Experimental measurement and modeling of the solubility of fluorinated compounds derived from dichlone in supercritical carbon dioxide

Dichlone, also known as 2,3-dichloronaphthalene-1,4-dione, is a solid organic substance employed in the field of agriculture for its fungicidal properties and as a retardant for vegetable decomposition. The bioactive properties of dichlone can be enhanced by modifying its structure, specifically through the synthesis of new derivatives achieved by replacing the functional groups within its molecular structure. Two new solid dichlone derivatives were synthesized in this work, namely 2-chloro-3-((4-fluorobenzyl)amino)naphthalene-1,4-dione (dCl-2B-F) and 2-chloro-3-((4-fluorophenethyl)amino)naphthalene-1,4-dione (dCl-3 P-F) and measured their solubility in supercritical carbon dioxide at (313, 323, and 333) K and pressures between (9. to 32) MPa. The results indicated that solubility ranged between 30.5 and 47.9 µmol of solute/mol of CO₂ for dCl-2B-F, and from 2.2 to 243.5 µmol of solute/mol of CO₂ for dCl-3 P-F. The solubility data of dichlone and its synthesized derivatives (dCl-2B-F, dCl-3 P-F, 2-chloro-3-((4-chlorobenzyl)amino)naphthalene-1,4-dione (dCl-2B-Cl), 2-chloro-3-((4-chlorophenethyl)amino)naphthalene-1,4-dione (dCl-3 P-Cl), 2-(benzylamino)-3-chloronaphthalene-1,4-dione (dCl-2B) and 2-chloro-3-(phenethylamino)naphthalene-1,4-dione (dCl-3 P)) was compared using the density-based correlation of Chrastil and the Statistical Associating Fluid Theory of Variable Range Mie-potential (SAFT-VR Mie) equation of state (EoS), to better comprehend the effects of the structural differences on the solubility. As a result, for the Chrastil model, a root mean square deviation (rmsd) of 3% was obtained for dCl-2B-F and 16% for dCl-3 P-F, whereas for the SAFT-VR Mie equation, it averaged 24% for dCl-2B-F and 28% for dCl-3 P-F. It was found that the solubility of the homologous compounds, differing only in one methylene group, increased with solute size (-2B derivatives were less soluble in CO₂ than the −3 P ones), contrary to the expected trend, which could be attributed to the increased probability of ring-to-ring interactions as the chain length connecting the rings decreases. This demonstrates that geometric factors, along with the pressure and temperature, affect the behavior of the solubility and these should be accurately represented in the predictive models.