SciELO - Scientific Electronic Library Online

vol.17 issue2Modeling and simulation of hydrodemetallation and hydrodesulfurization processes with transient catalytic efficiencyEquilibrium data on ethanol-water-solvent ternaries 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.2 São Paulo June 2000 

High-pressure vapor-liquid equilibrium data for CO2-orange peel oil


G.R. Stuart, C. Dariva and J. Vladimir Oliveira*
Programa de Engenharia Química/COPPE/UFRJ, Rio de Janeiro, RJ - Brazil,
21945-970, Fax: +55-21-5907135


(Received: August 26, 1999 ; Accepted: March 9, 2000)



Abstract - Recently, there has been a growing interest in fractionating orange peel oil by the use of supercritical carbon dioxide (SCCO2). However, progress in this area has been hindered by the lack of more comprehensive work concerning the phase equilibrium behavior of the SCCO2-orange peel oil system. In this context, the aim of this work is to provide new phase equilibrium data for this system over a wide range of temperatures and pressures, permitting the construction of coexistence PT-xy curves as well as the P-T diagram. The experiments were performed in a high-pressure variable-volume view cell in the temperature range of 50-70oC from 70 to 135 atm and in the CO2 mass fraction composition range of 0.35-0.98. Based on the experimental phase equilibrium results, appropriate operating conditions can be set for high-pressure fractionation purposes.
Keywords: phase equilibrium data, supercritical carbon dioxide, orange peel oil.




According to Nonino (1997) world production of citrus has grown approximately 70% while the processing of citrus has increased around 100%, pointing to the importance of using the by-products, especially the cold-pressed oil, as raw materials. Cold-pressed oils, also called citrus oils, consist of mixtures of hydrocarbons of the terpene and sesquiterpene groups, oxygenated compounds and nonvolatile residues. Terpenes do not contribute much to the flavor or fragrance of the oil; since they are mostly unsaturated compounds, they are unstable to heat and light, and rapidly oxidize in air. It is the oxygenated compounds that provide much of the characteristic flavor of citrus oil.

Over the past few years, the possibility of fractionating citrus oils has received widespread attention due to the direct applications in the food and pharmaceutical industries of the high-value-added products generated. Though fractionation of such oils has been commonly performed using the so-called conventional techniques, such as vacuum distillation, efforts in this area are heading toward the use of SCCO2 with different sorts of equipment operating in different modes (Sato et al., 1995; Sato et al., 1996; Budich et al., 1999). Also, a variety of simulations have been conducted in an attempt to establish the best operating conditions and to provide a first approach to scale-up in such processes (Vieira de Melo, 1997; Mizutani, 1999).

Nevertheless, this step has generally been accomplished by using thermodynamic properties and phase equilibrium data involving only the most representative compounds present in the citrus oil. Perhaps, the best example is orange peel oil, where limonene and linalool are commonly considered the key components in this complex mixture, allowing the calculation of selectivity and capacity from binary or ternary phase equilibrium data containing SCCO2. A recent and very good compilation on this subject is provided by Vieira de Melo et al. (1999). It is well known that the quality of the oil is primarily determined by the content of the oxygenated compounds, such as linalool, decanal, etc. However, considering the fact that the concentrations of some of these compounds in the raw oil are quite similar and their vapor pressures are very close, the choice of linalool as a cut component is strictly empirical, and therefore a realistic study of the fractionating process should be based on phase equilibrium of the whole oil rather than its key components.

Unfortunately, only a small amount of experimental data concerning SCCO2-orange cold-pressed peel oil has been published (Budich et al., 1999 - given in graph form only). Thus, new high-pressure vapor-liquid equilibrium data for this system are reported in this work permitting the construction of PT-xy and P-T diagrams the temperature and pressure ranges for practical interest.

The experimental technique used to study phase equilibrium can be classified as the synthetic or indirect method, in which equilibrium compositions might be determined without sampling (Stuart, 1999). As mentioned by Fornari et al. (1990), the advantage of an equilibrium cell with visual capabilities is that phase separation can be observed directly, thus eliminating the possibility of missing an additional equilibrium phase. Note also that the use of a variable-volume cell is especially useful near critical conditions where pressure drops, and hence large disturbances can occur when withdrawing samples from the cell. Though our equilibrium cell has two sampling exits, at this stage we are more concerned with the construction of the vapor-liquid phase boundary in order to set the appropriate operating conditions to avoid one-phase regions in our high-pressure extraction column. Excellent reviews of experimental techniques for studying high-pressure phase equilibrium can be found in the work of Fornari et al. (1990), Dohrn and Brunner (1995), Vieira de Melo (1997) and Cassel (1998).



Materials and Methods

Cold-pressed orange peel oil was kindly supplied by Citrosuco S.A. and used without further treatment. These oils were expressed by FMC in-line juice extractors during orange juice extraction. A crude oil emulsion was obtained from the juice extractor, and subsequently this oil emulsion was centrifuged to separate the oil from the water. Carbon dioxide, of a grade higher than 99.9% was purchased from AGA S.A.

Gas chromatography analyses were performed by injecting 0.04 ml samples with a 100:1 split into a Hewlett-Packard 5890 Series II gas chromatograph equipped with a flame ionization detector and connected to a mass spectrometer Hewlett-Packard 5970. A fused silica capillary column, 25m length 0.2mm i.d., cross-linked with polyethylene glycol-TPA modified (FFAP), was used. Column temperature was programmed from 60 to 200oC at 5oC/min. Hydrogen flow rate was 1ml/min through the column. Injection port temperature was 250oC and detector temperature 280oC. Percentages were calculated by the internal normalization method.

Apparatus and Procedure

The phase equilibrium experiments (cloud points) were performed in a high-pressure variable-volume view cell. A schematic diagram of the apparatus is presented in Figure 1. Basically, it consists of a view cell with three sapphire windows for visual observations, an absolute pressure transducer (Smar LD 301) with a precision of ± 0.118 atm, a portable programmer (Smar, HT 201) for pressure data acquisition and a syringe pump (ISCO 260D). The equilibrium cell has a maximum internal volume of 28cm3 and contains a movable piston that permits pressure control inside the cell. Phase transitions were recorded visually as bubble or dew points by varying the pressure behind the piston with the syringe pump and CO2 used as the pressurizing fluid. The cell was equipped with an electrical heater and a PID temperature controller (Dextron, DTS4) connected to a thermocouple in direct contact with the fluid mixture inside the cell body. This arrangement provided temperature control with a precision of 0.5oC.


a6f1.gif (4424 bytes)


Depending on the desired global composition, an amount of oil from 0.1 to 7g was weighed on a sensitive scale (Ohaus Analytical Standard with an accuracy of 0.0001g) and loaded into the cell. Then the cell and all lines were flushed with low-pressure carbon dioxide to remove any air. Afterwards, carbon dioxide was pumped into the cell to obtain a pre-established global composition. The amount of CO2 charged was monitored by the change in the total mass of the transfer vessel of the pump, maintained at 24oC and 78.46 atm. The mass of CO2 transferred was then computed, based on recent experimental PVT data for this component (Gokmenoglu et al., 1996). Precision was estimated to be within 0.005 wt%. During the charging process, no pressure was applied to the piston to ensure that experiments were started when the cell was at its maximum volume. Then the cell content was continuously stirred by means of a magnetic stirrer and a Teflon-coated stirring bar.

After the desired temperature had been reached, the pressure was increased by applying pressure to the back of the piston with the syringe pump until the presence of one phase was observed; subsequently it was decreased until the incipient formation of a two-phase system was observed. The equilibrium pressure was then recorded after repeating this procedure at least five times, leading to an average reproducibility of 0.68 atm. When tests were completed at a given temperature, the cell was heated to a new temperature and the procedure was repeated.



Since no experimental vapor-liquid equilibrium data for the system under study in this work are available for a direct comparison, two tests were performed to check the reliability of our experimental measurements. First, the vapor pressure of pure carbon dioxide was determined at 26 and 27oC, giving 66.00 and 69.17 atm, respectively, which are in good agreement with the values reported by Angus et al. (1976), 66.28 and 70.60 atm. The second and more important test consisted in obtaining bubble points for the CO2-ethanol system at 40oC. Figure 2 depicts a comparison between our data at four global compositions and those published in the literature (Chang et al., 1997; Suzuki et al., 1990; Day et al., 1996). Though there is some divergence in the experimental data, which is very common when dealing with high-pressure phase equilibrium, our experimental procedure seems to be reliable. In both cases, pure CO2 and CO2-ethanol, mean values were obtained after triplicate runs with a standard deviation of around 0.5 atm.


a6f2.gif (4387 bytes)


Afterwards, phase equilibrium data for CO2-orange peel oil were measured at temperatures of 50, 55, 60, 65 and 70oC and at global CO2 compositions (wt%) of 0.35, 0.50, 0.65, 0.80, 0.95 and 0.98. Table 1 presents all the experimental data where good reproducibility of the pressure values can be observed. In the last column of this table, the transition types observed at each temperature, characterized by either a dew point (DP) or a bubble point (BP), are shown. In our experiments, DP transitions were characterized by the formation of drops of the heavier phase at the bottom of the cell, preceded by a complete opalescence of the solution, while in BP transitions, small bubbles arise from the middle of the cell with a subsequent formation of a lighter phase at the top of the cell.


a6t1.gif (9039 bytes)


It is also worth observing in this table and the PT diagram in Figure 3 that at CO2 concentrations of 0.95 and 0.98 the phase transition changes from a BP to a DP as temperature increases. Thus the critical points for these compositions are believed to be very close to 47oC and 88 atm and 43oC and 81.5 atm, respectively. These results seem to be coherent since, as the oil has a much lower vapor pressure than carbon dioxide, it is expected that the binary critical point temperature increase with an enhancement of the CO2 content. Moreover, it should be noted that the liquid phase isoplets do not intercept themselves but they may intercept more than one vapor-phase isoplet curve, suggesting that our results are thermodynamically coherent. We have also experimentally observed the intriguing phenomenon of first-kind retrograde condensation when working with the highest CO2 compositions, 0.95 and 0.98. Decompression at constant temperature beginning at the DP led to the appearance of a liquid phase, and with further decompression the amount of liquid formed was gradually diminished.


a6f3.gif (4842 bytes)


In Figure 4 the PT-xy diagram for CO2-orange peel oil is presented. Our experimental data are in good qualitative agreement with those published by Budich et al. (1999). It is interesting to note from this figure that near the critical point, slight changes in pressure lead to large variations in liquid phase compositions at lower temperatures, whereas as temperature is raised the plateau becomes less pronounced. This might be a very important point since more stable operating regions may be reached at higher temperatures if one carries out a SCCO2 fractionation step.


a6f4.gif (3982 bytes)


The experimental phase equilibrium data were modeled using the Peng-Robinson equation of state (PR-EOS, Peng and Robinson, 1976) with the quadratic van der Waals mixing rule together with the three-parameter expression for the temperature dependence of the attractive term proposed by Aznar and Silva Telles (1997). The cross-interaction parameters presented in Table 2 (kij for the attractive term and lij for the repulsive one) were estimated by the Maximum Likelihood method, weighing the experimental values according to their precision. The objective function was minimized using the ESTIMA software (Noronha et al., 1987).


a6t2.gif (3210 bytes)


Since it is not our main objective to provide a rigorous thermodynamic treatment of the phase equilibria of such a system by considering all the binary interaction parameters involved in the multicomponent calculations, the oil was considered to be a pure pseudo component. The pertinent thermodynamic properties of the oil were evaluated, considering the following main compounds (in mass percent, wt/wt) as obtained by gas chromatography analysis: a -pinene (0.53), sabinene (0.21), myrcene (1.87), limonene (96.42), b -phellandrene (0.27), decanal (0.26) and linalool (0.45). The critical properties were estimated using the group contribution method suggested by Somayajulu (1989), resulting in the following values for the whole oil (molecular weight: 136.35g.gmol-1): critical pressure (27.49 atm); critical temperature (376.88oC) and accentric factor (0.3918). The Aznar and Silva Telles (1997) parameters for the whole oil were obtained by weighing the parameters of the main compounds present in the oil, which were evaluated using the Willmann and Teja (1985) procedure, resulting in the following values: m= 0.492, n=0.186 and G =0.685. For carbon dioxide, the values reported by Aznar and Silva Telles (1997) were employed.

In order to check the performance of the PR-EOS and to provide a better visualization of the experimental data for the CO2-orange peel oil, Figure 5 presents a comparison between experimental and calculated values, obtained by flash calculations at three temperatures. One can see from this figure that the thermodynamic model is capable of a good estimation of both the liquid and vapor phase boundaries, except near the critical points.


a6f5.gif (6739 bytes)




In this work we have reported new phase equilibrium data for the CO2-orange peel oil system obtained in a high-pressure variable-volume view cell over a temperature range of practical interest. The experimental procedure presented here is shown to be reliable, as one can consider a comparison between our vapor-liquid equilibrium data for CO2-ethanol and those taken from the literature. Moreover our results for the CO2-orange peel oil system are in good qualitative agreement with those reported by Budich et al. (1999). The experimental apparatus built in this work is of great value since it allows the investigation of different phase transitions as well as the identification of important phenomena such as retrograde condensation. With this equipment, liquid to liquid-liquid and then to vapor-liquid-liquid, phase transition experiments for polymer/solvent systems are currently under way in our laboratory.



The authors would like to thank FINEP and CNPq for their financial support. The authors gratefully acknowledge CITROSUCO S.A. for providing the orange peel oil. The authors are indebted to Dr. Sandro R.P. Rocha (University of Austin-Texas) for his helpful comments and suggestions during this work.


Angus, S., Armstrong, B. and de Reuck, K.M., Eds. International Thermodynamic Tables of the Fluid State. Carbon Dioxide. Pergamon Press, New York (1976).        [ Links ]

Aznar, M. and Silva Telles, A., A Data Bank of Parameters for the Attractive Coefficient of the Peng-Robinson Equation of State, Brazilian Journal of Chemical Engineering, 14, 19 (1997).        [ Links ]

Budich, M., Heilig, S., Wesse, T., Leifbkuchler, V. and Brunner, G., Countercurrent deterpenation of citrus oils with supercritical CO2, J. Supercrit. Fluids, 14, 105 (1999).        [ Links ]

Cassel, E., Equilíbrio de fases a altas pressões de sistemas contendo hidrocarbonetos e dióxido de carbono, D.Sc. Thesis, PEQ/COPPE/UFRJ (1998).        [ Links ]

Chang, J.C., Day, C.Y., Ko, C.M. and Chiu, K.L., Densities and P-x-y diagrams for carbon dioxide dissolution in methanol, ethanol, and acetone mixtures, Fluid Phase Equilibria, 131, 243 (1997).        [ Links ]

Day, C.Y., Chang, J.C. and Chen, C.Y., Phase Equilibrium of Ethanol + CO2 and Acetone + CO2 at Elevated Pressures, J. Chem. Eng. Data, 41, 839 (1996).        [ Links ]

Dohrn, R. and Brunner, G., High Pressure Fluid Phase Equilibria: Experimental Methods and Systems Investigated (1988-1993), Fluid Phase Equilibria, 106, 213 (1995).        [ Links ]

Fornari, R.E., Alessi, P. and Kikic, I., High Pressure Fluid Phase Equilibria: Experimental Methods and Systems Investigated (1978-1987), Fluid Phase Equilibria, 57, 1 (1990).        [ Links ]

Gokmenoglu, Z., Xiong, Y. and Kiran, E., Volumetric Properties of Carbon Dioxide + Sulfur Hexafluoride at High Pressures, J. Chem. Eng. Data, 41, 354 (1996).        [ Links ]

Mizutani, F.T., Dimensionamento e Análise de Processo de Extração Usando Fluidos Supercríticos, M.Sc. Thesis, PEQ/COPPE/UFRJ, Brazil (1999).        [ Links ]

Nonino, E., Where is the Citrus Industry Going? Perfumer and Flavorist, 22, 53 (1997).        [ Links ]

Noronha, F.B., Monteiro, J.L., Lobão, M.W., Santos, T.J. and Pinto, J.C.S., ESTIMA: Um Pacote Computacional para Estimação de Parâmetros e Projeto de Experimentos, PEQ/COPPE/UFRJ (1987).        [ Links ]

Peng, D.Y. and Robinson, D.B., A new two-constant equation of state, Ind. Eng. Chem. Fundam., 15, 59 (1976).        [ Links ]

Sato, M., Goto, M. and Hirose, T., Fractional Extraction with Supercritical Carbon Dioxide for the Removal of Terpenes from Citrus Oil, Ind. Eng. Chem. Res., 34, 3941 (1995).        [ Links ]

Sato, M., Goto, M. and Hirose, T., Supercritical Fluid Extraction in Semibatch Mode for the Removal of Terpenes in Citrus Oil, Ind. Eng. Chem. Res., 35, 1906 (1996).        [ Links ]

Somayajulu, G.R., Estimation Procedures for Critical Constants, J. Chem. Eng. Data, 34, 106 (1989).        [ Links ]

Stuart, G.R., Estudo do Fracionamento do Óleo de Laranja por Destilação a Vácuo e Dióxido de Carbono Supercrítico, D.Sc. Thesis, PEQ/COPPE/UFRJ, Brazil (1999).        [ Links ]

Suzuki, K., Sue, H., Itou, M., Smith, L.R., Inomata, H., Arai, K. and Saito, S., Isothermal Vapor-Liquid Equilibrium Data for Binary Systems at High Pressures: Carbon Dioxide-Methanol, Carbon Dioxide-Ethanol, Carbon Dioxide-1-Propanol, Methane-Ethanol, Methane-1-Propanol, Ethane-Ethanol, and Ethane-1-Propanol Systems, J. Chem. Eng. Data, 35, 63 (1990).        [ Links ]

Vieira de Melo, S.B., Desterpenação do Óleo Essencial de Laranja Usando CO2 Supercrítico, D.Sc. Thesis, PEQ/COPPE/UFRJ, Brazil (1997).        [ Links ]

Vieira de Melo, S.B., Pallado, P., Guarise, G.B. and Bertucco, A., High-Pressure Vapor-Liquid Equilibrium Data for Binary and Ternary Systems Formed by Supercritical CO2, Limonene and Linalool, Brazilian Journal of Chemical Engineering, 16, 7 (1999).        [ Links ]

Willmann, B. and Teja, A.S., Method for the Prediction of Pure Component Vapor Pressure in the Range of 1KPa to the Critical Pressure, Ind. Eng. Chem. Proc. Dev., 24, 1033 (1985).        [ Links ]



* 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