High-pressure Vapor-liquid Equilibrium Data for Ternary Systems Co 2 + Organic Solvent + Curcumin

In the present work phase equilibrium data for the ternary systems CO 2 + ethanol + curcumin and CO 2 + ethyl acetate + curcumin, at different concentrations of curcumin in the organic solvent (0.01 gcm-3 , 0.005 gcm-3 for ethanol and 0.01 gcm-3 , 0.0025 gcm-3 for ethyl acetate) are reported. The static synthetic method, using a variable-volume view cell was employed for obtaining the experimental data in the temperature range of (303 to 333) K and pressures up to 11 MPa. Vapor-liquid phase transitions were observed as bubble and dew points for the overall compositions investigated. The phase equilibria of the ternary systems were fitted to the Peng-Robinson equation of state (PR-EoS) using only binary system information. Experimental data together with modeling results may constitute relevant information for the precipitation of curcumin using the supercritical antisolvent process.


INTRODUCTION
Curcuma longa L. is a perennial herbaceous plant with medicinal properties that belongs to the family of Zingiberaceae.Such plant is cultivated in Indian, China and other countries with tropical weather, is considered one of the main pigments produced in Brazil.The rhizome is the part of the plant used in medical applications and it has a yellow color and may be used as a food coloring.Curcumin ((E,E)-1,7-bis(4-Hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5dione) is the major non-volatile active compound of the rhizome.This curcuminoid has anti-oxidative, anti-carcinogenic, anti-mutagenic, anti-inflammatory and anti-fungicidal effects.The application of supercritical fluids as antisolvents for precipitation of curcumin can be an attractive technique due to the low solubility of this substance in supercritical fluid in comparison with conventional methods that use organic solvents.In order to develop or to optimize these processes for producing micronized particles, the knowledge of the phase behavior of solute + solvent + antisolvent is a crucial aspect.
Since ancient times, curcumin present in turmeric is known by Ayurvedic and Chinese medicines.New researches showed that curcumin has therapeutic effects on various diseases like Alzheimer, AIDS and cancer.The numerous therapeutic activities of curcumin can be found in the works of professor Aggarwal and collaborators [1].According to Baskaya [2] curcumin (1,7-bis(4-hydroxy-3-methoxy-*Address correspondence to this author at the DEQ/UEM, Bloco D-90, Avenida Colombo, 5790, PR 87020-900, Maringá, PR, Brazil; Tel: +55 44 32614749; E-mail: cardozo@deq.uem.brfenil)-1,6-heptadiene-3,5-dione) is a highly polar substance, insoluble in water or neutral pH, poorly soluble in hydrocarbon solvents, and soluble in alcohols.However, clinical application of curcumin is limited due to poor aqueous solubility, and consequently, minimal systemic bioavailability.In vitro absorption studies showed that nanoparticles of curcumin have the potential to bypass poor solubility and poor systemic bioavailability [3].A method that is an alternative to conventional precipitation techniques such as organic solvents or spray drying is supercritical fluids technology [4].
The application of supercritical CO 2 as antisolvents for curcumin precipitation can be adequate because it is a strongly polar substance, with a high molecular mass, and two phenyl and two carbonyl groups.The knowledge of phase behavior of the solute + solvent + antisolvent system plays an important role to determine appropriate operating conditions to conduct precipitation and understanding the kinetic nucleation mechanism and growth of the particles.Due to the potential applications of curcumin in the chemical, pharmaceutical and cosmetic industries there are a lot of work available in literature.However, the majority of the studies that use supercritical technology are related to supercritical fluid extraction [5][6][7] and few works provide phase equilibrium data [2,8].
In this context, the aims of the present work were: (1) to measure phase equilibrium (transition points) data for the ternary systems CO 2 (1) + ethanol (2) + curcumin (3) and CO 2 (1) + ethyl acetate (2) + curcumin (3), at different concentrations of curcumin in the organic solvent and at different temperatures and (2) to fit the experimental data to the Peng-Robinson equation of state (PR-EoS) using quadratic mixing rule.

EXPERIMENTAL Material
Ethyl acetate (99.5%) was purchased from VETEC (Rio de Janeiro, Brazil), and ethanol (99.5%) was obtained from Merck (Darmstadt, Germany).Curcumin with a minimum purity of 95% was purchased from Sigma-Aldrich (USA), and CO 2 (99.9% pure) was supplied by White Martins S.A.All materials were used as received, without any further purification.The organic stock solutions were previously prepared using two concentrations of curcumin in ethyl acetate (0.0025 g cm -3 and 0.010 g cm -3 ) and in ethanol (0.005 g cm -3 and 0.010 g cm -3 ).

Apparatus and Experimental Procedure
Phase equilibrium experiments were conducted employing the static synthetic method in a high-pressure variablevolume view cell.The experimental apparatus and procedure have been used in a variety of studies [9][10][11][12][13][14].Briefly, the experimental apparatus consists of a variable-volume view cell, with a maximum internal volume of 27 cm 3 , with two sapphire windows for visual observation, an absolute pressure transducer (Smar LD 301, Sertãozinho, Brazil), with a precision of 0.03 MPa, a portable programmer (Smar, HT 201, Sertãozinho, Brazil) for the pressure data acquisition, and a syringe pump (ISCO 260D, Lincoln, USA).The equilibrium cell contains a movable piston, which permits pressure control inside the cell.Phase transitions were recorded visually through manipulation of the pressure by the syringe pump using the solvent as a pneumatic fluid.Initially, a known mass of the solute (either the organic solvent or a mixture of curcumin + organic solvent) was weighed on a precision scale balance (Marte AM-220, ± 0.0001 g, Santa Rita do Sapucai, Brazil) and loaded into the equilibrium cell.The cell was then flushed with low-pressure CO 2 to remove any residual air.The charge of a known volume of CO 2 was performed with the help of the syringe pump (resulting in accuracy of ± 0.005 g in CO 2 loadings) until a desired global composition was achieved.On the basis of the uncertainty in CO 2 loading and other compound weight, we estimate that the uncertainty in global mass fraction of the mixtures was lower than 0.005 %.The cell content was kept under continuous agitation with the help of a magnetic stirrer and a Teflon-coated stirring bar.The temperature control was then turned on, and once the desired temperature was reached and controlled within ± 0.5 K, the pressure system was increased until the visualization of a one-phase system in the cell.At this point, the system was kept for at least 30 min to allow stabilization, and then the pressure was slowly decreased, typically at a rate of (0.1 to 0.5) MPa min -1 , until incipient formation of a new phase.In the bubble point transition, a small vapor bubble appears, while in the dew point transition, a small amount of dew, or fog, is formed in the cell.This procedure was repeated at least three times for each temperature and global composition.After completion of the measurement at a given temperature, the cell temperature was stabilized at a new value and the experimental procedure was repeated.

Thermodynamic Modeling
The experimental data of liquid-vapor transition were modeled by the Peng-Robinson equation of state with van der Waals quadratic mixing rule (vdW2).The isofugacity approach was used in the vapor-liquid equilibrium (VLE) calculations using the EoS-PR.The EoS-PR is given by the following equation: where P is the absolute pressure of the system, T is the absolute temperature, and v is the molar volume.The twoparameter of vdW2 mixing rules are given by: where k i, j and l i, j are the binary interaction parameters of the components in the mixture.The pure component parameters a i and b i are given by the equations: where Tc i , Pc i , i are the critical temperature, critical pressure, and the acentric factor of the i-component, respectively.Table 1 presents the critical properties, acentric factor of pure compounds used in this work and the melting temperature (T fus ) and enthalpy of fusion ( h fus ) of curcumin [2,[14][15][16].
In this approach only binary information (binary interactions parameters, k ij and l ij ) were used to model the VLE of the ternary systems.The binary interaction parameters for system CO 2 + ethanol (k 12 = 0.0703 and l 12 = -0.0262)and CO 2 + ethyl acetate (k 12 = -0.0373and l 12 = -0.0639)were taken from literature [14].The binary interaction parameters for CO 2 + curcumin, ethanol + curcumin and ethyl acetate + curcumin were set to zero.

RESULTS AND DISCUSSIONS
In order to measure the phase equilibrium data two stock solutions with different concentrations of curcumin in ethanol and ethyl acetate were prepared.For the ternary system CO 2 (1) + ethanol (2) + curcumin (3) the concentrations of curcumin in ethanol were used (0.010 g cm -3 and 0.005 g cm -3 ), whereas for the ternary system CO 2 (1) + ethyl acetate (2) + curcumin (3) the concentrations of curcumin in ethyl acetate were of 0.010 g cm -3 and 0.0025 g cm -3 ; the solution concentration uncertainty was lower than 0.0005 g cm -3 .
Tables 2 and 3 present the results for the ternary system CO 2 (1) + ethanol (2) + curcumin (3) at the two investigated concentrations of curcumin.In Table 2 the experimental data refer to vapor-liquid coexistence curve at curcumin concentration of 0.010 g cm -3 , where one can notice the occurrence of biphasic vapor-liquid transition (bubble or dew point transitions type) in the presence of solid phase was observed at all CO 2 mole fraction investigated in this work.It must be emphasized that the solid was precipitated from the CO 2 fed into the cell.In Table 3, which refers to curcumin concentration of 0.005 g cm -3 , the vapor-liquid transitions with precipitated curcumin were observed from CO 2 mole fraction x 1 = 0.6110 up to x 1 = 0.9522.Tables 4 and 5 present the results for the ternary system CO 2 (1) + ethyl acetate (2) + curcumin (3) at the two investigated concentrations of curcumin in organic solutions, 0.010 g cm -3 and 0.0025 g cm -3 , respectively.Vapor-liquid transitions were observed at all composition range investigated in this work at concentration of curcumin (in the original stock solutions) at 0.010 g cm -3 , as indicated in Table 4 by symbol BP(S) or DP(S) (bubble point or dew point in the presence of precipitated curcumin).In Table 5, at curcumin concentration 0.0025 g cm -3 , it can be noted that vapor-liquid transitions in the presence of solid phase were observed for x 1 0.6839.Below this mixture composition vapor-liquid (BP) transitions were observed.Fig.
(1) provides a comparison in the pressurecomposition diagram between experimental data obtained in this work for the ternary system CO 2 + ethanol + curcumin for the two curcumin concentrations (0.010 g cm -3 and 0.005 g cm -3 ) and binary data from literature at 313 K [17][18][19]; Fig.
(2) presents the same comparisons for the temperatures of 303 K and 333 K.

Figs. (1 and 2)
show that the presence of did not lead to significant changes in the pressure transitions when compared to the corresponding binary systems; the presence of solute did not modify the shape of vapor-liquid region.The thermodynamic model fitted well the experimental data.This work at 0.005 g.mL -1 at 0.010 g.mL -1 Literature binary CO 2 +ethanol Joung et al. [17] Chang et al. [18] Chiu et al. [19] In Figs.
(3(a) and 3(b)) are indicated regions where the vapor-liquid transitions were observed in the presence of solid (curcumin precipitated) (as also indicated in Table 2 and Table 3).Vertical dashed lines are the qualitatively indication of these distinct regions.Taking into account this experimental evidence, it can be seen the anti-solvent effect of CO 2 in the organic phase (solvent + curcumin).The CO 2 amount necessary to precipitate the curcumin from organic solution is higher according to decreasing curcumin concentration in ethanol.As it can be observed from Tables 2 and 3 this anti-solvency effect of CO 2 was observed at all temperature range investigated in this work (303 to 333 K).Fig. (4) provides a comparison between experimental data obtained in this work with experimental data from the literature for the binary system of CO 2 (1) + ethyl acetate (2) at 303 and 333 K [14].
Data on the curcumin saturation pressure are scarce in literature however its saturation pressure is approximately zero.Therefore, systems with higher-pressure transition are generally those with lower concentrations of curcumin.
In Figs.(5(a) and 5(b)) are indicated the regions where the vapor-liquid transition were observed in the presence of solid (curcumin precipitated) (see also Table 4 and Table 5).For curcumin concentration of 0.010 g cm -3 (Fig. 5(a)) vapor-liquid transitions occurred in the presence of solid phase (curcumin).When the concentration of curcumin in ethyl acetate was 0.0025 g cm -3 the solid phase occurrence was observed for x 1 0.6839 as indicated by the vertical dashed line in Fig. 5(b).This behavior is similar to that observed for the system with ethanol (Figs.3(a) and 3(b)), i.e., smaller CO 2 amounts are necessary to precipitate curcumin from organic phase as the curcumin concentration in organic solvent (ethanol or ethyl acetate) increased.The anti-solvency effect of CO 2 was observed for all temperatures in the range investigated in this work (303 to 333 K) (Tables 4 and 5).

CONCLUSIONS
Phase transition measurements for the ternary systems CO 2 (1) + ethanol (2) + curcumin (3) and CO 2 (1) + ethyl acetate (2) + curcumin (3) were reported in this work and modeled with the PR-EoS with quadratic mixing rules.It was experimentally observed that the addition of curcumin to the binary systems of CO 2 /organic solvent did not lead to changes in the pressure transitions.The approach employed to predict phase equilibrium data of the ternary systems us-ing only binary (CO 2 + organic solvents -ethanol or ethyl acetate) information proved to be reliable, affording satisfactory agreement between experiment and theory.Results obtained in this work may be relevant for those interested in processing curcumin using an innovative high-pressure antisolvent technique.

Table 3 . Experimental Vapor-Liquid Transition Data for the System CO 2 (1) + Ethanol (2) + Curcumin (3) at a Curcumin Concen- tration of 0.005 g cm -3 in Ethanol (CO 2 -free Basis)
a Experimental standard deviation in pressure; b BP = bubble point; DP = dew point, BP(S) and DP(S) refers to phase transitions in the presence of solid (precipitated).

Table 4 . Experimental Vapor-Liquid Transition Data for the System CO 2 (1) + Ethyl Acetate (2) + Curcumin (3) at a Curcumin Con- centration of 0.01 g cm -3 in Ethyl Acetate (CO 2 -Free Basis)
a Experimental standard deviation in pressure; b BP = bubble point; DP = dew point, BP(S) and DP(S) refers to phase transitions in the presence of solid (precipitated).

Table 5 . Experimental Transition Vapor-Liquid Data for the CO 2 (1) + Ethyl Acetate (2) + Curcumin (3) at a Curcumin Concentra- tion of 0.0025 g cm -3 in Ethyl Acetate (CO 2 -Free Basis)
a Experimental standard deviation in pressure; b BP = bubble point; DP = dew point, BP(S) and DP(S) refers to phase transitions in the presence of solid (precipitated).