Studying the rifampin solubility in supercritical CO2 with/without co-solvent: Experimental data, modeling and machine learning approach

Abstract

Rifampin is an effective tuberculosis drug. This study examines drug solubility in supercritical carbon dioxide at 308 K to 338 K and 12 MPa to 30 MPa. This study investigates Rifampin solubility using ethanol as a co-solvent. The experimental data is then used to develop a solute-solvent system model using semi-empirical correlations, the Peng-Robinson equation of state, and machine learning models. A study showed that the mole fraction of rifampin without a co-solvent ranges from 0.147 × 10⁻⁵ to 1.972 × 10⁻⁵. Using ethanol as a co-solvent enhances Rifampin mole fraction, resulting in a range of 1.880 × 10⁻⁵ to 1.403 × 10⁻⁴. This study shows that ethanol can be used in this particular system as a co-solvent. The findings demonstrate that semi-empirical, equation of state-based, and machine learning methods may accurately determine Rifampin solubility. Moreover, the results indicate that the incorporation of a co-solvent enhances solubility and may facilitate the application of SCF-based micronization techniques, particularly the rapid expansion of supercritical solution (RESS) and other SCF-based methods for nanoparticle production. However, the use of a liquid co-solvent can complicate the recovery of a dry drugs powder.