Effects of Specific Interactions in Supercritical Fluid Solutions

Chapter 18

Effects of Specific Interactions in Supercritical Fluid Solutions A Chromatographic Study 1

2

2,3

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on April 4, 2016 | http://pubs.acs.org Publication Date: December 17, 1992 | doi: 10.1021/bk-1992-0514.ch018

Michael P. Ekart , Karen L. Bennett , and Charles A. Eckert 1

Department of Chemical Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801 School of Chemical Engineering and Specialty Separations Center, Georgia Institute of Technology, Atlanta, GA 30332-0100

2

The addition of a small amount of a second component, a cosolvent, to a supercritical fluid can alter phase equilibria dramatically, giving engineers the capability to fine tune the properties of the solvent over a wide range. However, this ability remains largely unexploited in engineering applications due in part to the scarcity of experimental data and lack of accurate models. We have developed a new chromatographic technique to measure these cosolvent effects rapidly. This approach was verified by comparison to literature data obtained by other methods. With this technique, cosolvent effects can be measured in a variety of systems leading to a better understanding of the special intermolecular interactions that occur in supercritical fluid solutions. Supercritical fluids (SCFs) have many special properties that make them ideal candidates for separation processes (i); however, selectivities in commonly used fluids such as carbon dioxide depend primarily on volatilities (2). One promising approach for manipulating selectivities is the addition of a small amount of cosolvent to the SCF. For example, the solubility of salicylic acid increases by more than an order of magnitude in C O 2 with 3.5 mol% methanol relative to pure C O 2 (3), but the solubility of anthracene increases only slightly (2). Nevertheless, current applications of large-scale cosolvent separations are both sparse and empirical, partly because of a lack of experimental data and accurate models. One reason that data are scarce is that conventional techniques to measure solubilities in SCFs aretime-consumingand require large amounts of pure solute. In this work, we test a new chromatographic technique to measure the effects of cosolvents in SCF solutions on solubilities. This approach is attractive because chromatographic measurements are rapid, require little solute, and inherently separate impurities from the solute. Several previous researchers have used supercritical fluid chromatography (SFC) to make thermodynamic measurements in pure SCFs. Solubility isotherms for a solute in pure solvents can be measured rapidly using SFC when an independently determined solubility point for the solute is available at the same temperature (4-6). Brown et al. (7) measured stationary phase Henry's constants for several solutes and 3

Corresponding author 0097-6156/93/0514-0228$06.00/0 © 1993 American Chemical Society

Kiran and Brennecke; Supercritical Fluid Engineering Science ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

18. EKARTETAL.

Interactions in Supercritical Fluid Solutions

229

correlated them with solute properties. This technique makes possible the measurement of solubilities from retention data where no additional solubility data exist. Shim and Johnston (8) measured the distribution coefficients for naphthalene and phenanthrene between SCF C O 2 and a bonded octadecane liquid phase in addition to solute partial molar volumes and enthalpies in the fluid phase.


Theory The retention time of a solute depends upon how it is distributed between the mobile phase (in this case, a SCF) and the stationary phase, and is characterized by *,·, the dimensionless capacity factor

where t is the retention time of solute ί and to is the retentiontimeof a solute that does not interact with the stationary phase. Assuming that equilibrium partitioning of the solute between the mobile and stationary phases is the sole mechanism of solute retention, the capacity factor is related to the equilibrium distribution of the solute between the two phases by t

*yy

t=

y/v

F

(2) where jc,- and y,- are the mole fractions of component 1 in the stationary and fluid phases respectively, W and vS are the molar volumes of the fluid and stationary phases, and V and V are the total volumes of the fluid and stationary phases. The validity of this assumption is discussed later. The fugacity of the solute in the fluid phase, ff is 7

F

s

9

(3) where is the fugacity coefficient of solute i at infinite dilution and Ρ is the pressure. Assuming Henry's law is valid in the stationary phase gives

(4) where Hi is the Henry's constant for solute i in a given stationary phase. At equilibrium, the fugacities are equal; substituting equations (3) and (4) into (2) then gives . Ψ

'

F

=

S

*./J. V V F

Pv

s

V

(5) Barde et al. (5,6) assumed that the ratio VFyS/VS depends on the column and should not vary significantly with pressure. To determine the fugacity of a solute along an



230

SUPERCRITICAL FLUID ENGINEERING SCIENCE

isotherm at a series of pressures in the fluid phase then requires knowledge of Henry's constant at that temperature. However, this technique has some limitations. Besides equilibrium partitioning, other mechanisms of solute retention must be considered, including adsorption of the solute at the interface between the mobile and stationary phases and interaction with active sites on the stationary phase support. In addition, the amount of fluid absorbed in the stationary phase can vary. These effects are pressure- and temperature-dependent, but because they are very difficult to quantify, previous investigators have ignored them when making thermodynamic measurements with chromatography. The size of the uncertainties is unknown and depends on the stationary phase, but Bartle et al. obtained good results. Our measurements of cosolvent effects eliminate the need for such an assumption by using a comparison technique that may cancel out the effects of retention mechanisms other than bulk absorption. In our work, we compare the retention of a solute at the same temperature and pressure in the same column, but with different fluids as the mobile phase. Taking the ratio of equation (5) for two different mobile phases (denoted by ' and ") gives 9

/ t

99 ~

φ;

99

/ F

φ~ _k

Η. v //

9

V

//

99

9 F

F

s

V 99

s

k, H v V v %

//

/ s

V

9 s

V

(6) Several simplifying assumptions may be made. First, because the volume of the nobile phase is much greater than the volume of the stationary phase in the capillary olumns that we used, swelling of the stationary phase has negligible effect on VF, hus it can be considered constant. Second, the ratio v W , or the total number of .noles of the stationary phase, should not depend greatly upon the mobile phase at constant temperature and pressure. 5

The third and key assumption is that Henry's constant for a solute is not altered by the changes in composition of the mobile phase. Several researchers have measured significant absorption of solvents and cosolvents into stationary phases (for example, 9-72). It has been proposed that the cosolvent also covers up active sites on the column; this would cause the solute to be retained less thus decreasing the retention time (13). The presence of cosolvent should have some effect on the Henry's constant; however, the magnitude of this effect is unknown and should depend on die stationary phase and the cosolvent Recent studies have demonstrated that the increase in solvating power of the fluid upon addition of the cosolvent is primarily responsible for the decrease in retention implying that the stationary phase environment around the solute, thus the Henry's constant, is largely unaffected by the cosolvent (14,15). Nevertheless, the assumption of a constant Henry's constant should be checked for each column; the best way to do this is to study systems where the fluid phase thermodynamics are already known. For the measurements to be meaningful, the pressure drop across the column must be small. In our experiments, we used capillary columns with estimated pressure drops ranging from 0.2 bars, at a column pressure of 90 bars, to 0.9 bars when operating at 350 bars. After making these assumptions, equation (6) becomes

v

#L_*, 99 —

φ;

F

// Τ

Κ

F V


(7)

18. EKARTETAL.

Interactions in Supercritical Fluid Solutions

231


The ratio of the molar volumes of the fluid phases should be near unity and can be estimated from an equation of state, from experimental data, or from the to and flow rate measurements. In this work, we used an equation of state to determine this ratio. Figure 1 demonstrates the meaning of the ratio of infinite dilution activity coefficients. The fugacity coefficient for a solute is shown as a function of composition up to the solubility limit in two different fluids at the same temperature and pressure. At saturation

yr _

Effects of Specific Interactions in Supercritical Fluid Solutions

Recommend Documents