Unexpected Interactions between Sol−Gel Silica Glass and Guest Molecules. Extraction of Aromatic Hydrocarbons into Polar Silica from Hydrophobic Solvents

Properties of a solute may differ greatly between a free solution and that solution confined in pores of a sol−gel glass. We studied the entry of various aromatic organic compounds from solution into the monolith of sol−gel silica immersed in this solution. Partitioning of the solute is quantified by the uptake coefficient, the ratio of its concentrations in the glass and in the surrounding solution at equilibrium. The dependence of this coefficient on the solvent gives insight into possible interactions between the solute and the silica matrix. We report the uptake of 31 compounds altogether: 18 halogen derivatives of benzene; 5 condensed (fused) aromatics; and stilbene and three substituted derivatives of it, each in both cis and trans configurations. When the solvent is hexane, the uptake coefficients are as follows: 1.0−1.9 for the halobenzenes; 3.0−4.6 for the hydrocarbons; and 3.3−4.9 for the stilbenes. When the solvent is carbon tetrachloride or dichloromethane, the uptake coefficients become 0.82−1.39 for the hydrocarbons and 0.90−1.25 for the stilbenes. The excessive uptake of organic compounds from hexane is unexpected, for it amounts to extraction of nonpolar or slightly polar solutes from a nonpolar solvent into a polar interior of silica glass. The solute−silica interactions responsible for this extraction are not of the van der Waals type. Our findings are consistent with hydrogen bonding between the aromatic π system in the solutes and the hydroxyl groups on the silica surface. Hexane cannot interact with this surface but dichloromethane and carbon tetrachloride can: they shield the hydroxyl groups from the organic solvents and thus suppress the hydrogen bonding. This explanation is supported by the emission spectra of the aromatic compound pyrene when it is dissolved in acetonitrile, dichloromethane, cyclohexyl chloride, and hexane and when it is taken up by monoliths of sol−gel silica from these four solutions. The relative intensities of the emission bands designated III and I change greatly when pyrene is taken up from hexane but remain unchanged when it is taken up from the other three solvents. Evidently, hexane does not, whereas the other three solvents do, line the silica surface and shield it from approach by pyrene molecules. Even though solute molecules are much smaller than the pores in the sol−gel glass, diffusion of these molecules into the monolith may result in an uneven partitioning at equilibrium. This fact must be taken