e kg of oil/kg of solute-free feed (-)

b parameter of empirical model s

t time s

elim e value for infinite extraction time or the maximum amount of solute present kg/kg

ecalc calculated e value kg/kg



To Cláudia Rezende and Vivian Ferreira from IQ/UFRJ, for help with the chromatographic analysis.



Adams, R. P., Identification of essential oil composition by gas chromatography/mass spectroscopy. Allured Publishing Corporation, Illinois-USA (1993).         [ Links ]

Carlson, L. H. C., Machado, R. A. F., Spricigo, C. B., Pereira, L. K., Bolzan, A., Extraction of lemongrass essential oil with dense carbon dioxide. Journal of Supercritical Fluids, 21, p. 33-39 (2001).         [ Links ]

Choochote, W., Chaithong, U., Kamsuk, K., Jitpakdi, A., Tippawangkosol, P., Tuetun, B., Champakaew, D., Pitasawat, B., Repellent activity of selected essential oils against Aedes aegypti. Fitoterapia, 78, p. 359-364 (2007).         [ Links ]

Esquível, M. M., Bernardo-Gil, M. G. and King, M. B., Mathematical models for supercritical extraction of olive husk oil. The Journal of Supercritical Fluids, 16, p. 43 (1999).         [ Links ]

Fradin, M. S., Mosquitoes and mosquito repellents: a clinician's guide. Ann. Int. Med. 128, p. 931-940 (1998).         [ Links ]

Katz, T. M., Miller, J. H., Hebert, A. A., Insect repellents: Historical perspectives and new developments. J. Am. Acad. Dermatol., v. 58, 5, p. 865-871 (2008).         [ Links ]

Magalhães, M. T., Composição Química de Óleos Essenciais. I Simpósio de Óleos Essenciais, São Paulo (1985).         [ Links ]

Mendes, M. F., Estudo do Processo de Concentração da Vitamina E Presente no Destilado da Desodorização do Óleo de soja usando CO2 Supercrítico. D. Sc. Thesis, PEQ/COPPE/UFRJ, Rio de Janeiro (2002).         [ Links ]

Mendes, M. F., Pessoa, F. L. P., Melo, S. B. V. de, Queiroz, E. M., Extraction modes IN Handbook of Food Products Manufacturing. Wiley, p. 147-156 (2007).         [ Links ]

Pereira, C. S. S, Avaliação de diferentes tecnologias na extração do óleo do pinhão-manso (Jatropha curcas L.). Tese de M. Sc., PPGEQ/UFRRJ, Seropédica, Rio de Janeiro, Brasil (2009).         [ Links ]

Nakahara, K., Alzoreky, N. S., Yoshihashi, T., Nguyen, H. T. T., Trakoontivakorn, G., Chemical composition and antifungal activity of essential oil from Cymbopogon nardus (citronella grass). Japan Agricultural Research Quaterly, v. 37, 4, p. 249-252 (2003).         [ Links ]

Naik, S. N., Lentz, H., Extraction of perfumes and flavours from plant materials with liquid carbon dioxide under liquid-vapour equilibrium conditions. Fluid Phase Equilibria, v. 49, p. 115- 126, 1988.         [ Links ]

Olivo, C. J., Carvalho, N. M. de, Silva, J. H. S. da, Vogel, F. F., Massariol, P., Meinerz, G., Agnolin, C., Morel, A. F., Viau, L. V., Óleo de citronella no controle do carrapato de bovinos. Revista Ciência Rural, v. 38, 2, p. 406-410 (2008).         [ Links ]

Radünz, L. L., Melo, E. C., Berbert, P. A., Barbosa, L. C. A., Rocha, R. P., Grandi, A. M., Efeitos da temperatura do ar de secagem sobre a qualidade do óleo essencial de alecrim pimenta (Lippia sidoides cham), Revista Brasileira de Armazenamento. 27, 2, p. 9-13 (2002).         [ Links ]

Reis, G. G., Peisino, A. L., Alberto, D. L., Mendes, M. F., Calçada, L. A., Estudo do efeito da secagem em convecção natural e forçada na composição do óleo essencial da citronela (Cymbopogon nardus). Revista Brasileira de Plantas Medicinais, v. 8, 4, p. 47-55 (2006).         [ Links ]

Reverchon, E., De Marco, I., Supercritical fluid extraction and fractionation of natural matter. Journal of Supercritical Fluids, 38, p. 146-166 (2006).         [ Links ]

Sakulku, U., Nuchuchua, O., Uawongyart, N., Puttipipatkhachorn, S., Soottitantawat, A., Ruktanonchai, U., Characterization and mosquito repellent activity of citronella oil nanoemulsion. International Journal of Pharmaceutics, 372, 1-2, p. 105-111 (2009).         [ Links ]

Silva, C. F., Mendes, M. F., Pessoa, F. L. P., Queiroz, E. M., Supercritical Carbon Dioxide Extraction of Macadamia (Macadamia integrifólia) Nut Oil: Experiments and Modeling. Brazilian Journal of Chemical Engineering, v. 25, 1, p. 175-181 (2008).         [ Links ]

Simic, A., Rancic, A., Sokovic, M. D., Ristic, M., Grujic-Jovanovic, S., Vukojevic, J., Marin, P. D., Essential oil composition of Cymbopogon winterianus and Carum carvi and their antimicrobial activities. Pharmaceutical Biology, v. 46, 6, p. 437-441 (2008).         [ Links ]

Vargas, C. E., Mendes, M. F., Azevedo, D. A., Pessoa, F. L. P., Uller, A. C., Extraction of the essential oil of abajeru (Chrysobalamus icaco) using supercritical CO2. Journal of Supercritical Fluids, v. 52, 2, p. 171-177 (2010).         [ Links ]

Trongtokit, Y., Rongsriyam, Y., Komalamisra, N., Apiwathnasorn, C., Comparative repellency of 38 essential oils against mosquito bites. Phytotherapy Research, v. 19, 4, p. 303-309 (2005).         [ Links ]

Wong, K. K. Y., Signal, F. A., Campion, S. H., Motion, R. L., Citronella as an insect repellent in food packaging. Journal of Agricultural and Food Chemistry, v. 53, 11, p. 4633-4636 (2005).         [ Links ]



(Submitted: August 20, 2009 ; Revised: February 16, 2011 ; Accepted: February 26, 2011)



* To whom correspondence should be addressed

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