Phase-equilibrium measurement of a carbon dioxide/toluene/polystyrene ternary system using laser turbidimetry

Abstract

The phase behavior of a carbon dioxide (CO₂)/toluene (Tol)/polystyrene (PS) ternary system was measured. The measurements were performed using a synthetic method combined with a reflected light intensity measurement. The bubble points were determined by visual observation, and the cloud points were determined from the change in the reflected light intensity. The phase boundaries were measured at temperatures ranging from 313.2 K to 373.2 K and CO₂ weight fractions ranging from 0.10 to 0.41. The homogeneous phase area decreased when the molecular weight of PS and/or the ratio of PS to Tol increased. It was found that the phase behavior of ternary systems, especially the liquid-liquid phase separation behavior, was altered by a change in the mutual solubility between the three components.

Introduction

There has been recent interest in the applications of supercritical fluids, particularly supercritical carbon dioxide (scCO₂), and polymers because they are safe for the environment and may reduce the use of costly fossil fuels. For example, scCO₂ can be used as a substitute for organic solvents and has been widely utilized in polymerization [1], dying [2], impregnation [2], encapsulation [3], nanoparticle synthesis [4], and so forth.

Unfortunately, it is difficult to obtain sufficient products with a system comprised of only supercritical carbon dioxide and polymers. Therefore, attempts have been made to add a solvent as a third component. In that regard, reports have been published indicating that performance can be changed more broadly with the addition of an additive. Dixon et al. [5] have succeeded in synthesizing polymer particles with a diameter of 0.1–20 μm by injecting scCO₂/polystyrene (PS)/toluene mixtures from a capillary. This method is called Gas Anti-Solvent Recrystallization (GAS), and it utilizes phase separation caused by the poor solvent effect. When scCO₂, which is a poor solvent for the polymer, is added to the polymer solution, the solvency of the solvent is reduced and the polymer is separated. There are many reported cases where liquid-liquid (LL) phase separation has occurred in this way [1,[5], [6], [7], [8]]. For example, in the supercritical polymerization process, polymerization proceeds and a LL phase separation occurs when the ratio of the polymer and monomer changes [1]. In addition, in preparation of a polymer foam in a solvent-based system using a supercritical fluid as a foaming agent, a LL phase separation (a polymer-rich and -lean phase) is a factor that affects the cell diameter of the foam [6]. Thus, it is important to understand the phase behavior of a CO₂/organic solvent/polymer system to properly design a target process.

Most of the phase equilibria of systems including polymers are described using PT phase diagrams [9,10]. Hasch et al. [9] reported the phase diagram of a CO₂/acetone/polyethylene ternary system and found that as the acetone concentration increased, the system changed from exhibiting UCST-type behavior to exhibiting U-LCST-type and LCST-type behaviors. They reasoned that the polyethylene precipitated due to an increasing acetone-acetone interaction as the solvent concentration increased. On the other hand, Matsukawa et al. [11] and Görnert et al. [12] reported Px diagrams of CO₂/organic solvent/polymer systems. The LL phase boundary (cloud point: CP) behavior with respect to concentration at constant temperature can be ascertained from Px diagrams. Matsukawa et al. [11] showed that the behavior differed depending on the concentration. When the CO₂ mass fraction was low, the bubble point (BP), at which the vapor-liquid (VL) phase separation occurs, was observed. On the other hand, when the CO₂ mass fraction was high, the LL and the vapor-liquid-liquid (VLL) phase boundary (also defined as BP) were observed. Unfortunately, there are fewer reported Px phase diagrams compared to PT phase diagrams, which do not provide sufficient information.

Methods for measuring a high-pressure LL phase equilibrium include the static method [13], circulation method [14], flow method [15], and synthetic method [10,11]. Of these, the synthetic method is the standard for measuring a system including polymers. For example, McHugh et al. [10] and Matsukawa et al. [11] employed the synthetic method for their measurements. In this method, the CO₂/organic solvent/polymer mixture is observed through a visible window of the cell with changing pressure and temperature. In many cases, the cloud point pressure is defined as the pressure at which the solution becomes so opaque that it is no longer possible to see the magnetic stirring bar in the cell. This method has drawbacks because it is affected by the operator's subjectivity.

In order to solve this problem, we considered using laser light. In the past, some researchers obtained cloud points at low pressure by measuring light intensity. Ochi et al. [16] reported the LL phase equilibrium of binary systems such as a butanol/water. By irradiating the view cell with the light of a tungsten lamp, Ochi determined cloud points by evaluating the scattered light intensity as a function of the turbidity of the solution. Nishida et al. [17] also measured cloud points by determining the decrease in light intensity at low pressure. However, there are few reports of measurements at high pressure. One such report by Zhu et al. [18] measured cloud points at high pressure using transmitted laser light intensity through two view windows, but this measurement method required a dedicated high-pressure resistant cell.

In this study, we constructed an easy cloud point measurement system based upon a synthetic method using a laser light without a dedicated cell. The cloud points were obtained by measuring the reflected light intensity. Furthermore, to deepen the knowledge of the phase behavior of CO₂/organic solvent/polymer systems, the phase behavior of a CO₂/toluene/PS ternary system was measured. The PT diagrams of this ternary system were reported by Kim et al. [19]. However, the Px diagrams and the effect of the PS molecular weight for the phase behavior was not reported. Therefore, the effect of each parameter, such as composition, temperature and polystyrene molecular weight, for the Px diagrams were examined.